1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.623 2010/06/06 00:27:52 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.
43 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
44 $data_path, and $header_path variables; fixed documentation bugs and
47 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
48 exipick to access non-standard spools, including the "frozen" queue
51 NM/01 Bugzilla 965: Support mysql stored procedures.
52 Patch from Alain Williams
54 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
56 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
57 Patch from Andreas Metzler
59 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
60 Patch from Kirill Miazine
62 NM/05 Bugzilla 671: Added umask to procmail example.
64 JJ/03 installed exipick 20100323.0, fixing doc bug
66 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
67 directory. Notification and patch from Dan Rosenberg.
69 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
71 TK/02 Improve log output when DKIM signing operation fails.
73 MH/01 Treat the transport option dkim_domain as a colon separated
74 list, not as a single string, and sign the message with each element,
75 omitting multiple occurences of the same signer.
77 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
78 Bugzilla 985, 986. Patch by Simon Arlott
80 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
83 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
84 MBX locking. Notification from Dan Rosenberg.
90 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
92 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
94 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
96 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
98 NM/04 Bugzilla 915: Flag broken perl installation during build.
104 TK/01 Added patch by Johannes Berg that expands the main option
105 "spamd_address" if it starts with a dollar sign.
107 TK/02 Write list of recipients to X-Envelope-Sender header when building
108 the mbox-format spool file for content scanning (suggested by Jakob
111 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
112 (http://www.dcc-servers.net/) support via dccifd. Activated by
113 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
115 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
116 by Mark Daniel Reidel <mr@df.eu>.
118 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
119 When building exim an external PCRE library is now needed -
120 PCRE is a system library on the majority of modern systems.
121 See entry on PCRE_LIBS in EDITME file.
123 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
124 conversation. Added nologin parameter to request.
125 Patch contributed by Kirill Miazine
127 TF/01 Do not log submission mode rewrites if they do not change the address.
129 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
131 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
132 log files in place. Contributed by Roberto Lima
134 NM/04 Bugzilla 667: close socket used by dovecot authenticator
136 TF/03 Bugzilla 615: When checking the local_parts router precondition
137 after a local_part_suffix or local_part_prefix option, Exim now
138 does not use the address's named list lookup cache, since this
139 contains cached lookups for the whole local part.
141 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
142 Robert Millan. Documentation is in experimental-spec.txt
144 TF/04 Bugzilla 668: Fix parallel build (make -j).
146 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
148 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
149 Patch provided by Jan Srzednicki
151 TF/05 Leading white space used to be stripped from $spam_report which
152 wrecked the formatting. Now it is preserved.
154 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
155 that they are available at delivery time.
157 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
159 TF/08 TLS error reporting now respects the incoming_interface and
160 incoming_port log selectors.
162 TF/09 Produce a more useful error message if an SMTP transport's hosts
163 setting expands to an empty string.
165 NM/06 Bugzilla 744: EXPN did not work under TLS.
166 Patch provided by Phil Pennock
168 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
169 Patch provided by Richard Godbee
171 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
172 acl_smtp_notquit, added index entry.
174 NM/09 Bugzilla 787: Potential buffer overflow in string_format
175 Patch provided by Eugene Bujak
177 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
178 Patch provided by Maxim Dounin
180 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
181 Patch provided by Phil Pennock
183 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
185 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
186 Patch provided by Brad "anomie" Jorsch
188 NM/14 Bugzilla 622: Added timeout setting to transport filter.
189 Patch provided by Dean Brooks
191 TK/05 Add native DKIM support (does not depend on external libraries).
193 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
194 Patch provided by Graeme Fowler
196 NM/16 Bugzilla 851: Documentation example syntax fix.
198 NM/17 Changed NOTICE file to remove references to embedded PCRE.
200 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
202 NM/19 Bugzilla 745: TLS version reporting
203 Patch provided by Phil Pennock
205 NM/20 Bugzilla 167: bool: condition support
206 Patch provided by Phil Pennock
208 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
209 Patch provided by Phil Pennock
211 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
212 Patch provided by Brad "anomie" Jorsch
214 NM/23 Bugzilla 687: Fix misparses in eximstats
215 Patch provided by Heiko Schlittermann
217 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
218 Patch provided by Heiko Schlittermann
220 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
221 plus update to original patch.
223 NM/26 Bugzilla 799: Documentation correction for ratelimit
225 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
226 Patch provided by David Brownlee
228 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
230 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
232 NM/30 Bugzilla 888: TLS documentation bugfixes
234 NM/31 Bugzilla 896: Dovecot buffer overrun fix
236 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
237 Unlike the original bugzilla I have changed all shell scripts in src tree
239 NM/33 Bugzilla 898: Transport filter timeout fix
240 Patch by Todd Rinaldo
242 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
243 Patch by Serge Demonchaux
245 NM/35 Bugzilla 39: Base64 decode bug fixes
246 Patch by Jakob Hirsch
248 NM/36 Bugzilla 909: Correct connect() call in dcc code
250 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
252 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
254 NM/39 Bugzilla 911: Fixed MakeLinks build script
260 TK/01 Add preliminary DKIM support. Currently requires a forked version of
261 ALT-N's libdkim that I have put here:
262 http://duncanthrax.net/exim-experimental/
264 Note to Michael Haardt: I had to rename some vars in sieve.c. They
265 were called 'true' and it seems that C99 defines that as a reserved
266 keyword to be used with 'bool' variable types. That means you could
267 not include C99-style headers which use bools without triggering
268 build errors in sieve.c.
270 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
271 as mailq or other aliases. Changed the --help handling significantly
272 to do whats expected. exim_usage() emits usage/help information.
274 SC/01 Added the -bylocaldomain option to eximstats.
276 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
278 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
280 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
286 PH/01 Another patch from the Sieve maintainer.
288 PH/02 When an IPv6 address is converted to a string for single-key lookup
289 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
290 dots are used instead of colons so that keys in lsearch files need not
291 contain colons. This was done some time before quoting was made available
292 in lsearch files. However, iplsearch files do require colons in IPv6 keys
293 (notated using the quote facility) so as to distinguish them from IPv4
294 keys. This meant that lookups for IP addresses in host lists did not work
295 for iplsearch lookups.
297 This has been fixed by arranging for IPv6 addresses to be expressed with
298 colons if the lookup type is iplsearch. This is not incompatible, because
299 previously such lookups could never work.
301 The situation is now rather anomolous, since one *can* have colons in
302 ordinary lsearch keys. However, making the change in all cases is
303 incompatible and would probably break a number of configurations.
305 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
308 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
309 conversion specification without a maximum field width, thereby enabling
310 a rogue spamd server to cause a buffer overflow. While nobody in their
311 right mind would setup Exim to query an untrusted spamd server, an
312 attacker that gains access to a server running spamd could potentially
313 exploit this vulnerability to run arbitrary code as the Exim user.
315 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
316 $primary_hostname instead of what libspf2 thinks the hosts name is.
318 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
319 a directory entry by the name of the lookup key. Previously, if a
320 symlink pointed to a non-existing file or a file in a directory that
321 Exim lacked permissions to read, a lookup for a key matching that
322 symlink would fail. Now it is enough that a matching directory entry
323 exists, symlink or not. (Bugzilla 503.)
325 PH/03 The body_linecount and body_zerocount variables are now exported in the
328 PH/04 Added the $dnslist_matched variable.
330 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
331 This means they are set thereafter only if the connection becomes
334 PH/06 Added the client_condition to authenticators so that some can be skipped
335 by clients under certain conditions.
337 PH/07 The error message for a badly-placed control=no_multiline_responses left
338 "_responses" off the end of the name.
340 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
342 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
343 (without spaces) instead of just copying the configuration text.
345 PH/10 Added the /noupdate option to the ratelimit ACL condition.
347 PH/11 Added $max_received_linelength.
349 PH/12 Added +ignore_defer and +include_defer to host lists.
351 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
352 way in which PCRE > 7.0 is built.
354 PH/14 Implemented queue_only_load_latch.
356 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
357 MAIL command. The effect was to mangle the value on 64-bit systems.
359 PH/16 Another patch from the Sieve maintainer.
361 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
363 PH/18 If a system quota error occurred while trying to create the file for
364 a maildir delivery, the message "Mailbox is full" was not appended to the
365 bounce if the delivery eventually timed out. Change 4.67/27 below applied
366 only to a quota excession during the actual writing of the file.
368 PH/19 It seems that peer DN values may contain newlines (and other non-printing
369 characters?) which causes problems in log lines. The DN values are now
370 passed through string_printing() before being added to log lines.
372 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
373 and InterBase are left for another time.)
375 PH/21 Added message_body_newlines option.
377 PH/22 Guard against possible overflow in moan_check_errorcopy().
379 PH/23 POSIX allows open() to be a macro; guard against that.
381 PH/24 If the recipient of an error message contained an @ in the local part
382 (suitably quoted, of course), incorrect values were put in $domain and
383 $local_part during the evaluation of errors_copy.
389 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
390 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
393 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
394 issue a MAIL command.
396 PH/02 In an ACL statement such as
398 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
400 if a client was not listed at all, or was listed with a value other than
401 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
402 the condition was not true (as it should be), so access was not denied.
403 The bug was that the ! inversion was incorrectly passed on to the second
404 item. This has been fixed.
406 PH/03 Added additional dnslists conditions == and =& which are different from
407 = and & when the dns lookup returns more than one IP address.
409 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
410 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
412 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
413 FSYNC, which compiles an option called disable_fsync that allows for
414 bypassing fsync(). The documentation is heavily laced with warnings.
416 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
418 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
419 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
420 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
421 including adding "make clean"; (3) Added -fPIC when compiling the test
422 dynamically loaded module, to get rid of a warning.
424 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
425 message fails, move_frozen_messages = true and ignore_bounce_errors_after
426 = 0s. The bug is otherwise harmless.
428 PH/07 There was a bug in the dovecot authenticator such that the value of
429 $auth1 could be overwritten, and so not correctly preserved, after a
430 successful authentication. This usually meant that the value preserved by
431 the server_setid option was incorrect.
433 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
435 PH/09 Installed PCRE release 7.0.
437 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
438 run for batched SMTP input. It is now run at the start of every message
439 in the batch. While fixing this I discovered that the process information
440 (output by running exiwhat) was not always getting set for -bs and -bS
441 input. This is fixed, and it now also says "batched" for BSMTP.
443 PH/11 Added control=no_pipelining.
445 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
446 patch, slightly modified), and move the expansion of helo_data till after
447 the connection is made in the smtp transport (so it can use these
450 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
452 PH/14 Added log_selector = +pid.
454 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
456 PH/16 Add ${if forany and ${if forall.
458 PH/17 Added dsn_from option to vary the From: line in DSNs.
460 PH/18 Flush SMTP output before performing a callout, unless control =
461 no_callout_flush is set.
463 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
464 was true (the default) a successful delivery failed to delete the retry
465 item, thus causing premature timeout of the address. The bug is now
468 PH/20 Added hosts_avoid_pipelining to the smtp transport.
470 PH/21 Long custom messages for fakedefer and fakereject are now split up
471 into multiline reponses in the same way that messages for "deny" and
472 other ACL rejections are.
474 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
475 with slight modification.
477 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
478 draft, changing the syntax and factoring some duplicate code".
480 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
481 for deliveries of the second and subsequent messages over the same SMTP
484 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
485 ${reduce, with only minor "tidies".
487 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
489 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
490 expansion side effects.
492 PH/27 When a message times out after an over-quota error from an Exim-imposed
493 quota, the bounce message says "mailbox is full". This message was not
494 being given when it was a system quota that was exceeded. It now should
497 MH/03 Made $recipients available in local_scan(). local_scan() already has
498 better access to the recipient list through recipients_list[], but
499 $recipients can be useful in postmaster-provided expansion strings.
501 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
502 in the case of a MAIL command with additional options following the
503 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
504 were accidentally chopped off.
506 PH/29 SMTP synchronization checks are implemented when a command is read -
507 there is a check that no more input is waiting when there shouldn't be
508 any. However, for some commands, a delay in an ACL can mean that it is
509 some time before the response is written. In this time, more input might
510 arrive, invalidly. So now there are extra checks after an ACL has run for
511 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
512 pipelining has not been advertised.
514 PH/30 MH's patch to allow iscntrl() characters to be list separators.
516 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
517 returned in the SMTP response when smtp_return_error_details was false.
520 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
521 instead of the C I/O that was originally supplied, because problems were
524 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
525 Exim which did not show up earlier: it was assuming that a call to
526 SSL_CTX_set_info_callback() might give an error value. In fact, there is
527 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
528 was a macro that became an assignment, so it seemed to work. This has
529 changed to a proper function call with a void return, hence the compile
530 error. Exim's code has been fixed.
532 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
535 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
537 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
539 PH/37 If a message is not accepted after it has had an id assigned (e.g.
540 because it turns out to be too big or there is a timeout) there is no
541 "Completed" line in the log. When some messages of this type were
542 selected by exigrep, they were listed as "not completed". Others were
543 picked up by some special patterns. I have improved the selection
544 criteria to be more general.
546 PH/38 The host_find_failed option in the manualroute router can now be set
547 to "ignore", to completely ignore a host whose IP address cannot be
548 found. If all hosts are ignored, the behaviour is controlled by the new
549 host_all_ignored option.
551 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
552 homing or because of multiple MX records with /mx) generated more than
553 one IP address, and the following item turned out to be the local host,
554 all the secondary addresses of the first item were incorrectly removed
555 from the list, along with the local host and any following hosts (which
556 is what is supposed to happen).
558 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
559 whoever called Exim into the -H file. In the case of the daemon it was
560 behaving confusingly. When first started, it used values for whoever
561 started the daemon, but after a SIGHUP it used the Exim user (because it
562 calls itself on a restart). I have changed the code so that it now always
565 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
566 message are rejected with the same error (e.g. no authentication or bad
567 sender address), and a DATA command is nevertheless sent (as can happen
568 with PIPELINING or a stupid MUA), the error message that was given to the
569 RCPT commands is included in the rejection of the DATA command. This is
570 intended to be helpful for MUAs that show only the final error to their
573 PH/42 Another patch from the Sieve maintainer.
575 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
576 Eximstats - Fixed some broken HTML links and added missing column headers
578 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
579 columns for Rejects, Temp Rejects, Ham, and Spam rows.
581 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
583 PH/43 Yet another patch from the Sieve maintainer.
585 PH/44 I found a way to check for a TCP/IP connection going away before sending
586 the response to the final '.' that terminates a message, but only in the
587 case where the client has not sent further data following the '.'
588 (unfortunately, this is allowed). However, in many cases there won't be
589 any further data because there won't be any more messages to send. A call
590 to select() can be used: if it shows that the input is "ready", there is
591 either input waiting, or the socket has been closed. An attempt to read
592 the next input character can distinguish the two cases. Previously, Exim
593 would have sent an OK response which the client would never have see.
594 This could lead to message repetition. This fix should cure that, at
595 least in a lot of common cases.
597 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
598 advertised in response to EHLO.
604 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
605 fixed by 4.65/MH/01 (is this a record?) are fixed:
607 (i) An empty string was always treated as zero by the numeric comparison
608 operators. This behaviour has been restored.
610 (ii) It is documented that the numeric comparison operators always treat
611 their arguments as decimal numbers. This was broken in that numbers
612 starting with 0 were being interpreted as octal.
614 While fixing these problems I realized that there was another issue that
615 hadn't been noticed. Values of message_size_limit (both the global option
616 and the transport option) were treated as octal if they started with 0.
617 The documentation was vague. These values are now always treated as
618 decimal, and I will make that clear in the documentation.
624 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
625 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
628 MH/01 Don't check that the operands of numeric comparison operators are
629 integers when their expansion is in "skipping" mode (fixes bug
630 introduced by 4.64-PH/07).
632 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
633 child addresses, Exim now panics and dies. Previously, because the count
634 is held in a short int, deliveries were likely to be lost. As such a
635 large number of recipients for a single message is ridiculous
636 (performance will be very, very poor), I have chosen to impose a limit
637 rather than extend the field.
643 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
644 leftover -K file (the existence of which was triggered by #402).
645 While we were at it, introduced process PID as part of the -K
646 filename. This should rule out race conditions when creating
649 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
650 processing considerably. Previous code took too long for large mails,
651 triggering a timeout which in turn triggers #401.
653 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
654 in the DK code in transports.c. sendfile() is not really portable,
655 hence the _LINUX specificness.
657 TF/01 In the add_headers option to the mail command in an Exim filter,
658 there was a bug that Exim would claim a syntax error in any
659 header after the first one which had an odd number of characters
662 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
663 callout verification, Exim cached a "reject" for the entire domain. This
664 is correct for most verifications, but it is not correct for a recipient
665 verification with use_sender or use_postmaster set, because in that case
666 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
667 case of MAIL FROM:<> rejection from other early rejections (e.g.
668 rejection of HELO). When verifying a recipient using a non-null MAIL
669 address, the cache is ignored if it shows MAIL FROM:<> rejection.
670 Whatever the result of the callout, the value of the domain cache is
671 left unchanged (for any other kind of callout, getting as far as trying
672 RCPT means that the domain itself is ok).
674 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
677 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
678 manifest itself as EPIPE rather than ECONNECT. When tidying away a
679 session, the daemon ignores ECONNECT errors and logs others; it now
680 ignores EPIPE as well.
682 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
683 (quoted-printable decoding).
685 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
686 later the small subsequent patch to fix an introduced bug.
688 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
690 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
692 PH/08 An error is now given if message_size_limit is specified negative.
694 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
695 to be given (somewhat) arbitrary names.
697 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
700 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
701 miscellaneous code fixes
703 PH/10 Added the log_reject_target ACL modifier to specify where to log
706 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
707 hostname. This is wrong, because it relates to the incoming message (and
708 probably the interface on which it is arriving) and not to the outgoing
709 callout (which could be using a different interface). This has been
710 changed to use the value of the helo_data option from the smtp transport
711 instead - this is what is used when a message is actually being sent. If
712 there is no remote transport (possible with a router that sets up host
713 addresses), $smtp_active_hostname is used.
715 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
716 tweaks were necessary in order to get it to work (see also 21 below):
717 (a) The code assumed that strncpy() returns a negative number on buffer
718 overflow, which isn't the case. Replaced with Exim's string_format()
720 (b) There were several signed/unsigned issues. I just did the minimum
721 hacking in of casts. There is scope for a larger refactoring.
722 (c) The code used strcasecmp() which is not a standard C function.
723 Replaced with Exim's strcmpic() function.
724 (d) The code set only $1; it now sets $auth1 as well.
725 (e) A simple test gave the error "authentication client didn't specify
726 service in request". It would seem that Dovecot has changed its
727 interface. Fortunately there's a specification; I followed it and
728 changed what the client sends and it appears to be working now.
730 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
733 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
734 address A is aliased to B and C, where B exists and C does not. Without
735 -v the output is "A verified" because verification stops after a
736 successful redirection if more than one address is generated. However,
737 with -v the child addresses are also verified. Exim was outputting "A
738 failed to verify" and then showing the successful verification for C,
739 with its parentage. It now outputs "B failed to verify", showing B's
740 parentage before showing the successful verification of C.
742 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
743 look up a TXT record in a specific list after matching in a combined
746 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
747 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
748 they consult the DNS. I had assumed they would set it the way they
749 wanted; and indeed my experiments on Linux seem to show that in some
750 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
751 To be on the safe side, however, I have now made the interface to
752 host_find_byname() similar to host_find_bydns(), with an argument
753 containing the DNS resolver options. The host_find_byname() function now
754 sets these options at its start, just as host_find_bydns() does. The smtp
755 transport options dns_qualify_single and dns_search_parents are passed to
756 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
757 of host_find_byname() use the default settings of RES_DEFNAMES
758 (qualify_single) but not RES_DNSRCH (search_parents).
760 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
761 spool_read_header() do less string testing, by means of a preliminary
762 switch on the second character of optional "-foo" lines. (This is
763 overdue, caused by the large number of possibilities that now exist.
764 Originally there were few.) While I was there, I also converted the
765 str(n)cmp tests so they don't re-test the leading "-" and the first
766 character, in the hope this might squeeze out yet more improvement.
768 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
769 flag allowing group syntax was set by the header_syntax check but not
770 turned off, possible causing trouble later; (2) The flag was not being
771 set at all for the header_verify test, causing "group"-style headers to
772 be rejected. I have now set it in this case, and also caused header_
773 verify to ignore an empty address taken from a group. While doing this, I
774 came across some other cases where the code for allowing group syntax
775 while scanning a header line wasn't quite right (mostly, not resetting
776 the flag correctly in the right place). These bugs could have caused
777 trouble for malformed header lines. I hope it is now all correct.
779 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
780 with the "reply" argument non-NULL. The code, however (which originally
781 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
782 but it didn't always do it. This confused somebody who was copying the
783 code for some other use. I have removed all the tests.
785 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
786 feature that was used to support insecure browsers during the U.S. crypto
787 embargo. It requires special client support, and Exim is probably the
788 only MTA that supported it -- and would never use it because real RSA is
789 always available. This code has been removed, because it had the bad
790 effect of slowing Exim down by computing (never used) parameters for the
791 RSA_EXPORT functionality.
793 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
794 authenticator to fail if there's a tab character in the incoming data
795 (there should never be unless someone is messing about, as it's supposed
796 to be base64-encoded). Also added, on Timo's advice, the "secured" option
797 if the connection is using TLS or if the remote IP is the same as the
798 local IP, and the "valid-client-cert option" if a client certificate has
801 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
802 authenticators. This can be used for authorization after authentication
803 succeeds. (In the case of plaintext, it servers for both authentication
806 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
807 if any retry times were supplied.
809 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
810 connection, where there is no client IP address to check. In this
811 situation, the verify now always succeeds.
813 PH/25 Applied John Jetmore's -Mset patch.
815 PH/26 Added -bem to be like -Mset, but loading a message from a file.
817 PH/27 In a string expansion for a processed (not raw) header when multiple
818 headers of the same name were present, leading whitespace was being
819 removed from all of them, but trailing whitespace was being removed only
820 from the last one. Now trailing whitespace is removed from each header
821 before concatenation. Completely empty headers in a concatenation (as
824 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
825 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
827 PH/29 [Removed. This was a change that I later backed out, and forgot to
828 correct the ChangeLog entry (that I had efficiently created) before
829 committing the later change.]
831 PH/30 Exim was sometimes attempting to deliver messages that had suffered
832 address errors (4xx response to RCPT) over the same connection as other
833 messages routed to the same hosts. Such deliveries are always "forced",
834 so retry times are not inspected. This resulted in far too many retries
835 for the affected addresses. The effect occurred only when there were more
836 hosts than the hosts_max_try setting in the smtp transport when it had
837 the 4xx errors. Those hosts that it had tried were not added to the list
838 of hosts for which the message was waiting, so if all were tried, there
839 was no problem. Two fixes have been applied:
841 (i) If there are any address or message errors in an SMTP delivery, none
842 of the hosts (tried or untried) are now added to the list of hosts
843 for which the message is waiting, so the message should not be a
844 candidate for sending over the same connection that was used for a
845 successful delivery of some other message. This seems entirely
846 reasonable: after all the message is NOT "waiting for some host".
847 This is so "obvious" that I'm not sure why it wasn't done
848 previously. Hope I haven't missed anything, but it can't do any
849 harm, as the worst effect is to miss an optimization.
851 (ii) If, despite (i), such a delivery is accidentally attempted, the
852 routing retry time is respected, so at least it doesn't keep
853 hammering the server.
855 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
856 in ${readsocket because some servers need this prod.
858 PH/32 Added some extra debug output when updating a wait-xxx database.
860 PH/33 The hint "could be header name not terminated by colon", which has been
861 given for certain expansion errors for a long time, was not being given
862 for the ${if def:h_colon_omitted{... case.
864 PH/34 The spec says: "With one important exception, whenever a domain list is
865 being scanned, $domain contains the subject domain." There was at least
866 one case where this was not true.
868 PH/35 The error "getsockname() failed: connection reset by peer" was being
869 written to the panic log as well as the main log, but it isn't really
870 panic-worthy as it just means the connection died rather early on. I have
871 removed the panic log writing for the ECONNRESET error when getsockname()
874 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
875 runs only) independently of the message's sender address. This meant
876 that, if the 4xx error was in fact related to the sender, a different
877 message to the same recipient with a different sender could confuse
878 things. In particualar, this can happen when sending to a greylisting
879 server, but other circumstances could also provoke similar problems.
880 I have changed the default so that the retry time for these errors is now
881 based a combination of the sender and recipient addresses. This change
882 can be overridden by setting address_retry_include_sender=false in the
885 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
886 remote server are returned as part of bounce messages. This was not
887 happening for LMTP over a pipe (the lmtp transport), but now it is the
888 same for both kinds of LMTP.
890 PH/38 Despite being documented as not happening, Exim was rewriting addresses
891 in header lines that were in fact CNAMEs. This is no longer the case.
893 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
894 and queue runs started by the daemon processed all messages. This has
895 been fixed so that -R and -S can now usefully be given with -q<time>.
897 PH/40 Import PCRE release 6.7 (fixes some bugs).
899 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
901 PH/42 Give an error if -q is specified more than once.
903 PH/43 Renamed the variables $interface_address and $interface_port as
904 $received_ip_address and $received_port, to make it clear that these
905 values apply to message reception, and not to the outgoing interface when
906 a message is delivered. (The old names remain recognized, of course.)
908 PH/44 There was no timeout on the connect() call when using a Unix domain
909 socket in the ${readsocket expansion. There now is.
911 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
912 be meaningful with "accept".
914 SC/01 Eximstats V1.43
915 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
917 SC/02 Eximstats V1.44
918 Use a glob alias rather than an array ref in the generated
919 parser. This improves both readability and performance.
921 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
922 Collect SpamAssassin and rejection statistics.
923 Don't display local sender or destination tables unless
924 there is data to show.
925 Added average volumes into the top table text output.
927 SC/04 Eximstats V1.46
928 Collect data on the number of addresses (recipients)
929 as well as the number of messages.
931 SC/05 Eximstats V1.47
932 Added 'Message too big' to the list of mail rejection
933 reasons (thanks to Marco Gaiarin).
935 SC/06 Eximstats V1.48
936 Mainlog lines which have GMT offsets and are too short to
937 have a flag are now skipped.
939 SC/07 Eximstats V1.49 (Alain Williams)
940 Added the -emptyok flag.
942 SC/08 Eximstats V1.50
943 Fixes for obtaining the IP address from reject messages.
945 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
946 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
947 whitesspace changes from 4.64-PH/27
949 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
952 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
953 are found, allow negative numbers in numeric criteria)
955 JJ/06 exipick.20061117.2, added new $message_body_missing variable
957 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
958 to match changes made in 4.64-PH/43
960 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
962 PH/47 Put in an explicit test for a DNS lookup of an address record where the
963 "domain" is actually an IP address, and force a failure. This locks out
964 those revolvers/nameservers that support "A-for-A" lookups, in
965 contravention of the specifications.
967 PH/48 When a host name was looked up from an IP address, and the subsequent
968 forward lookup of the name timed out, the host name was left in
969 $sender_host_name, contrary to the specification.
971 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
972 restricted to single-key lookups, Exim was not diagnosing an error if
973 * or *@ was used with a query-style lookup.
975 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
977 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
978 long ago, but noone interested enough thought of it. Let's just say that
979 the "1.1" means that there are some new functions that weren't there at
980 some point in the past.
982 PH/51 Error processing for expansion failure of helo_data from an smtp
983 transport during callout processing was broken.
985 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
986 tested/used via the -bh/-bhc/-bs options.
988 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
989 bug, fixed in subsequent PCRE releases).
991 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
992 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
994 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1000 SC/01 Use a glob alias rather than an array ref in eximstats generated
1001 parser. This improves both readability and performance.
1003 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1004 Don't display local sender or destination tables in eximstats unless
1005 there is data to show.
1006 Added average volumes into the eximstats top table text output.
1008 SC/03 Collect data on the number of addresses (recipients) as well
1009 as the number of messages in eximstats.
1011 TF/01 Correct an error in the documentation for the redirect router. Exim
1012 does (usually) call initgroups() when daemonizing.
1014 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1015 with consistent privilege compared to when running as a daemon.
1017 TF/03 Note in the spec that $authenticated_id is not set for local
1018 submissions from trusted users.
1020 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1021 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1023 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1024 by adding some example configuration directives to the default
1025 configuration file. A little bit of work is required to uncomment the
1026 directives and define how usernames and passwords are checked, but
1027 there is now a framework to start from.
1029 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1030 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1031 without this. I don't know how relevant this is to other LDAP libraries.
1033 PH/02 Add the verb name to the "unknown ACL verb" error.
1035 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1037 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1039 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1040 directory not expanded when it should be if an expanded home directory
1041 was set for the address (which is overridden by the transport).
1043 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1046 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1047 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1048 because it is too late at that time, and has no effect.
1050 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1051 security issue with \' (bugzilla #107). I could not use the
1052 PQescapeStringConn() function, because it needs a PGconn value as one of
1055 PH/08 When testing addresses using -bt, indicate those final addresses that
1056 are duplicates that would not cause an additional delivery. At least one
1057 person was confused, thinking that -bt output corresponded to deliveries.
1058 (Suppressing duplicates isn't a good idea as you lose the information
1059 about possibly different redirections that led to the duplicates.)
1061 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1062 systems where poll() doesn't work, in particular OS X.
1064 PH/10 Added more information to debugging output for retry time not reached.
1066 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1067 operations in malware.c.
1069 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1072 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1073 syslog_duplication set false, log lines that would normally be written
1074 both the the main log and to the reject log were not written to syslog at
1077 PH/14 In the default configuration, change the use of "message" in ACL warn
1078 statements to "add_header".
1080 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1081 not followed by a command (e.g. "seen endif").
1083 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1084 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1087 PH/17 Added extra conditions to the default value of delay_warning_condition
1091 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1092 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1093 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1096 The Auto-Submitted: and various List- headers are standardised, whereas I
1097 don't think Precedence: ever was.
1099 PH/18 Refactored debugging code in route_finduser() to show more information,
1100 in particular, the error code if getpwnam() issues one.
1102 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1103 This is apparently needed in addition to the PH/07 change above to avoid
1104 any possible encoding problems.
1106 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1107 but not after initializing Perl.
1109 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1110 output them only if debugging. By default they are written stderr,
1111 apparently, which is not desirable.
1113 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1116 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1119 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1121 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1122 authenticated or an ident call has been made. Suppress the default
1123 values for $authenticated_id and $authenticated_sender (but permit -oMai
1124 and -oMas) when testing with -bh.
1126 PH/24 Re-jigged the order of the tests in the default configuration so that the
1127 tests for valid domains and recipients precede the DNS black list and CSA
1128 tests, on the grounds that those ones are more expensive.
1130 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1131 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1132 This bug exists in every version of Exim that I still have, right back to
1135 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1136 However, an attempt to turn on case-sensitivity in a regex key by
1137 including (?-i) didn't work because the subject string was already
1138 lowercased, and the effects were non-intuitive. It turns out that a
1139 one-line patch can be used to allow (?-i) to work as expected.
1145 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1146 other effects) broke the use of negated acl sub-conditions.
1148 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1151 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1152 "Deny" causes Exim to reject the incoming connection with a 554 error.
1153 Unfortunately, if there is a major crisis, such as a disk failure,
1154 tcp-wrappers gives "deny", whereas what one would like would be some
1155 kind of temporary error. A kludge has been added to help with this.
1156 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1157 554 error is used if errno is still zero or contains ENOENT (which occurs
1158 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1161 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1163 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1164 errors. Otherwise a message that provokes a temporary error (when other
1165 messages do not) can cause a whole host to time out.
1167 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1168 addresses were routed directly to files or pipes from a redirect router.
1169 File deliveries just didn't batch; pipe deliveries might have suffered
1172 PH/06 A failure to get a lock for a hints database would erroneously always say
1173 "Failed to get write lock", even when it was really a read lock.
1175 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1176 of 0600. This has been changed to use the value of the lockfile_mode
1177 option (which defaults to 0600).
1179 PH/08 Applied small patch from the Sieve maintainer.
1181 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1182 folder from quota calculations, a direct delivery into this folder messed
1183 up the contents of the maildirsize file. This was because the regex was
1184 used only to exclude .Trash (or whatever) when the size of the mailbox
1185 was calculated. There was no check that a delivery was happening into an
1186 excluded directory. This bug has been fixed by ignoring all quota
1187 processing for deliveries into excluded directories.
1189 PH/10 Added the maildirfolder_create_regex option to appendfile.
1195 PH/01 The code for finding all the local interface addresses on a FreeBSD
1196 system running IPv6 was broken. This may well have applied to all BSD
1197 systems, as well as to others that have similar system calls. The broken
1198 code found IPv4 interfaces correctly, but gave incorrect values for the
1199 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1200 that it would not match correctly against @[] and not recognize the IPv6
1203 PH/02 The ipliteral router was not recognizing addresses of the form user@
1204 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1206 PH/03 Added disable_ipv6.
1208 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1209 decoded form, because it is most often used to construct To: headers
1210 lines in autoreplies, and the decoded form may well be syntactically
1211 invalid. However, $reply_address has leading white space removed, and all
1212 newlines turned into spaces so that the autoreply transport does not
1215 PH/05 If group was specified without a user on a router, and no group or user
1216 was specified on a transport, the group from the router was ignored.
1218 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1219 for visible compile-time settings that can be used to change these
1220 numbers, for those that want even more. Backwards compatibility with old
1221 spool files has been maintained. However, going back to a previous Exim
1222 release will lost any variables that are in spool files.
1224 PH/07 Two small changes when running in the test harness: increase delay when
1225 passing a TCP/IP connection to a new process, in case the original
1226 process has to generate a bounce, and remove special handling of
1227 127.0.0.2 (sic), which is no longer necessary.
1229 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1230 be the same on different OS.
1232 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1235 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1236 whether --show-vars was specified or not
1238 JJ/02 exipick: Added support for new ACL variable spool format introduced
1241 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1242 syntactically invalid From: or Reply-to: line, and a filter used this to
1243 generate an autoreply, and therefore failed to obtain an address for the
1244 autoreply, Exim could try to deliver to a non-existent relative file
1245 name, causing unrelated and misleading errors. What now happens is that
1246 it logs this as a hard delivery error, but does not attempt to create a
1249 PH/11 The exinext utility has a -C option for testing purposes, but although
1250 the given file was scanned by exinext itself; it wasn't being passed on
1251 when Exim was called.
1253 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1254 an end-of-file indication when reading a command response.
1256 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1257 compiled. In many other places in Exim, IPv6 addresses are always
1258 recognized, so I have changed this. It also means that IPv4 domain
1259 literals of the form [IPV4:n.n.n.n] are now always recognized.
1261 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1262 used if the router is not running as root, for example, when verifying at
1263 ACL time, or when using -bh. The debugging output from this situation was
1264 non-existent - all you got was a failure to exec. I have made two
1267 (a) Failures to set uid/gid, the current directory, or a process leader
1268 in a subprocess such as that created by queryprogram now generate
1269 suitable debugging ouput when -d is set.
1271 (b) The queryprogram router detects when it is not running as root,
1272 outputs suitable debugging information if -d is set, and then runs
1273 the subprocess without attempting to change uid/gid.
1275 PH/15 Minor change to Makefile for building test_host (undocumented testing
1278 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1279 additional section of a DNS packet that returns MX or SRV records.
1280 Instead, it always explicitly searches for A/AAAA records. This avoids
1281 major problems that occur when a DNS server includes only records of one
1282 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1283 fixed another bug: if SRV records were looked up and the corresponding
1284 address records were *not* found in the additional section, the port
1285 values from the SRV records were lost.
1287 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1288 using the correct key (the original address) when searching the retry
1289 rules in order to find which one to use for generating the retry hint.
1291 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1292 adding the default one. Similarly, if it contains a Reply-To: header, the
1293 errors_reply_to option, if set, is not used.
1295 PH/19 When calculating a retry time, Exim used to measure the "time since
1296 failure" by looking at the "first failed" field in the retry record. Now
1297 it does not use this if it is later than than the arrival time of the
1298 message. Instead it uses the arrival time. This makes for better
1299 behaviour in cases where some deliveries succeed, thus re-setting the
1300 "first failed" field. An example is a quota failure for a huge message
1301 when small messages continue to be delivered. Without this change, the
1302 "time since failure" will always be short, possible causing more frequent
1303 delivery attempts for the huge message than are intended.
1304 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1306 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1307 $1, $2, $3) because the numerical variables can be reset during some
1308 expansion items (e.g. "match"), thereby losing the authentication data.
1310 PH/21 Make -bV show the size of off_t variables so that the test suite can
1311 decide whether to run tests for quotas > 2G.
1313 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1314 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1315 filecount value is greater than 2G or if a quota value is greater than 2G
1316 on a system where the size of off_t is not greater than 4, a panic error
1319 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1320 never match. The debug and -bh output now contains an explicit error
1321 message indicating a malformed IPv4 address or mask.
1323 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1324 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1325 PH/23 above applies.
1327 PH/25 Do not write to syslog when running in the test harness. The only
1328 occasion when this arises is a failure to open the main or panic logs
1329 (for which there is an explicit test).
1331 PH/26 Added the /no_tell option to "control=freeze".
1333 PH/27 If a host name lookup failed very early in a connection, for example, if
1334 the IP address matched host_lookup and the reverse lookup yielded a name
1335 that did not have a forward lookup, an error message of the form "no IP
1336 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1337 could be logged. Now it outputs the IP address instead of "NULL".
1339 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1340 allows the sender and the authenticated sender to be set when
1341 submitting a message from within Exim. Since child_open_exim() is
1342 documented for local_scan(), the new function should be too.
1344 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1345 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1346 results in an empty string is now treated as unset.
1348 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1350 PH/31 Added sender_verify_fail logging option.
1352 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1353 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1354 tidied the source and removed it altogether.
1356 PH/33 When a queue run was abandoned because the load average was too high, a
1357 log line was always written; now it is written only if the queue_run log
1358 selector is set. In addition, the log line for abandonment now contains
1359 information about the queue run such as the pid. This is always present
1360 in "start" and "stop" lines but was omitted from the "abandon" line.
1362 PH/34 Omit spaces between a header name and the colon in the error message that
1363 is given when verify = headers_syntax fails (if there are lots of them,
1364 the message gets confusing).
1366 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1367 names, as there are now some PTR records that contain slashes. This check
1368 is only to protect against broken name servers that fall over on strange
1369 characters, so the fact that it applies to all lookups doesn't matter.
1371 PH/36 Now that the new test suite is complete, we can remove some of the
1372 special code in Exim that was needed for the old test suite. For example,
1373 sorting DNS records because real resolvers return them in an arbitrary
1374 order. The new test suite's fake resolver always returns records in the
1377 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1378 bounces) except when queue_only is set, to avoid logging races between
1379 the different processes.
1381 PH/38 Panic-die if .include specifies a non-absolute path.
1383 PH/39 A tweak to the "H" retry rule from its user.
1385 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1386 a label. They prevented compilation on older perls.
1388 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1389 a warning to be raised on newish perls.
1391 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1392 on queue. Changes to match documented behaviour of showing count of
1393 messages matching specified criteria.
1395 PH/40 Changed the default ident timeout from 30s to 5s.
1397 PH/41 Added support for the use of login_cap features, on those BSD systems
1398 that have them, for controlling the resources used by pipe deliveries.
1400 PH/42 The content-scanning code uses fopen() to create files in which to put
1401 message data. Previously it was not paying any attention to the mode of
1402 the files. Exim runs with umask(0) because the rest of the code creates
1403 files with open(), and sets the required mode explicitly. Thus, these
1404 files were ending up world-writeable. This was not a big issue, because,
1405 being within the spool directory, they were not world-accessible. I have
1406 created a function called modefopen, which takes an additional mode
1407 argument. It sets umask(777), creates the file, chmods it to the required
1408 mode, then resets the umask. All the relevant calls to fopen() in the
1409 content scanning code have been changed to use this function.
1411 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1412 to 24 hours. This avoids potential overflow problems when processing G
1413 and H retry rules. I suspect nobody ever tinkers with this value.
1415 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1417 PH/45 When the plaintext authenticator is running as a client, the server's
1418 challenges are checked to ensure they are valid base64 strings. By
1419 default, the authentication attempt is cancelled if an invalid string is
1420 received. Setting client_ignore_invalid_base64 true ignores these errors.
1421 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1422 they are received. Thus, the responses can be made to depend on the
1423 challenges. If an invalid string is ignored, an empty string is placed in
1426 PH/46 Messages that are created by the autoreply transport now contains a
1427 References: header, in accordance with RFCs 2822 and 3834.
1429 PH/47 Added authenticated_sender_force to the smtp transport.
1431 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1433 PH/49 Installed latest patch from the Sieve maintainer.
1435 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1436 was also set, the appendfile transport was unnecessarily scanning a
1437 directory of message files (e.g. for maildir delivery) to find the count
1438 of files (along with the size), even though it did not need this
1439 information. It now does the scan only if it needs to find either the
1440 size of the count of files.
1442 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1444 PH/52 Two bugs concerned with error handling when the smtp transport is
1447 (i) Exim was not creating retry information for temporary errors given
1448 for individual recipients after the DATA command when the smtp transport
1449 was used in LMTP mode. This meant that they could be retried too
1450 frequently, and not timed out correctly.
1452 (ii) Exim was setting the flag that allows error details to be returned
1453 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1454 recipients that were returned after the DATA command.
1456 PH/53 This is related to PH/52, but is more general: for any failing address,
1457 when detailed error information was permitted to be returned to the
1458 sender, but the error was temporary, then after the final timeout, only
1459 "retry timeout exceeded" was returned. Now it returns the full error as
1460 well as "retry timeout exceeded".
1462 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1463 do this, and (what is worse) MTAs that accept it.
1465 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1466 will now be deprecated.
1468 PH/56 New os.c-cygwin from the Cygwin maintainer.
1470 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1471 formats (previously only available in exim formats via -bpr, -bpru,
1472 and -bpra. Now also available in native and exiqgrep formats)
1474 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1475 with very large, slow to parse queues
1477 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1479 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1481 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1482 responses to authentication challenges, though it was showing the
1483 challenges; (ii) I've removed the CR characters from the debug output for
1486 PH/58 Allow for the insertion of a newline as well as a space when a string
1487 is turned into more than one encoded-word during RFC 2047 encoding. The
1488 Sieve code now uses this.
1490 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1491 data_4xx, lost_connection, tls_required.
1493 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1494 message was being sent as an SMTP response.
1496 PH/61 Add -l and -k options to exicyclog.
1498 PH/62 When verifying, if an address was redirected to one new address, so that
1499 verification continued, and the new address failed or deferred after
1500 having set something in $address_data, the value of $address_data was not
1501 passed back to the ACL. This was different to the case when no
1502 redirection occurred. The value is now passed back in both cases.
1504 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1505 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1506 use login_cap.h, so on its own it isn't the distinguishing feature. The
1507 new name refers directly to the setclassresources() function.
1509 PH/65 Added configuration files for NetBSD3.
1511 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1513 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1514 is preferred over IPv4.
1516 PH/68 The bounce_return_message and bounce_return_body options were not being
1517 honoured for bounces generated during the reception of non-SMTP messages.
1518 In particular, this applied to messages rejected by the ACL. This bug has
1519 been fixed. However, if bounce_return_message is true and bounce_return_
1520 body is false, the headers that are returned for a non-SMTP message
1521 include only those that have been read before the error was detected.
1522 (In the case of an ACL rejection, they have all been read.)
1524 PH/69 The HTML version of the specification is now built in a directory called
1525 spec_html instead of spec.html, because the latter looks like a path with
1526 a MIME-type, and this confuses some software.
1528 PH/70 Catch two compiler warnings in sieve.c.
1530 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1531 function verify_get_ident() calls ip_connect() to connect a socket, but
1532 if the "connect()" function timed out, ip_connect() used to close the
1533 socket. However, verify_get_ident() also closes the socket later, and in
1534 between Exim writes to the log, which may get opened at this point. When
1535 the socket was closed in ip_connect(), the log could get the same file
1536 descriptor number as the socket. This naturally causes chaos. The fix is
1537 not to close the socket in ip_connect(); the socket should be closed by
1538 the function that creates it. There was only one place in the code where
1539 this was missing, in the iplookup router, which I don't think anybody now
1540 uses, but I've fixed it anyway.
1542 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1543 well as to direct DNS lookups. Otherwise the handling of names in host
1544 lists is inconsistent and therefore confusing.
1550 PH/01 Two changes to the default runtime configuration:
1552 (1) Move the checks for relay_from_hosts and authenticated clients from
1553 after to before the (commented out) DNS black list checks.
1555 (2) Add control=submission to the relay_from_hosts and authenticated
1556 clients checks, on the grounds that messages accepted by these
1557 statements are most likely to be submissions.
1559 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1561 (1) Generate an error if the third argument for the ${prvs expansion is
1564 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1567 (3) Reset the variables that are obtained from the first argument of
1568 ${prvscheck and used in the second argument before leaving the code,
1569 because their memory is reclaimed, so using them afterwards may do
1572 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1573 one (it's much easier than Tom thought :-).
1575 (5) Because of (4), we can now allow for the use of $prvscheck_result
1576 inside the third argument.
1578 PH/03 For some reason, the default setting of PATH when running a command from
1579 a pipe transport was just "/usr/bin". I have changed it to
1582 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1583 anything to be listed in the output from -bV.
1585 PH/05 When a filter generated an autoreply, the entire To: header line was
1586 quoted in the delivery log line, like this:
1588 => >A.N.Other <ano@some.domain> <original@ddress> ...
1590 This has been changed so that it extracts the operative address. There
1591 may be more than one such address. If so, they are comma-separated, like
1594 => >ano@some.domain,ona@other.domain <original@ddress> ...
1596 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1597 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1598 not being looked up in the rDNS to get a host name, Exim was showing the
1599 IP address twice in Received: lines, even though the IP addresses were
1600 identical. For example:
1602 Received: from [1.2.3.4] (helo=[1.2.3.4])
1604 However, if the real host name was known, it was omitting the HELO data
1605 if it matched the actual IP address. This has been tidied up so that it
1606 doesn't show the same IP address twice.
1608 PH/07 When both +timestamp and +memory debugging was on, the value given by
1609 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1610 called by the debug printing, thereby overwriting the timestamp buffer.
1611 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1613 PH/08 When the original message was included in an autoreply transport, it
1614 always said "this is a copy of the message, including all the headers",
1615 even if body_only or headers_only was set. It now gives an appropriate
1618 PH/09 Applied a patch from the Sieve maintainer which:
1620 o fixes some comments
1621 o adds the (disabled) notify extension core
1622 o adds some debug output for the result of if/elsif tests
1623 o points to the current vacation draft in the documentation
1624 and documents the missing references header update
1628 o fixes a bug in processing the envelope test (when testing
1629 multiple envelope elements, the last element determinted the
1632 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1633 Electronic Mail") by including:
1635 Auto-submitted: auto-generated
1637 in the messages that it generates (bounce messages and others, such as
1638 warnings). In the case of bounce messages for non-SMTP mesages, there was
1639 also a typo: it was using "Auto_submitted" (underscore instead of
1640 hyphen). Since every message generated by Exim is necessarily in response
1641 to another message, thes have all been changed to:
1643 Auto-Submitted: auto-replied
1645 in accordance with these statements in the RFC:
1647 The auto-replied keyword:
1649 - SHOULD be used on messages sent in direct response to another
1650 message by an automatic process,
1652 - MUST NOT be used on manually-generated messages,
1654 - MAY be used on Delivery Status Notifications (DSNs) and Message
1655 Disposition Notifications (MDNs),
1657 - MUST NOT be used on messages generated by automatic or periodic
1658 processes, except for messages which are automatic responses to
1661 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1662 to the default Received: header definition.
1664 PH/12 Added log selector acl_warn_skipped (default on).
1666 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1667 variables because (a) they are in the wrong storage pool and (b) even if
1668 they were copied, it wouldn't work properly because of the caching.
1670 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1671 checking when decoding. Apparently there are clients that generate
1672 overlong encoded strings. Why am I not surprised?
1674 PH/15 If the first argument of "${if match_address" was not empty, but did not
1675 contain an "@" character, Exim crashed. Now it writes a panic log message
1676 and treats the condition as false.
1678 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1680 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1681 extension "envelope-auth". The code is finished and in agreement with
1682 other implementations, but there is no documentation so far and in fact,
1683 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1684 not changing the active code.
1686 Print executed "if" and "elsif" statements when debugging is used. This
1687 helps a great deal to understand what a filter does.
1689 Document more things not specified clearly in RFC3028. I had all this
1690 sorted out, when out of a sudden new issues came to my mind. Oops."
1692 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1695 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1696 canonical form (as documented). However, after a host name lookup from
1697 the IP address, check_host() was doing a simple string comparison with
1698 addresses acquired from the DNS when checking that the found name did
1699 have the original IP as one of its addresses. Since any found IPv6
1700 addresses are likely to be in abbreviated form, the comparison could
1701 fail. Luckily, there already exists a function for doing the comparison
1702 by converting both addresses to binary, so now that is used instead of
1703 the text comparison.
1705 PH/20 There was another similar case to PH/19, when a complete host name was
1706 given in a host list; looking up its IP address could give an abbreviated
1707 form, whereas the current host's name might or might not be abbreviated.
1708 The same fix has been applied.
1714 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1715 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1718 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1719 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1721 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1723 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1724 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1725 Also, for "Auto-Submitted", treat anything other than "no" as
1726 non-personal, in accordance with RFC 3834. (Previously it treated
1727 anything starting "auto-" as non-personal.)
1729 TF/01 The control=submission/name=... option had a problem with syntax
1730 errors if the name included a slash character. The /name= option
1731 now slurps the rest of the string, so it can include any characters
1732 but it must come last in the list of options (after /sender_retain
1735 PH/05 Some modifications to the interface to the fake nameserver for the new
1743 TK/01 Added the "success_on_redirect" address verification option. See
1744 NewStuff for rationale and an example.
1746 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1748 PH/02 Patch to exigrep to allow it to work on syslog lines.
1750 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1751 fread() to skip over the body file's header line, because in Cygwin the
1752 header line is locked and is inaccessible.
1754 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1755 co-exist for some time) to make it clear that it is the Exim ID that is
1756 referenced, not the Message-ID: header line.
1758 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1759 string_format() function, because snprintf() does not exist on all
1762 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1763 ${stat: expansion item.
1765 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1766 protocol synchronization error", to keep the pedants happy.
1768 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1769 well as for IRIX systems, when gcc is being used. See the host.c source
1772 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1774 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1777 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1778 transport and to the smtp transport in LMTP mode.
1780 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1782 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1783 run a filter in a subprocess. This could lead to confusion in subsequent
1784 lookups in the parent process. There should also be a search_tidyup() at
1785 the end of the subprocess.
1787 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1788 only if the host matched helo_try_verify_hosts, which caused the
1789 verification to occur when the EHLO/HELO command was issued. The ACL just
1790 tested the remembered result. Now, if a previous verification attempt has
1791 not happened, "verify = helo" does it there and then.
1793 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1795 TK/03 Fix log output including CR from clamd.
1797 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1798 exist provoked a memory error which could cause a segfault.
1800 PH/15 Installed PCRE 6.2
1802 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1804 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1805 of the problem. Specifically, suggested +O2 rather than +O1 for the
1808 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1810 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1811 record, the port was not correctly logged when the outgoing_port log
1812 selector was set (it logged the transort's default port).
1814 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1815 fallback_hosts, and "hosts" in the smtp transport.
1817 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1818 host errors such as "Connection refused".
1820 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1821 authentication with radiusclient 0.4.9:
1823 - Error returned from rc_read_config was caught wrongly
1824 - Username/password not passed on to radius server due to wrong length.
1826 The presumption is that some radiusclient API changes for 4.51/PH/17
1827 were not taken care of correctly. The code is still untested by me (my
1828 Linux distribution still has 0.3.2 of radiusclient), but it was
1829 contributed by a Radius user.
1831 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1832 expanding the "port" option of the smtp transport.
1834 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1835 while reading a MIME header. Thanks to Tom Hughes for a patch.
1837 PH/24 Include config.h inside local_scan.h so that configuration settings are
1840 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1841 that in an ACL for RCPT (for example), you can examine exactly what was
1844 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1845 commands, but it was not correctly comparing the address with the actual
1846 client host address. Thus, it would show the EHLO address in Received:
1847 header lines when this was not necessary.
1849 PH/27 Added the % operator to ${eval:}.
1851 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1852 it should be ignoring failures (because with -C, for example, it has lost
1853 privilege). It wasn't ignoring creation failures other than "already
1856 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1859 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1860 because an input line was too long, either on its own, or by virtue of
1861 too many continuations, the temporary file was not being removed, and the
1862 return code was incorrect.
1864 PH/31 Missing "BOOL" in function definition in filtertest.c.
1866 PH/32 Applied Sieve patches from the maintainer.
1868 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1870 PH/33 Added "verify = not_blind".
1872 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1873 Local/Makefile (with some defaults set). These are used in built scripts
1874 such as exicyclog, but they have never been used in the exim_install
1875 script (though there are many overriding facilities there). I have
1876 arranged that the exim_install script now takes note of these two
1879 PH/35 Installed configuration files for Dragonfly.
1881 PH/36 When a locally submitted message by a trusted user did not contain a
1882 From: header, and the sender address was obtained from -f or from an SMTP
1883 MAIL command, and the trusted user did not use -F to supply a sender
1884 name, $originator_name was incorrectly used when constructing a From:
1885 header. Furthermore, $originator_name was used for submission mode
1886 messages from external hosts without From: headers in a similar way,
1887 which is clearly wrong.
1889 PH/37 Added control=suppress_local_fixups.
1891 PH/38 When log_selector = +received_sender was set, and the addition of the
1892 sender made the log line's construction buffer exactly full, or one byte
1893 less than full, an overflow happened when the terminating "\n" was
1896 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1897 when the result of a list match is failure because a DNS lookup failed.
1899 PH/40 RM_COMMAND is now used in the building process.
1901 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1902 the "build-* directories that it finds.
1904 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1905 address in a domain literal was a prefix of an interface address.
1907 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1908 when verifying a sender address, unless rewrite_headers is false.
1910 PH/44 Wrote a long comment about why errors_to addresses are verified as
1911 recipients, not senders.
1913 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1914 the ratelimit ACL was added.
1916 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1918 PH/46 Added extra information about PostgreSQL errors to the error string.
1920 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1921 avoiding the need to install special zones in a real server. This is
1922 backwards compatible; if it can't find the fake resolver, it drops back.
1923 Thus, both old and new test suites can be run.
1925 TF/02 Added util/ratelimit.pl
1927 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1928 clock is set back in time.
1930 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1931 Candler <B.Candler@pobox.com>.
1933 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1934 OK for addresses that are the result of redirections.
1936 PH/48 A number of further additions for the benefit of the new test suite,
1937 including a fake gethostbyname() that interfaces to the fake DNS resolver
1940 TF/06 The fix for widen_domains has also been applied to qualify_single and
1941 search_parents which are the other dnslookup options that can cause
1944 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1947 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1949 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1950 service" if it encountered an MX record with an empty target hostname.
1951 The message is now "an MX or SRV record indicated no SMTP service".
1953 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1954 if the DNS of the sending site is misconfigured. This is quite a
1955 common situation. This change restores the behaviour of treating a
1956 helo verification defer as a failure.
1958 PH/51 If self=fail was set on a router, the bounce message did not include the
1959 actual error message.
1965 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1967 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1968 command itself ended in error, the underlying message about the transport
1969 filter timeout was being overwritten with the pipe command error. Now the
1970 underlying error message should be appended to the second error message.
1972 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1974 PH/02 Reduce the amount of output that "make" produces by default. Full output
1975 can still be requested.
1977 PH/03 The warning log line about a condition test deferring for a "warn" verb
1978 was being output only once per connection, rather than after each
1979 occurrence (because it was using the same function as for successful
1980 "warn" verbs). This seems wrong, so I have changed it.
1982 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1983 it should not have, which might have caused a crash in the right
1984 circumstances, but probably never did.
1986 PH/04 Installed a modified version of Tony Finch's patch to make submission
1987 mode fix the return path as well as the Sender: header line, and to
1988 add a /name= option so that you can make the user's friendly name appear
1991 TF/03 Added the control = fakedefer ACL modifier.
1993 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1994 Mark Lowes for thorough testing.
1996 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1998 TK/03 Merged latest SRS patch from Miles Wilton.
2000 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2001 with the definition in sysexits.h (which is #included earlier).
2002 Fortunately, Exim does not actually use EX_OK. The code used to try to
2003 preserve the sysexits.h value, by assumimg that macro definitions were
2004 scanned for macro replacements. I have been disabused of this notion,
2005 so now the code just undefines EX_OK before #including unistd.h.
2007 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2008 in the smtp transport. When a block could not be written in a single
2009 write() function, the timeout was being re-applied to each part-write.
2010 This seems wrong - if the receiver was accepting one byte at a time it
2011 would take for ever. The timeout is now adjusted when this happens. It
2012 doesn't have to be particularly precise.
2014 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2015 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2017 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2018 without a domain if the check to <postmaster@domain> fails.
2020 SC/01 Eximstats: added -xls and the ability to specify output files
2021 (patch written by Frank Heydlauf).
2023 SC/02 Eximstats: use FileHandles for outputing results.
2025 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2027 SC/04 Eximstats: fixed display of large numbers with -nvr option
2029 SC/05 Eximstats: fixed merging of reports with empty tables.
2031 SC/06 Eximstats: added the -include_original_destination flag
2033 SC/07 Eximstats: removed tabs and trailing whitespace.
2035 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2037 TK/06 MBOX spool code: Add real "From " MBOX separator line
2038 so the .eml file is really in mbox format (even though
2039 most programs do not really care). Patch from Alex Miller.
2041 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2042 The latter is generated from $received_to and is only set if the
2043 message has one envelope recipient. SA can use these headers,
2044 obviously out-of-the-box. Patch from Alex Miller.
2046 PH/08 The ${def test on a variable was returning false if the variable's
2047 value was "0", contrary to what the specification has always said!
2048 The result should be true unless the variable is empty.
2050 PH/09 The syntax error of a character other than { following "${if
2051 def:variable_name" (after optional whitespace) was not being diagnosed.
2052 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2053 accidental colon was present, for example, could give incorrect results.
2055 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2056 result is used (not including appendfile, where other changes are about
2059 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2060 This involved changing a lot of size variables from int to off_t. It
2061 should work with maildirs and everything.
2063 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2064 spamd dying while we are connected to it.
2066 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2069 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2070 function for BDB 4.3.
2072 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2074 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2075 the sources for PCRE have been re-arranged and more files are now
2078 PH/15 The code I had for printing potentially long long variables in PH/11
2079 above was not the best (it lost precision). The length of off_t variables
2080 is now inspected at build time, and an appropriate printing format (%ld
2081 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2082 to be "long long int" or "long int". This is needed for the internal
2083 formatting function string_vformat().
2085 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2086 the configuration file to be ":syslog", then the script "guesses" where
2087 the logs files are, rather than using the compiled in default. In our
2088 case the guess is not the same as the compiled default, so the script
2089 suddenly stopped working when I started to use syslog. The patch checks
2090 to see if log_file_path is "". If so, it attempts to read it from exim
2091 with no configuration file to get the compiled in version, before it
2092 falls back to the previous guessing code."
2094 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2095 implementing BATV in an Exim configuration. See NewStuff for the gory
2098 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2099 Makefile that are specific to HP-UX.
2101 PH/18 If the "use_postmaster" option was set for a recipient callout together
2102 with the "random" option, the postmaster address was used as the MAIL
2103 FROM address for the random test, but not for the subsequent recipient
2104 test. It is now used for both.
2106 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2107 patch removes a few documentation additions to RFC 3028, because the
2108 latest draft now contains them. It adds the new en;ascii-case comparator
2109 and a new error check for 8bit text in MIME parts. Comparator and
2110 require names are now matched exactly. I enabled the subaddress
2111 extension, but it is not well tested yet (read: it works for me)."
2113 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2114 rework some of the code of TK/09 above to avoid the hardwired use of
2115 "%lld" and "long long". Replaced the call to snprintf() with a call to
2118 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2119 records point to non-existent hosts", "retry timeout exceeded", and
2120 "retry time not reached for any host after a long failure period".
2122 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2123 experimental DomainKeys support:
2125 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2126 (2) On an error such as an illegally used "control", the wrong name for
2127 the control was given.
2129 These problems did NOT occur unless DomainKeys support was compiled.
2131 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2133 PH/24 Added ${if match_ip condition.
2135 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2136 databases so that it will be absolutely obvious if a crash occurs in the
2137 DB library. This is a regular occurrence (often caused by mis-matched
2140 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2141 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2142 picked up on a user's system that detects such things. There doesn't seem
2143 to be a gcc warning option for this - only an attribute that has to be
2144 put on the function's prototype. It seems that in Fedora Core 4 they have
2145 set this on a number of new functions. No doubt there will be more in due
2148 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2149 specify a transport. However, if an address that was verified by such a
2150 router was the subject of a callout, Exim crashed because it tried to
2151 read the rcpt_include_affixes from the non-existent transport. Now it
2152 just assumes that the setting of that option is false. This bug was
2153 introduced by 4.51/PH/31.
2155 PH/28 Changed -d+all to exclude +memory, because that information is very
2156 rarely of interest, but it makes the output a lot bigger. People tend to
2157 do -d+all out of habit.
2159 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2160 code in os-type was giving problems when libc.so lives in lib64, like on
2163 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2164 aren't the modern standard, and it seems that some systems' include files
2165 don't always have them. Exim was already checking for some of the newer
2166 ones like T_AAAA, and defining it itself. I've added checks for all the
2167 record types that Exim uses.
2169 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2170 not automatically generating a new one, as it is supposed to. This
2171 prevented TLS from working. If the file did exist, but contained invalid
2172 data, a new version was generated, as expected. It was only the case of a
2173 non-existent file that was broken.
2175 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2176 with a change in libdomainkeys > 0.64.
2178 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2179 from DNS. If the selector record carries the flag, it now has
2180 precedence over the domain-wide flag.
2182 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2184 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2185 the use of a port name that isn't defined in /etc/services) occurred, the
2186 message was deferred as in a normal delivery, and thus remained on the
2187 spool, instead of being failed because of the mua_wrapper setting. This
2188 is now fixed, and I tidied up some of the mua_wrapper messages at the
2191 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2192 the messages in a hash of arrays rather than using individual hashes.
2193 This is a bit cleaner and results in dramatic memory savings, albeit
2194 at a slight CPU cost.
2196 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2197 as requested by Marc Sherman.
2199 SC/10 Eximstats: added histograms for user specified patterns as requested
2202 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2204 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2205 fopen() in the content-scanning modules that did not already have it.
2211 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2212 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2214 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2216 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2218 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2219 to test Sieve filters that use "vacation".
2221 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2222 that changes the way the GnuTLS parameters are stored in the cache file.
2223 The new format can be generated externally. For backward compatibility,
2224 if the data in the cache doesn't make sense, Exim assumes it has read an
2225 old-format file, and it generates new data and writes a new file. This
2226 means that you can't go back to an older release without removing the
2229 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2230 work if there are any delivery delays because "one_time" forces the
2231 parent to be marked "delivered", so its unseen clone is never tried
2232 again. For this reason, Exim now forbids the simultaneous setting of
2235 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2236 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2237 entry if you want to know the details. The fix, however, neglected to
2238 consider the case when local delivery batching is involved. The test for
2239 "previously delivered" was not happening when checking to see if an
2240 address could be batched with a previous (undelivered) one; under
2241 certain circumstances this could lead to multiple deliveries to the same
2244 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2245 in its include files, and this causes problems building Exim.
2247 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2248 header_syntax) but Exim was just ignoring anything given after a slash.
2249 In particular, this caused confusion with an attempt to use "verify =
2250 reverse_host_lookup/defer_ok". An error is now given when options are
2251 supplied for verify items that do not have them. (Maybe reverse_host_
2252 lookup should have a defer_ok option, but that's a different point.)
2254 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2255 defined by RFC 821) to 2048, because there were problems with some AUTH
2256 commands, and RFC 1869 says the size should be increased for extended
2257 SMTP commands that take arguments.
2259 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2262 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2263 "unknown" error; now it says that the functionality isn't in the binary.
2265 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2266 an address' error message when a string expansion fails (syntax or
2267 whatever). Otherwise the password may appear in the log. Following change
2268 PH/42 below, there is no longer a chance of it appearing in a bounce
2271 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2273 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2274 of its addresses was ever tried. (Bugzilla bug #2.)
2276 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2277 the result incorrectly in the debug output. (It correctly added a newline
2278 to what was transported.)
2280 TF/01 Added $received_time.
2282 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2283 commented out examples of how to interface to a virus scanner and to
2284 SpamAssassin. Also added commented examples of av_scanner and
2285 spamd_address settings.
2287 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2288 and controls are allowed in which ACLs. There were a couple of minor
2289 errors. Some of the entries in the conditions table (which is a table of
2290 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2291 negation of where the condition IS allowed.
2293 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2295 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2296 header file does not have a version number, so I've had to invent a new
2297 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2298 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2299 radiusclient), but it was contributed by a Radius user.
2301 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2302 files or directories, for OpenSSL.
2304 PH/19 When an Exim process that is running as root has to create an Exim log
2305 file, it does so in a subprocess that runs as exim:exim so as to get the
2306 ownership right at creation (otherwise, other Exim processes might see
2307 the file with the wrong ownership). There was no test for failure of this
2308 fork() call, which would lead to the process getting stuck as it waited
2309 for a non-existent subprocess. Forks do occasionally fail when resources
2310 run out. I reviewed all the other calls to fork(); they all seem to check
2313 PH/20 When checking for unexpected SMTP input at connect time (before writing
2314 the banner), Exim was not dealing correctly with a non-positive return
2315 from the read() function. If the client had disconnected by this time,
2316 the result was a log entry for a synchronization error with an empty
2317 string after "input=" when read() returned zero. If read() returned -1
2318 (an event I could not check), uninitialized data bytes were printed.
2319 There were reports of junk text (parts of files, etc) appearing after
2322 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2324 PH/22 Added support for macro redefinition, and (re)definition in between
2325 driver and ACL definitions.
2327 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2328 forgetting to use the resulting value; it was using the unexpanded value.
2330 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2331 hadn't been configured. The fix is from Juergen Kreileder, who
2332 understands it better than I do:
2334 "Here's what I see happening with three configured cyrus_sasl
2335 authenticators configured (plain, login, cram-md5):
2337 On startup auth_cyrus_sasl_init() gets called for each of these.
2338 This means three calls to sasl_listmech() without a specified mech_list.
2339 => SASL tests which mechs of all available mechs actually work
2340 => three warnings about OTP not working
2341 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2343 With the patch, sasl_listmech() also gets called three times. But now
2344 SASL's mech_list option is set to the server_mech specified in the the
2345 authenticator. Or in other words, the answer from sasl_listmech()
2346 gets limited to just the mech you're testing for (which is different
2348 => the return list contains just 'plain' or 'login', 'cram-md5' or
2349 nothing depending on the value of ob->server_mech.
2351 I've just tested the patch: Authentication still works fine,
2352 unavailable mechs specified in the exim configuration are still
2353 caught, and the auth.log warnings about OTP are gone."
2355 PH/25 When debugging is enabled, the contents of the command line are added
2356 to the debugging output, even when log_selector=+arguments is not
2359 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2360 answer is "GNU", and only if the return is "GNU/something" is the answer
2363 PH/27 $acl_verify_message is now set immediately after the failure of a
2364 verification in an ACL, and so is available in subsequent modifiers. In
2365 particular, the message can be preserved by coding like this:
2367 warn !verify = sender
2368 set acl_m0 = $acl_verify_message
2370 Previously, $acl_verify_message was set only while expanding "message"
2371 and "log_message" when a very denied access.
2373 PH/28 Modified OS/os.c-Linux with
2375 -#ifndef OS_LOAD_AVERAGE
2376 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2378 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2379 nomenclature these days.)
2381 PH/29 Installed patch from the Sieve maintainer that adds the options
2382 sieve_useraddress and sieve_subaddress to the redirect router.
2384 PH/30 In these circumstances:
2385 . Two addresses routed to the same list of hosts;
2386 . First host does not offer TLS;
2387 . First host accepts first address;
2388 . First host gives temporary error to second address;
2389 . Second host offers TLS and a TLS session is established;
2390 . Second host accepts second address.
2391 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2392 and peerdn, if requested) that were in fact used only for the second
2395 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2396 attention to any local part prefix or suffix that was matched by the
2397 router that accepted the address. It now behaves in the same way as it
2398 does for delivery: the affixes are removed from the local part unless
2399 rcpt_include_affixes is set on the transport.
2401 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2402 timeout during the DATA phase of an incoming message.
2404 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2405 applied a patch sanctioned by the Sieve maintainer.
2407 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2408 the uid or gid is negative. A case of a negative gid caused this to be
2409 noticed. The fix allows for either to be negative.
2411 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2412 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2415 PH/36 The change PH/12 above was broken. Fixed it.
2417 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2418 the grounds that routing the same address twice would always produce the
2419 same answer. This might have been true once, but it is certainly no
2420 longer true now. Routing a child address may depend on the previous
2421 routing that produced that child. Some complicated redirection strategies
2422 went wrong when messages had multiple recipients, and made Exim's
2423 behaviour dependent on the order in which the addresses were given.
2425 I have moved the duplicate checking until after the routing is complete.
2426 Exim scans the addresses that are assigned to local and remote
2427 transports, and removes any duplicates. This means that more work will be
2428 done, as duplicates will always all be routed, but duplicates are
2429 presumably rare, so I don't expect this is of any significance.
2431 For deliveries to pipes, files, and autoreplies, the duplicate checking
2432 still happens during the routing process, since they are not going to be
2435 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2436 It corrects a timeout issue with spamd. This is Ian's comment: "The
2437 background is that sometimes spamd either never reads data from a
2438 connection it has accepted, or it never writes response data. The exiscan
2439 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2440 blindly assumes that writes won't block so it may never time out."
2442 PH/39 Allow G after quota size as well as K and M.
2444 PH/40 The value set for $authenticated_id in an authenticator may not contain
2445 binary zeroes or newlines because the value is written to log lines and
2446 to spool files. There was no check on this. Now the value is run through
2447 the string_printing() function so that such characters are converted to
2448 printable escape sequences.
2450 PH/41 $message_linecount is a new variable that contains the total number of
2451 lines in the message. Compare $body_linecount, which is the count for the
2454 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2455 bounce and delay warning messages, except in certain special cases, which
2458 (a) An SMTP error message from a remote host;
2459 (b) A message specified in a :fail: redirection;
2460 (c) A message specified in a "fail" command in a system filter;
2461 (d) A message specified in a FAIL return from the queryprogram router;
2462 (e) A message specified by the cannot_route_message router option.
2464 In these cases only, Exim does include the error details in bounce and
2465 warning messages. There are also a few cases where bland messages such
2466 as "unrouteable address" or "local delivery error" are given.
2468 PH/43 $value is now also set for the "else" part of a ${run expansion.
2470 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2471 being worked on, but at least Exim now implements the latest version to
2474 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2475 process to complete was treated as a delivery failure, a timeout while
2476 writing the message to the pipe was logged, but erroneously treated as a
2477 successful delivery. Such timeouts include transport filter timeouts. For
2478 consistency with the overall process timeout, these timeouts are now
2479 treated as errors, giving rise to delivery failures by default. However,
2480 there is now a new Boolean option for the pipe transport called
2481 timeout_defer, which, if set TRUE, converts the failures into defers for
2482 both kinds of timeout. A transport filter timeout is now identified in
2485 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2486 systems where "make" and "gmake" are different, calling "gmake" at top
2487 level broke things. I've arranged for the value of $(MAKE) to be passed
2488 from the Makefile to this script so that it can call the same version of
2492 A note about Exim versions 4.44 and 4.50
2493 ----------------------------------------
2495 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2496 changes of various kinds. As a consequence, a big documentation update was
2497 needed. This delayed the release for rather longer than seemed good, especially
2498 in the light of a couple of (minor) security issues. Therefore, the changes
2499 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2500 release. So 4.44 and 4.50 are in effect two different branches that both start
2503 I have left the 4.50 change log unchanged; it contains all the changes since
2504 4.43. The change log for 4.44 is below; many of its items are identical to
2505 those for 4.50. This seems to be the most sensible way to preserve the
2506 historical information.
2512 1. Minor wording change to the doc/README.SIEVE file.
2514 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2515 computation of the current number of files was incorrect.
2517 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2518 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2521 4. Give more explanation in the error message when the command for a transport
2522 filter fails to execute.
2524 5. There are several places where Exim runs a non-Exim command in a
2525 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2526 was being done only for the command run by the queryprogram router. It is
2527 now done for all such subprocesses. The other cases are: ${run, transport
2528 filters, and the commands run by the lmtp and pipe transports.
2530 6. Added CONFIGURE_GROUP build-time option.
2532 7. Some older OS have a limit of 256 on the maximum number of file
2533 descriptors. Exim was using setrlimit() to set 1000 as a large value
2534 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2535 systems. I've change it so that if it can't get 1000, it tries for 256.
2537 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2538 was an oversight, and furthermore, ever since the addition of extra
2539 controls (e.g. 4.43/32), the checks on when to allow different forms of
2540 "control" were broken. There should now be diagnostics for all cases when a
2541 control that does not make sense is encountered.
2543 9. Added the /retain_sender option to "control=submission".
2545 10. $recipients is now available in the predata ACL (oversight).
2547 11. Tidy the search cache before the fork to do a delivery from a message
2548 received from the command line. Otherwise the child will trigger a lookup
2549 failure and thereby defer the delivery if it tries to use (for example) a
2550 cached ldap connection that the parent has called unbind on.
2552 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2553 of $address_data from the recipient verification was clobbered by the
2554 sender verification.
2556 13. The value of address_data from a sender verification is now available in
2557 $sender_address_data in subsequent conditions in the ACL statement.
2559 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2561 15. Added a new option "connect=<time>" to callout options, to set a different
2564 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2565 was its contents. (It was OK if the option was not defined at all.)
2567 17. A "Completed" log line is now written for messages that are removed from
2568 the spool by the -Mrm option.
2570 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2571 information about exactly what failed.
2573 19. Added -dd to debug only the daemon process.
2575 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2576 handles timeouts, both on the server side and network timeouts. Renamed the
2577 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2579 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2580 It is now set to "smtps".
2582 22. $host_address is now set to the target address during the checking of
2583 ignore_target_hosts.
2585 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2586 being passed; this would have caused $sender_host_name to have been used if
2587 matching the list had actually called for a host name (not very likely,
2588 since this list is usually IP addresses). A host name is now passed as
2591 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2592 code that specifies a non-restarting handler (typically sigaction() in
2593 modern systems) in an attempt to fix a rare and obscure crash bug.
2595 25. Narrowed the window for a race in the daemon that could cause it to ignore
2596 SIGCHLD signals. This is not a major problem, because they are used only to
2597 wake it up if nothing else does.
2599 26. A malformed maildirsize file could cause Exim to calculate negative values
2600 for the mailbox size or file count. Odd effects could occur as a result.
2601 The maildirsize information is now recalculated if the size or filecount
2604 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2605 support for a long time. Removed HAVE_SYS_VFS_H.
2607 28. Installed the latest version of exipick from John Jetmore.
2609 29. In an address list, if the pattern was not a regular expression, an empty
2610 subject address (from a bounce message) matched only if the pattern was an
2611 empty string. Non-empty patterns were not even tested. This was the wrong
2612 because it is perfectly reasonable to use an empty address as part of a
2613 database query. An empty address is now tested by patterns that are
2614 lookups. However, all the other forms of pattern expect the subject to
2615 contain a local part and a domain, and therefore, for them, an empty
2616 address still always fails if the pattern is not itself empty.
2618 30. Exim went into a mad DNS loop when attempting to do a callout where the
2619 host was specified on an smtp transport, and looking it up yielded more
2620 than one IP address.
2622 31. Re-factored the code for checking spool and log partition space into a
2623 function that finds that data and another that does the check. The former
2624 is then used to implement four new variables: $spool_space, $log_space,
2625 $spool_inodes, and $log_inodes.
2627 32. The RFC2047 encoding function was originally intended for short strings
2628 such as real names; it was not keeping to the 75-character limit for
2629 encoded words that the RFC imposes. It now respects the limit, and
2630 generates multiple encoded words if necessary. To be on the safe side, I
2631 have increased the buffer size for the ${rfc2047: expansion operator from
2634 33. It is now permitted to omit both strings after an "if" condition; if the
2635 condition is true, the result is "true". As before, when the second string
2636 is omitted, a false condition yields an empty string. This makes it less
2637 cumbersome to write custom ACL and router conditions.
2639 34. Failure to deliver a bounce message always caused it to be frozen, even if
2640 there was an errors_to setting on the router. The errors_to setting is now
2643 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2644 canonical form (fully expanded) before being placed in
2645 $sender_host_address.
2647 36. The table in the code that translates DNS record types into text (T_A to
2648 "A" for instance) was missing entries for NS and CNAME. It is just possible
2649 that this could have caused confusion if both these types were looked up
2650 for the same domain, because the text type is used as part of Exim's
2651 per-process caching. But the chance of anyone hitting this buglet seems
2654 37. The dnsdb lookup has been extended in a number of ways.
2656 (1) There is a new type, "zns", which walks up the domain tree until it
2657 finds some nameserver records. It should be used with care.
2659 (2) There is a new type, "mxh", which is like "mx" except that it returns
2660 just the host names, not the priorities.
2662 (3) It is now possible to give a list of domains (or IP addresses) to be
2663 looked up. The behaviour when one of the lookups defers can be
2664 controlled by a keyword.
2666 (4) It is now possible to specify the separator character for use when
2667 multiple records are returned.
2669 38. The dnslists ACL condition has been extended: it is now possible to supply
2670 a list of IP addresses and/or domains to be looked up in a particular DNS
2673 39. Added log_selector=+queue_time_overall.
2675 40. When running the queue in the test harness, wait just a tad after forking a
2676 delivery process, to get repeatability of debugging output.
2678 41. Include certificate and key file names in error message when GnuTLS fails
2679 to set them up, because the GnuTLS error message doesn't include the name
2680 of the failing file when there is a problem reading it.
2682 42. Allow both -bf and -bF in the same test run.
2684 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2686 44. The "Exiscan patch" is now merged into the mainline Exim source.
2688 45. Sometimes the final signoff response after QUIT could fail to get
2689 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2690 < 0 before doing a fflush(). This bug looks as though it goes back to the
2691 introduction of TLS in release 3.20, but "sometimes" must have been rare
2692 because the tests only now provoked it.
2694 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2695 (this can affect the format of dates).
2697 47. exim_tidydb, when checking for the continued existence of a message for
2698 which it has found a message-specific retry record, was not finding
2699 messages that were in split spool directories. Consequently, it was
2700 deleting retry records that should have stayed in existence.
2702 48. Steve fixed some bugs in eximstats.
2704 49. The SPA authentication driver was not abandoning authentication and moving
2705 on to the next authenticator when an expansion was forced to fail,
2706 contradicting the general specification for all authenticators. Instead it
2707 was generating a temporary error. It now behaves as specified.
2709 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2710 (the order specifies the preference for clients). The order is now AES256,
2711 AES128, 3DES, ARCFOUR128.
2713 51. Small patch to Sieve code - explicitly set From: when generating an
2716 52. Exim crashed if a remote delivery caused a very long error message to be
2717 recorded - for instance if somebody sent an entire SpamAssassin report back
2718 as a large number of 550 error lines. This bug was coincidentally fixed by
2719 increasing the size of one of Exim's internal buffers (big_buffer) that
2720 happened as part of the Exiscan merge. However, to be on the safe side, I
2721 have made the code more robust (and fixed the comments that describe what
2724 53. Now that there can be additional text after "Completed" in log lines (if
2725 the queue_time_overall log selector is set), a one-byte patch to exigrep
2726 was needed to allow it to recognize "Completed" as not the last thing in
2729 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2730 patch that reportedly fixes this has been added. I am not expert enough to
2731 create a test for it. This is what the patch creator wrote:
2733 "I found a little strange behaviour of ldap code when working with
2734 Windows 2003 AD Domain, where users was placed in more than one
2735 Organization Units. When I tried to give exim partial DN, the exit code
2736 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2737 But simultaneously result of request was absolutely normal ldap result,
2738 so I produce this patch..."
2740 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2741 REFERENCE, so I have modified the code to exclude the patch when that macro
2744 55. Some experimental protocols are using DNS PTR records for new purposes. The
2745 keys for these records are domain names, not reversed IP addresses. The
2746 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2747 leaves it alone. Component reversal etc. now happens only for IP addresses.
2750 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2752 57. Double the size of the debug message buffer (to 2048) so that more of very
2753 long debug lines gets shown.
2755 58. The exicyclog utility now does better if the number of log files to keep
2756 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2758 59. Two changes related to the smtp_active_hostname option:
2760 (1) $smtp_active_hostname is now available as a variable.
2761 (2) The default for smtp_banner uses $smtp_active_hostname instead
2762 of $primary_hostname.
2764 60. The host_aton() function is supposed to be passed a string that is known
2765 to be a valid IP address. However, in the case of IPv6 addresses, it was
2766 not checking this. This is a hostage to fortune. Exim now panics and dies
2767 if the condition is not met. A case was found where this could be provoked
2768 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2769 components; fortuitously, this particular loophole had already been fixed
2770 by change 4.50/55 above.
2772 If there are any other similar loopholes, the new check in host_aton()
2773 itself should stop them being exploited. The report I received stated that
2774 data on the command line could provoke the exploit when Exim was running as
2775 exim, but did not say which command line option was involved. All I could
2776 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2777 running as the user.
2780 61. There was a buffer overflow vulnerability in the SPA authentication code
2781 (which came originally from the Samba project). I have added a test to the
2782 spa_base64_to_bits() function which I hope fixes it.
2785 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2786 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2788 63. The daemon start-up calls getloadavg() while still root for those OS that
2789 need the first call to be done as root, but it missed one case: when
2790 deliver_queue_load_max is set with deliver_drop_privilege. This is
2791 necessary for the benefit of the queue runner, because there is no re-exec
2792 when deliver_drop_privilege is set.
2794 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2795 This has been fixed.
2797 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2798 was in use, was not putting the data itself into the right store pool;
2799 consequently, it could be overwritten for a subsequent message in the same
2800 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2803 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2805 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2806 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2807 the function were treating the return as a boolean value, which happened to
2808 work because 0=false and not-0=true, but is not correct code.
2810 68. The host_aton() function was not handling scoped IPv6 addresses (those
2811 with, for example, "%eth0" on the end) correctly.
2813 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2814 negated items (that is, ~something) in unsigned ints. Some compilers
2815 apparently mutter when there is no cast.
2817 70. If an address verification called from an ACL failed, and did not produce a
2818 user-specific message (i.e. there was only a "system" message), nothing was
2819 put in $acl_verify_message. In this situation, it now puts the system
2822 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2823 session; change 4.31/43 added the unwanted input to the log line - except
2824 that it did not do this in the start of session case. It now does.
2826 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2827 This is wrong and can cause the other end to generate a synchronization
2828 error if it is another Exim or anything else that does the synchronization
2829 check. A QUIT command is no longer sent after a timeout.
2831 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2832 during host lookups.
2834 74. The defer_ok option of callout verification was not working if it was used
2835 when verifying addresses in header lines, that is, for this case:
2837 verify = header_sender/callout=defer_ok
2839 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2840 those file descriptors could be used for SMTP connections. If anything
2841 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2842 could be sent to the SMTP client, causing chaos. The daemon now opens
2843 stdin, stdout, and stderr to /dev/null when it puts itself into the
2846 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2847 log by default. The user can override this with suitable Perl magic.
2849 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2850 the log message when discard triggers, was not working for the DATA ACL or
2851 for the non-SMTP ACL.
2853 78. Error message wording change in sieve.c.
2855 79. If smtp_accept_max_per_host was set, the number of connections could be
2856 restricted to fewer than expected, because the daemon was trying to set up
2857 a new connection before checking whether the processes handling previous
2858 connections had finished. The check for completed processes is now done
2859 earlier. On busy systems, this bug wouldn't be noticed because something
2860 else would have woken the daemon, and it would have reaped the completed
2863 80. If a message was submitted locally by a user whose login name contained one
2864 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2865 It caused a spool format error. I have fixed the spool reading code. A
2866 related problem was that the "from" clause in the Received: line became
2867 illegal because of the space(s). It is now covered by ${quote_local_part.
2869 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2872 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2874 83. Patch from Sieve maintainer to fix unterminated string problem in
2875 "vacation" handling.
2877 84. Some minor changes to the Linux configuration files to help with other
2878 OS variants using glibc.
2880 85. One more patch for Sieve to update vacation handling to latest spec.
2883 ----------------------------------------------------
2884 See the note above about the 4.44 and 4.50 releases.
2885 ----------------------------------------------------
2891 1. Change 4.43/35 introduced a bug that caused file counts to be
2892 incorrectly computed when quota_filecount was set in an appendfile
2895 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2896 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2899 3. Give more explanation in the error message when the command for a transport
2900 filter fails to execute.
2902 4. There are several places where Exim runs a non-Exim command in a
2903 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2904 was being done only for the command run by the queryprogram router. It is
2905 now done for all such subprocesses. The other cases are: ${run, transport
2906 filters, and the commands run by the lmtp and pipe transports.
2908 5. Some older OS have a limit of 256 on the maximum number of file
2909 descriptors. Exim was using setrlimit() to set 1000 as a large value
2910 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2911 systems. I've change it so that if it can't get 1000, it tries for 256.
2913 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2914 was an oversight, and furthermore, ever since the addition of extra
2915 controls (e.g. 4.43/32), the checks on when to allow different forms of
2916 "control" were broken. There should now be diagnostics for all cases when a
2917 control that does not make sense is encountered.
2919 7. $recipients is now available in the predata ACL (oversight).
2921 8. Tidy the search cache before the fork to do a delivery from a message
2922 received from the command line. Otherwise the child will trigger a lookup
2923 failure and thereby defer the delivery if it tries to use (for example) a
2924 cached ldap connection that the parent has called unbind on.
2926 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2927 of $address_data from the recipient verification was clobbered by the
2928 sender verification.
2930 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2931 was its contents. (It was OK if the option was not defined at all.)
2933 11. A "Completed" log line is now written for messages that are removed from
2934 the spool by the -Mrm option.
2936 12. $host_address is now set to the target address during the checking of
2937 ignore_target_hosts.
2939 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2940 being passed; this would have caused $sender_host_name to have been used if
2941 matching the list had actually called for a host name (not very likely,
2942 since this list is usually IP addresses). A host name is now passed as
2945 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2946 code that specifies a non-restarting handler (typically sigaction() in
2947 modern systems) in an attempt to fix a rare and obscure crash bug.
2949 15. Narrowed the window for a race in the daemon that could cause it to ignore
2950 SIGCHLD signals. This is not a major problem, because they are used only to
2951 wake it up if nothing else does.
2953 16. A malformed maildirsize file could cause Exim to calculate negative values
2954 for the mailbox size or file count. Odd effects could occur as a result.
2955 The maildirsize information is now recalculated if the size or filecount
2958 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2959 support for a long time. Removed HAVE_SYS_VFS_H.
2961 18. Updated exipick to current release from John Jetmore.
2963 19. Allow an empty sender to be matched against a lookup in an address list.
2964 Previously the only cases considered were a regular expression, or an
2967 20. Exim went into a mad DNS lookup loop when doing a callout where the
2968 host was specified on the transport, if the DNS lookup yielded more than
2971 21. The RFC2047 encoding function was originally intended for short strings
2972 such as real names; it was not keeping to the 75-character limit for
2973 encoded words that the RFC imposes. It now respects the limit, and
2974 generates multiple encoded words if necessary. To be on the safe side, I
2975 have increased the buffer size for the ${rfc2047: expansion operator from
2978 22. Failure to deliver a bounce message always caused it to be frozen, even if
2979 there was an errors_to setting on the router. The errors_to setting is now
2982 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2983 canonical form (fully expanded) before being placed in
2984 $sender_host_address.
2986 24. Updated eximstats to version 1.33
2988 25. Include certificate and key file names in error message when GnuTLS fails
2989 to set them up, because the GnuTLS error message doesn't include the name
2990 of the failing file when there is a problem reading it.
2992 26. Expand error message when OpenSSL has problems setting up cert/key files.
2995 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2996 (this can affect the format of dates).
2998 28. exim_tidydb, when checking for the continued existence of a message for
2999 which it has found a message-specific retry record, was not finding
3000 messages that were in split spool directories. Consequently, it was
3001 deleting retry records that should have stayed in existence.
3003 29. eximstats updated to version 1.35
3004 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3005 1.35 - bugfix such that pie charts by volume are generated correctly
3007 30. The SPA authentication driver was not abandoning authentication and moving
3008 on to the next authenticator when an expansion was forced to fail,
3009 contradicting the general specification for all authenticators. Instead it
3010 was generating a temporary error. It now behaves as specified.
3012 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3013 (the order specifies the preference for clients). The order is now AES256,
3014 AES128, 3DES, ARCFOUR128.
3016 31. Small patch to Sieve code - explicitly set From: when generating an
3019 32. Exim crashed if a remote delivery caused a very long error message to be
3020 recorded - for instance if somebody sent an entire SpamAssassin report back
3021 as a large number of 550 error lines. This bug was coincidentally fixed by
3022 increasing the size of one of Exim's internal buffers (big_buffer) that
3023 happened as part of the Exiscan merge. However, to be on the safe side, I
3024 have made the code more robust (and fixed the comments that describe what
3027 33. Some experimental protocols are using DNS PTR records for new purposes. The
3028 keys for these records are domain names, not reversed IP addresses. The
3029 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3030 leaves it alone. Component reversal etc. now happens only for IP addresses.
3033 34. The host_aton() function is supposed to be passed a string that is known
3034 to be a valid IP address. However, in the case of IPv6 addresses, it was
3035 not checking this. This is a hostage to fortune. Exim now panics and dies
3036 if the condition is not met. A case was found where this could be provoked
3037 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3038 components; fortuitously, this particular loophole had already been fixed
3039 by change 4.50/55 or 4.44/33 above.
3041 If there are any other similar loopholes, the new check in host_aton()
3042 itself should stop them being exploited. The report I received stated that
3043 data on the command line could provoke the exploit when Exim was running as
3044 exim, but did not say which command line option was involved. All I could
3045 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3046 running as the user.
3049 35. There was a buffer overflow vulnerability in the SPA authentication code
3050 (which came originally from the Samba project). I have added a test to the
3051 spa_base64_to_bits() function which I hope fixes it.
3054 36. The daemon start-up calls getloadavg() while still root for those OS that
3055 need the first call to be done as root, but it missed one case: when
3056 deliver_queue_load_max is set with deliver_drop_privilege. This is
3057 necessary for the benefit of the queue runner, because there is no re-exec
3058 when deliver_drop_privilege is set.
3060 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3061 was in use, was not putting the data itself into the right store pool;
3062 consequently, it could be overwritten for a subsequent message in the same
3063 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3066 38. Sometimes the final signoff response after QUIT could fail to get
3067 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3068 < 0 before doing a fflush(). This bug looks as though it goes back to the
3069 introduction of TLS in release 3.20, but "sometimes" must have been rare
3070 because the tests only now provoked it.
3076 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3077 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3078 Now it returns an integer. A number of calls were still expecting a T/F
3079 return. Fortuitously, in all cases, the tests worked in OK situations,
3080 which is the norm. However, things would have gone wrong on any write
3081 failures on the smtp file descriptor. This function is used when sending
3082 messages over SMTP and also when doing verify callouts.
3084 2. When Exim is called to do synchronous delivery of a locally submitted
3085 message (the -odf or -odi options), it no longer closes stderr before doing
3088 3. Implemented the mua_wrapper option.
3090 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3092 5. Implemented the functions header_remove(), header_testname(),
3093 header_add_at_position(), and receive_remove_recipient(), and exported them
3096 6. If an ACL "warn" statement specified the addition of headers, Exim already
3097 inserted X-ACL-Warn: at the start if there was no header name. However, it
3098 was not making this test for the second and subsequent header lines if
3099 there were newlines in the string. This meant that an invalid header could
3100 be inserted if Exim was badly configured.
3102 7. Allow an ACL "warn" statement to add header lines at the start or after all
3103 the Received: headers, as well as at the end.
3105 8. Added the rcpt_4xx retry error code.
3107 9. Added postmaster_mailfrom=xxx to callout verification option.
3109 10. Added mailfrom=xxxx to the callout verification option, for verify=
3112 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3113 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3115 12. Inserted some casts to stop certain compilers warning when using pointer
3116 differences as field lengths or precisions in printf-type calls (mostly
3117 affecting debugging statements).
3119 13. Added optional readline() support for -be (dynamically loaded).
3121 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3122 same clock tick as a message's arrival, so that its received time was the
3123 same as the "first fail" time on the retry record, and that message
3124 remained on the queue past the ultimate address timeout, every queue runner
3125 would try a delivery (because it was past the ultimate address timeout) but
3126 after another failure, the ultimate address timeout, which should have then
3127 bounced the address, did not kick in. This was a "< instead of <=" error;
3128 in most cases the first failure would have been in the next clock tick
3129 after the received time, and all would be well.
3131 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3132 being recognized when the domain list was tested by the match_domain
3133 condition in an expansion string.
3135 16. Added the ${str2b64: operator.
3137 17. Exim was always calling setrlimit() to set a large limit for the number of
3138 processes, without checking whether the existing limit was already
3139 adequate. (It did check for the limit on file descriptors.) Furthermore,
3140 errors from getrlimit() and setrlimit() were being ignored. Now they are
3141 logged to the main and panic logs, but Exim does carry on, to try to do its
3142 job under whatever limits there are.
3144 18. Imported PCRE 5.0.
3146 19. Trivial typo in log message " temporarily refused connection" (the leading
3149 20. If the log selector return_path_on_delivery was set and an address was
3150 redirected to /dev/null, the delivery process crashed because it assumed
3151 that a return path would always be set for a "successful" delivery. In this
3152 case, the whole delivery is bypassed as an optimization, and therefore no
3155 21. Internal re-arrangement: the function for sending a challenge and reading
3156 a response while authentication was assuming a zero-terminated challenge
3157 string. It's now changed to take a pointer and a length, to allow for
3158 binary data in such strings.
3160 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3162 23. Exim was not respecting finduser_retries when seeking the login of the
3163 uid under which it was called; it was always trying 10 times. (The default
3164 setting of finduser_retries is zero.) Also, it was sleeping after the final
3165 failure, which is pointless.
3167 24. Implemented tls_on_connect_ports.
3169 25. Implemented acl_smtp_predata.
3171 26. If the domain in control=submission is set empty, Exim assumes that the
3172 authenticated id is a complete email address when it generates From: or
3173 Sender: header lines.
3175 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3176 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3177 chown and chgrp in /bin and hostname in /usr/bin.
3179 28. Exim was keeping the "process log" file open after each use, just as it
3180 does for the main log. This opens the possibility of it remaining open for
3181 long periods when the USR1 signal hits a daemon. Occasional processlog
3182 errors were reported, that could have been caused by this. Anyway, it seems
3183 much more sensible not to leave this file open at all, so that is what now
3186 29. The long-running daemon process does not normally write to the log once it
3187 has entered its main loop, and it closes the log before doing so. This is
3188 so that log files can straightforwardly be renamed and moved. However,
3189 there are a couple of unusual error situations where the daemon does write
3190 log entries, and I had neglected to close the log afterwards.
3192 30. The text of an SMTP error response that was received during a remote
3193 delivery was being truncated at 512 bytes. This is too short for some of
3194 the long messages that one sometimes sees. I've increased the limit to
3197 31. It is now possible to make retry rules that apply only when a message has a
3198 specific sender, in particular, an empty sender.
3200 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3201 it possible to be selective about when SMTP synchronization is enforced.
3203 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3205 32. Implemented hosts_connection_nolog.
3207 33. Added an ACL for QUIT.
3209 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3212 35. Added mailbox_size and mailbox_filecount to appendfile.
3214 36. Added control = no_multiline_responses to ACLs.
3216 37. There was a bug in the logic of the code that waits for the clock to tick
3217 in the case where the clock went backwards by a substantial amount such
3218 that the microsecond fraction of "now" was more than the microsecond
3219 fraction of "then" (but the whole seconds number was less).
3221 38. Added support for the libradius Radius client library this is found on
3222 FreeBSD (previously only the radiusclient library was supported).
3228 1. When certain lookups returned multiple values in the form name=value, the
3229 quoting of the values was not always being done properly. Specifically:
3230 (a) If the value started with a double quote, but contained no whitespace,
3232 (b) If the value contained whitespace other than a space character (i.e.
3233 tabs or newlines or carriage returns) it was not quoted.
3234 This fix has been applied to the mysql and pgsql lookups by writing a
3235 separate quoting function and calling it from the lookup code. The fix
3236 should probably also be applied to nisplus, ibase and oracle lookups, but
3237 since I cannot test any of those, I have not disturbed their existing code.
3239 2. A hit in the callout cache for a specific address caused a log line with no
3240 reason for rejecting RCPT. Now it says "Previous (cached) callout
3241 verification failure".
3243 3. There was an off-by-one bug in the queryprogram router. An over-long
3244 return line was truncated at 256 instead of 255 characters, thereby
3245 overflowing its buffer with the terminating zero. As well as fixing this, I
3246 have increased the buffer size to 1024 (and made a note to document this).
3248 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3249 when Exim is waiting for an SMTP response from a remote server, Exim
3250 restarts its select() call on the socket, thereby resetting its timeout.
3251 This is not a problem when such interrupts are rare. Somebody set up a cron
3252 job to run exiwhat every 2 minutes, which is less than the normal select()
3253 timeout (5 or 10 minutes). This meant that the select() timeout never
3254 kicked in because it was always reset. I have fixed this by comparing the
3255 time when an interrupt arrives with the time at the start of the first call
3256 to select(). If more time than the timeout has elapsed, the interrupt is
3257 treated as a timeout.
3259 5. Some internal re-factoring in preparation for the addition of Sieve
3260 extensions (by MH). In particular, the "personal" test is moved to a
3261 separate function, and given an option for scanning Cc: and Bcc: (which is
3262 not set for Exim filters).
3264 6. When Exim created an email address using the login of the caller as the
3265 local part (e.g. when creating a From: or Sender: header line), it was not
3266 quoting the local part when it contained special characters such as @.
3268 7. Installed new OpenBSD configuration files.
3270 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3271 try to make them clearer.
3273 9. Callout options, other than the timeout value, were being ignored when
3274 verifying sender addresses in header lines. For example, when using
3276 verify = header_sender/callout=no_cache
3278 the cache was (incorrectly) being used.
3280 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3281 only the Cygwin environment.
3283 11. When return_path_on_delivery was set as a log selector, if different remote
3284 addresses in the same message used different return paths and parallel
3285 remote delivery occurred, the wrong values would sometimes be logged.
3286 (Whenever a remote delivery process finished, the return path value from
3287 the most recently started remote delivery process was logged.)
3289 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3290 lines when AUTH and/or TLS are in use. This is the "received protocol"
3291 field. Exim used to use "asmtp" for authenticated SMTP, without any
3292 indication (in the protocol name) for TLS use. Now it follows the RFC and
3293 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3294 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3295 names appear in log lines as well as in Received: header lines.
3297 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3298 extensions, and comparison tests, and to fix some bugs.
3300 14. Changes to the "personal" filter test:
3302 (1) The test was buggy in that it was just doing the equivalent of
3303 "contains" tests on header lines. For example, if a user's address was
3304 anne@some.where, the "personal" test would incorrectly be true for
3306 To: susanne@some.where
3308 This test is now done by extracting each address from the header in turn,
3309 and checking the entire address. Other tests that are part of "personal"
3310 are now done using regular expressions (for example, to check local parts
3311 of addresses in From: header lines).
3313 (2) The list of non-personal local parts in From: addresses has been
3314 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3315 taken from the Sieve specification recommendations.
3317 (3) If the message contains any header line starting with "List-" it is
3318 treated as non-personal.
3320 (4) The test for "circular" in the Subject: header line has been removed
3321 because it now seems ill-conceived.
3323 15. Minor typos in src/EDITME comments corrected.
3325 16. Installed latest exipick from John Jetmore.
3327 17. If headers_add on a router specified a text string that was too long for
3328 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3329 of string_sprintf() is now avoided.
3331 18. $message_body_size was not set (it was always zero) when running the DATA
3332 ACL and the local_scan() function.
3334 19. For the "mail" command in an Exim filter, no default was being set for
3335 the once_repeat time, causing a random time value to be used if "once" was
3336 specified. (If the value happened to be <= 0, no repeat happened.) The
3337 default is now 0s, meaning "never repeat". The "vacation" command was OK
3338 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3339 (I found it when inspecting the code).
3341 20. There is now an overall timeout for performing a callout verification. It
3342 defaults to 4 times the callout timeout, which applies to individual SMTP
3343 commands during the callout. The overall timeout applies when there is more
3344 than one host that can be tried. The timeout is checked before trying the
3345 next host. This prevents very long delays if there are a large number of
3346 hosts and all are timing out (e.g. when the network connections are timing
3347 out). The value of the overall timeout can be changed by specifying an
3348 additional sub-option for "callout", called "maxwait". For example:
3350 verify = sender/callout=5s,maxwait=20s
3352 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3353 to the end before writing, but this should make it even safer).
3355 22. Exim was forgetting that it had advertised PIPELINING for the second and
3356 subsequent messages on an SMTP connection. It was also not resetting its
3357 memory on STARTTLS and an internal HELO.
3359 23. When Exim logs an SMTP synchronization error within a session, it now
3360 records whether PIPELINING has been advertised or not.
3362 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3363 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3364 rather than long int.
3366 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3368 26. Added the never_mail option to autoreply.
3374 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3375 crash if the getsockname() call failed; this can happen if a connection is
3376 closed very soon after it is established. The problem was simply in the
3377 order in which certain operations were done, causing Exim to try to write
3378 to the SMTP stream before it had set up the file descriptor. The bug has
3379 been fixed by making things happen in the correct order.
3385 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3386 before the connection was closed, thus losing the rejection response.
3388 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3389 some early Solaris releases, but causes trouble in current releases where
3390 socklen_t is defined.
3392 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3395 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3398 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3399 that releasing the top of it at the end releases what was used for sub-
3400 expansions (unless the block got too big). However, discard this block if
3401 the first thing is a variable or header, so that we can use its block when
3402 it is dynamic (useful for very large $message_headers, for example).
3404 6. Lookups now cache *every* query, not just the most recent. A new, separate
3405 store pool is used for this. It can be recovered when all lookup caches are
3406 flushed. Lookups now release memory at the end of their result strings.
3407 This has involved some general refactoring of the lookup sources.
3409 7. Some code has been added to the store_xxx() functions to reduce the amount
3410 of flapping under certain conditions.
3412 8. log_incoming_interface used to affect only the <= reception log lines. Now
3413 it causes the local interface and port to be added to several more SMTP log
3414 lines, for example "SMTP connection from", and rejection lines.
3416 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3418 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3420 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3421 could be overwritten at the end of the current message (or the start of a
3422 new message if it was set in a HELO ACL). The value is now preserved for
3423 the duration of the SMTP connection.
3425 12. If a transport had a headers_rewrite setting, and a matching header line
3426 contained an unqualified address, that address was qualified, even if it
3427 did not match any rewriting rules. The underlying bug was that the values
3428 of the flags that permit the existence of unqualified sender and recipient
3429 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3430 non-local messages, and by -bnq for local messages) were not being
3431 preserved with the message after it was received.
3433 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3434 "next input=" as part of the text comprising the host identity instead of
3435 the correct text. The code was using the same buffer for two different
3436 strings. However, depending on which order the printing function evaluated
3437 its arguments, the bug did not always show up. Under Linux, for example, my
3438 test suite worked just fine.
3440 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3441 doesn't work with some older versions of Perl. It has been changed to "my",
3442 which in any case is probably the better facility to use.
3444 15. A really picky compiler found some instances of statements for creating
3445 error messages that either had too many or two few arguments for the format
3448 16. The size of the buffer for calls to the DNS resolver has been increased
3449 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3450 for addresses that have a lot of PTR records. This alleviates a problem; it
3451 does not fully solve it.
3453 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3454 buffer now truncates the list and logs the incident, which is the same
3455 action as happens when Exim is looking up a host name and its aliases.
3456 Previously in this situation something unpredictable would happen;
3457 sometimes it was "internal error: store_reset failed".
3459 18. If a server dropped the connection unexpectedly when an Exim client was
3460 using GnuTLS and trying to read a response, the client delivery process
3461 crashed while trying to generate an error log message.
3463 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3464 string, for example:
3466 warn message = H1: something\nH2: something
3468 the text was added as a single header line from Exim's point of view
3469 though it ended up OK in the delivered message. However, searching for the
3470 second and subsequent header lines using $h_h2: did not work. This has been
3471 fixed. Similarly, if a system filter added multiple headers in this way,
3472 the routers could not see them.
3474 20. Expanded the error message when iplsearch is called with an invalid key to
3475 suggest using net-iplsearch in a host list.
3477 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3478 ACLs are no longer actually imposed (and a message to that effect is
3481 22. If a "gecos" field in a passwd entry contained escaped characters, in
3482 particular, if it contained a \" sequence, Exim got it wrong when building
3483 a From: or a Sender: header from that name. A second bug also caused
3484 incorrect handling when an unquoted " was present following a character
3485 that needed quoting.
3487 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3488 was not being matched caselessly.
3490 24. Arranged for all hyphens in the exim.8 source to be escaped with
3493 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3494 properly. Recipient callout cache records were still being keyed to include
3495 the sender, even when use_sender was set false. This led to far more
3496 callouts that were necessary. The sender is no longer included in the key
3497 when use_sender is false.
3499 26. Added "control = submission" modifier to ACLs.
3501 27. Added the ${base62d: operator to decode base 62 numbers.
3503 28. dnsdb lookups can now access SRV records.
3505 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3506 the configuration file.
3508 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3509 (-v) mode. This makes the output for a verbose queue run more intelligible.
3511 31. Added a use_postmaster feature to recipient callouts.
3513 32. Added the $body_zerocount variable, containing the number of binary zero
3514 bytes in the message body.
3516 33. The time of last modification of the "new" subdirectory is now used as the
3517 "mailbox time last read" when there is a quota error for a maildir
3520 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3522 35. Added +ignore_unknown as a special item in host lists.
3524 36. Code for decoding IPv6 addresses in host lists is now included, even if
3525 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3526 address was recognized as an IP address, but was then not correctly decoded
3527 into binary, causing unexpected and incorrect effects when compared with
3534 1. Very minor rewording of debugging text in manualroute to say "list of
3535 hosts" instead of "hostlist".
3537 2. If verify=header_syntax was set, and a header line with an unqualified
3538 address (no domain) and a large number of spaces between the end of the
3539 name and the colon was received, the reception process suffered a buffer
3540 overflow, and (when I tested it) crashed. This was caused by some obsolete
3541 code that should have been removed. The fix is to remove it!
3543 3. When running in the test harness, delay a bit after writing a bounce
3544 message to get a bit more predictability in the log output.
3546 4. Added a call to search_tidyup() just before forking a reception process. In
3547 theory, someone could use a lookup in the expansion of smtp_accept_max_
3548 per_host which, without the tidyup, could leave open a database connection.
3550 5. Added the variables $recipient_data and $sender_data which get set from a
3551 lookup success in an ACL "recipients" or "senders" condition, or a router
3552 "senders" option, similar to $domain_data and $local_part_data.
3554 6. Moved the writing of debug_print from before to after the "senders" test
3557 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3558 problems for message scanning, either using a data ACL, or using
3559 local_scan() because the Received: header was not generated till after they
3560 were called (in order to set the time as the time of reception completion).
3561 I have revised the way this works. The header is now generated after the
3562 body is received, but before the ACL or local_scan() are called. After they
3563 are run, the timestamp in the header is updated.
3569 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3570 before starting a queue runner without re-exec. This happened only when
3571 deliver_drop_privilege was set or when the Exim user was set to root. The
3572 effect of the bug was that timeouts during subsequent deliveries caused
3573 crashes instead of being properly handled. The handler is now left at its
3574 default (and expected) setting.
3576 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3577 message, again when deliver_drop_privilege is set or Exim is run as root.
3578 The bug described in (1) was not present in this case, but the tidying up
3579 of the other signals was missing. I have made the two cases consistent.
3581 3. The ignore_target_hosts setting on a manualroute router was being ignored
3582 for hosts that were looked up using the /MX notation.
3584 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3587 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3588 operated on the sender address. After changing the $sender_address to <>
3589 for the sender address verify, Exim was re-instated it as the original
3590 (before rewriting) address, but remembering that it had rewritten it, so it
3591 wasn't rewriting it again. This bug also had the effect of breaking the
3592 sender address verification caching when the sender address was rewritten.
3594 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3595 This has been changed so that if the ip literal address matches
3596 ignore_target_hosts, the router declines.
3598 7. Added expansion conditions match_domain, match_address, and match_local_
3599 part (NOT match_host).
3601 8. The placeholder for the Received: header didn't have a length field set.
3603 9. Added code to Exim itself and to exim_lock to test for a specific race
3604 condition that could lead to file corruption when using MBX delivery. The
3605 issue is with the lockfile that is created in /tmp. If this file is removed
3606 after a process has opened it but before that process has acquired a lock,
3607 there is the potential for a second process to recreate the file and also
3608 acquire a lock. This could lead to two Exim processes writing to the file
3609 at the same time. The added code performs the same test as UW imapd; it
3610 checks after acquiring the lock that its file descriptor still refers to
3611 the same named file.
3613 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3614 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3617 11. Added the smtp_active_hostname option. If used, this will typically be made
3618 to depend on the incoming interface address. Because $interface_address is
3619 not set up until the daemon has forked a reception process, error responses
3620 that can happen earlier (such as "too many connections") no longer contain
3623 12. If an expansion in a condition on a "warn" statement fails because a lookup
3624 defers, the "warn" statement is abandoned, and the next ACL statement is
3625 processed. Previously this caused the whole ACL to be aborted.
3627 13. Added the iplsearch lookup type.
3629 14. Added ident_timeout as a log selector.
3631 15. Added tls_certificate_verified as a log selector.
3633 16. Added a global option tls_require_ciphers (compare the smtp transport
3634 option of the same name). This controls incoming TLS connections.
3636 17. I finally figured out how to make tls_require_ciphers do a similar thing
3637 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3638 before starting the TLS session.
3640 18. Tabs are now shown as \t in -bP output.
3642 19. If the log selector return_path_on_delivery was set, Exim crashed when
3643 bouncing a message because it had too many Received: header lines.
3645 20. If two routers both had headers_remove settings, and the first one included
3646 a superfluous trailing colon, the final name in the first list and the
3647 first name in the second list were incorrectly joined into one item (with a
3648 colon in the middle).
3654 1. Added -C and -D options to the exinext utility, mainly to make it easier
3655 to include in the automated testing, but these could be helpful when
3656 multiple configurations are in use.
3658 2. The exinext utility was not formatting the output nicely when there was
3659 an alternate port involved in the retry record key, nor when there was a
3660 message id as well (for retries that were specific to a specific message
3661 and a specific host). It was also confused by IPv6 addresses, because of
3662 the additional colons they contain. I have fixed the IPv4 problem, and
3663 patched it up to do a reasonable job for IPv6.
3665 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3666 delivery, the log line now contains "pipelined" if PIPELINING was used.
3668 4. An SMTP transport process used to panic and die if the bind() call to set
3669 an explicit outgoing interface failed. This has been changed; it is now
3670 treated in the same way as a connect() failure.
3672 5. A reference to $sender_host_name in the part of a conditional expansion
3673 that was being skipped was still causing a DNS lookup. This no longer
3676 6. The def: expansion condition was not recognizing references to header lines
3677 that used bh_ and bheader_.
3679 7. Added the _cache feature to named lists.
3681 8. The code for checking quota_filecount in the appendfile transport was
3682 allowing one more file than it should have been.
3684 9. For compatibility with Sendmail, the command line option
3692 and sets the incoming protocol and host name (for trusted callers). The
3693 host name and its colon can be omitted when only the protocol is to be set.
3694 Note the Exim already has two private options, -pd and -ps, that refer to
3695 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3696 "s", but I don't think that's a major issue.
3698 10. A number of refactoring changes to the code, none of which should affect
3701 (a) The number of logging options was getting close to filling up the
3702 32-bit word that was used as a bit map. I have split them into two classes:
3703 those that are passed in the argument to log_write(), and those that are
3704 only ever tested independently outside of that function. These are now in
3705 separate 32-bit words, so there is plenty of room for expansion again.
3706 There is no change in the user interface or the logging behaviour.
3708 (b) When building, for example, log lines, the code previously used a
3709 macro that called string_cat() twice, in order to add two strings. This is
3710 not really sufficiently general. Furthermore, there was one instance where
3711 it was actually wrong because one of the argument was used twice, and in
3712 one call a function was used. (As it happened, calling the function twice
3713 did not affect the overall behaviour.) The macro has been replaced by a
3714 function that can join an arbitrary number of extra strings onto a growing
3717 (c) The code for expansion conditions now uses a table and a binary chop
3718 instead of a serial search (which was left over from when there were very
3719 few conditions). Also, it now recognizes conditions like "pam" even when
3720 the relevant support is not compiled in: a suitably worded error message is
3721 given if an attempt is made to use such a condition.
3723 11. Added ${time_interval:xxxxx}.
3725 12. A bug was causing one of the ddress fields not to be passed back correctly
3726 from remote delivery subprocesses. The field in question was not being
3727 subsequently used, so this caused to problems in practice.
3729 13. Added new log selectors queue_time and deliver_time.
3731 14. Might have fixed a bug in maildirsizefile handling that threw up
3732 "unexpected character" debug warnings, and recalculated the data
3733 unnecessarily. In any case, I expanded the warning message to give more
3736 15. Added the message "Restricted characters in address" to the statements in
3737 the default ACL that block characters like @ and % in local parts.
3739 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3740 Three changes have been made:
3742 (a) There was a serious bug; a negative response to MAIL caused the whole
3743 recipient domain to be cached as invalid, thereby blocking all messages
3744 to all local parts at the same domain, from all senders. This bug has
3745 been fixed. The domain is no longer cached after a negative response to
3746 MAIL if the sender used is not empty.
3748 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3751 (c) A new callout option, "use_sender" has been added for people who want
3752 the modified behaviour.
3758 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3761 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3762 indeed breaks things for older releases.
3764 3. Added additional logging to the case where there is a problem reading data
3765 from a filter that is running in a subprocess using a pipe, in order to
3766 try to track down a specific problem.
3768 4. Testing facility fudge: when running in the test harness and attempting
3769 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3770 getting "No route to host". Convert this to a timeout.
3772 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3775 6. Some OS don't have socklen_t but use size_t instead. This affects the
3776 fifth argument of getsockopt() amongst other things. This is now
3777 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3778 can be set for individual OS. I have set it for SunOS5, OSF1, and
3779 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3780 some earlier ones do not.
3782 7. Change 4.30/15 was not doing the test caselessly.
3784 8. The standard form for an IPv6 address literal was being rejected by address
3785 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3786 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3787 this, as well as the form without the "IPv6" on the front (but only when
3788 address literals are enabled, of course).
3790 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3792 10. Exim crashed if a message with an empty sender address specified by -f
3793 encountered a router with an errors_to setting. This could be provoked only
3794 by a command such as
3798 where an empty string was supplied; "<>" did not hit this bug.
3800 11. Installed PCRE release 4.5.
3802 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3803 remained set. It is now erased.
3805 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3806 times from message ids (which are base 36 rather than the normal 62).
3808 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3809 were being counted as actual protocol errors, and logged if the log
3810 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3811 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3812 and DATA following a set of rejected RCPTs do not count as protocol errors.
3813 In other words, Exim assumes they were pipelined, though this may not
3814 actually be the case. Of course, in all cases the client gets an
3815 appropriate error code.
3817 15. If a lookup fails in an ACL condition, a message about the failure may
3818 be available; it is used if testing the ACL cannot continue, because most
3819 such messages specify what the cause of the deferral is. However, some
3820 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3821 that caused an old message to be retained and used if a later statement
3822 caused a defer, replacing the real cause of the deferral.
3824 16. If an IP address had so many PTR records that the DNS lookup buffer
3825 was not large enough to hold them, Exim could crash while trying to process
3826 the truncated data. It now detects and logs this case.
3828 17. Further to 4.21/58, another change has been made: if (and only if) the
3829 first line of a message (the first header line) ends with CRLF, a bare LF
3830 in a subsequent header line has a space inserted after it, so as not to
3831 terminate the header.
3833 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3834 unnecessarily, used atoi() instead of strtol(), and didn't check the
3835 termination when getting file sizes from file names by regex.
3837 19. Completely re-implemented the support for maildirsize files, in the light
3838 of a number of problems with the previous contributed implementation
3839 (4.30/29). In particular:
3841 . If the quota is zero, the maildirsize file is maintained, but no quota is
3844 . If the maildir directory does not exist, it is created before any attempt
3845 to write a maildirsize file.
3847 . The quota value in the file is just a cache; if the quota is changed in
3848 the transport, the new value overrides.
3850 . A regular expression is available for excluding directories from the
3853 20. The autoreply transport checks the characters in options that define the
3854 message's headers; it allows continued headers, but it was checking with
3855 isspace() after an embedded newline instead of explicitly looking for a
3858 21. If all the "regular" hosts to which an address was routed had passed their
3859 expiry times, and had not reached their retry times, the address was
3860 bounced, even if fallback hosts were defined. Now Exim should go on to try
3863 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3864 equivalent code in the SMTP transport. Some hosts send humungous responses
3865 to HELO/EHLO, more than 1024 it seems.
3867 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3868 gives compiler warnings in some environments. I've now done it "properly",
3871 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3872 (because of problems with the built-in one) was declared to return uschar *
3873 instead of char *, causing compiler failure.
3875 25. Fixed a file descriptor leak when processing alias/forward files.
3877 26. Fixed a minor format string issue in dbfn.c.
3879 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3881 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3882 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3885 29. When a sender address is verified, it is cached, to save repeating the test
3886 when there is more than one recipient in a message. However, when the
3887 verification involves a callout, it is possible for different callout
3888 options to be set for different recipients. It is too complicated to keep
3889 track of this in the cache, so now Exim always runs a verification when a
3890 callout is required, relying on the callout cache for the optimization.
3891 The overhead is duplication of the address routing, but this should not be
3894 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3895 to be verified with callout=postmaster, and the main callout worked but the
3896 postmaster check failed, the verification correctly failed. However, if a
3897 subsequent RCPT command asked for sender verification *without* the
3898 postmaster check, incorrect caching caused this verification also to fail,
3901 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3902 it was not caching the DNS options (qualify_single, search_parents) that
3903 were used when the lookup failed. A subsequent lookup with different
3904 options therefore always gave the same answer, though there were cases
3905 where it should not have. (Example: a "domains = !$mx_any" option on a
3906 dnslookup router: the "domains" option is always processed without any
3907 widening, but the router might have qualify_single set.) Now Exim uses the
3908 cached value only when the same options are set.
3910 32. Added John Jetmore's "exipick" utility to the distribution.
3912 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3913 than a timeout (e.g. a certificate is required, and is not provided), an
3914 Exim server now closes the connection immediately. Previously it waited for
3915 the client to close - but if the client is SSL, it seems that they each
3916 wait for each other, leading to a delay before one of them times out.
3918 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3919 maintained 0.8.x compatibility because I don't think many are using it, and
3920 it is clearly obsolete.
3922 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3925 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3926 client certificate was expired. A simple patch fixes this, though I don't
3927 understand the full logic of why the verify callback is called multiple
3930 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3931 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3932 which causes problems with some clients (such as the Certicom SSL Plus
3933 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3934 disables the coutermeasure allowing Eudora to connect."
3936 38. Exim was not checking that a write() to a log file succeeded. This could
3937 lead to Bad Things if a log got too big, in particular if it hit a file
3938 size limit. Exim now panics and dies if it cannot write to a log file, just
3939 as it does if it cannot open a log file.
3941 39. Modified OS/Makefile-Linux so that it now contains
3943 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3945 The two -D definitions ensure that Exim is compiled with large file
3946 support, which makes it possible to handle log files that are bigger than
3949 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3950 instance) a domain was checked against a named list that involved a lookup,
3951 causing $domain_data to be set, then another domain was checked against the
3952 same list, then the first domain was re-checked, the value of $domain_data
3953 after the final check could be wrong. In particular, if the second check
3954 failed, it could be set empty. This bug probably also applied to
3957 41. The strip_trailing_dot option was not being applied to the address given
3958 with the -f command-line option.
3960 42. The code for reading a message's header from the spool was incrementing
3961 $received_count, but never initializing it. This meant that the value was
3962 incorrect (doubled) while delivering a message in the same process in which
3963 it was received. In the most common configuration of Exim, this never
3964 happens - a fresh exec is done - but it can happen when
3965 deliver_drop_privilege is set.
3967 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3968 it now includes up to 150 characters of the unexpected data in the log
3971 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3972 and building data strings. The size of both of these buffers was 10 000
3973 bytes - far larger than anybody would *ever* want, thought I. Needless to
3974 say, somebody hit the limit. I have increased the maximum line length to
3975 20 000 and the maximum data length of concatenated lines to 100 000. I have
3976 also fixed two bugs, because there was no checking on these buffers. Tsk,
3977 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3978 buffer is too small.
3980 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3981 lsearch lookups. Now it does.
3983 46. When parsing a route_list item in a manualroute router, a fixed-length
3984 buffer was used for the list of hosts. I made this 1024 bytes long,
3985 thinking that nobody would ever have a list of hosts that long. Wrong.
3986 Somebody had a whole pile of complicated expansion conditions, and the
3987 string was silently truncated, leading to an expansion error. It turns out
3988 that it is easier to change to an unlimited length (owing to other changes
3989 that have happened since this code was originally written) than to build
3990 structure for giving a limitation error. The length of the item that
3991 expands into the list of hosts is now unlimited.
3993 47. The lsearch lookup could not handle data where the length of text line was
3994 more than 4095 characters. Such lines were truncated, leading to shortened
3995 data being returned. It should now handle lines of any length.
3997 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3998 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4001 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4002 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4003 no longer happens in comment lines. A list of replacements is now placed
4004 at the head of all of the source files, except those whose only change is
4005 to replace PERL_COMMAND in the very first #! line.
4007 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4008 messages on the queue, with a bottom-up merge sort, using code contributed
4009 by Michael Haardt. This should make operations like -bp somewhat faster on
4010 large queues. It won't affect queue runners, except when queue_run_in_order
4013 51. Installed eximstats 1.31 in the distribution.
4015 52. Added support for SRV lookups to the dnslookup router.
4017 53. If an ACL referred to $message_body or $message_body_end, the value was not
4018 reset for any messages that followed in the same SMTP session.
4020 54. The store-handling optimization for building very long strings was not
4021 differentiating between the different store pools. I don't think this
4022 actually made any difference in practice, but I've tidied it.
4024 55. While running the routers to verify a sender address, $sender_address
4025 was still set to the sender address. This is wrong, because when routing to
4026 send a bounce to the sender, it would be empty. Therefore, I have changed
4027 it so that, while verifying a sender address, $sender_address is set to <>.
4028 (There is no change to what happens when verifying a recipient address.)
4030 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4031 target A or AAAA records (if not already returned) without resetting the
4032 qualify_single or search_parents options of the DNS resolver. These are
4033 inappropriate in this case because the targets of MX and SRV records must
4034 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4035 target that, when qualified, matched something in the local domain. These
4036 two options are now turned off when doing these lookups.
4038 57. It seems that at least some releases of Reiserfs (which does not have the
4039 concept of a fixed number of inodes) returns zero and not -1 for the
4040 number of available inodes. This interacted badly with check_spool_inodes,
4041 which assumed that -1 was the "no such thing" setting. What I have done is
4042 to check that the total number of inodes is greater than zero before doing
4043 the test of how many are available.
4045 58. When a "warn" ACL statement has a log_message modifier, the message is
4046 remembered, and not repeated. This is to avoid a lot of repetition when a
4047 message has many recipients that cause the same warning to be written.
4048 Howewer, Exim was preserving the list of already written lines for an
4049 entire SMTP session, which doesn't seem right. The memory is now reset if a
4050 new message is started.
4052 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4053 debugging output unless log_rewrite was also set.
4055 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4056 of (int)(handle) when we know that handle contains (void *)(-1).
4058 61. The Exim daemon panic-logs an error return when it closes the incoming
4059 connection. However "connection reset by peer" seems to be common, and
4060 isn't really an error worthy of noting specially, so that particular error
4063 62. When Exim is trying to find all the local interfaces, it used to panic and
4064 die if the ioctl to get the interface flags failed. However, it seems that
4065 on at least one OS (Solaris 9) it is possible to have an interface that is
4066 included in the list of interfaces, but for which you get a failure error
4067 for this call. This happens when the interface is not "plumbed" into a
4068 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4069 failure of the "get flags" call assumes that the interface is down.
4071 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4072 makes life easier for people who are doing arithmetic on fields extracted
4073 from dates, where you often get leading zeros that should not be
4074 interpreted as octal.
4076 64. Added qualify_domain to the redirect router, to override the global
4079 65. If a pathologically long header line contained very many addresses (the
4080 report of this problem mentioned 10 000) and each of them was rewritten,
4081 Exim could use up a very large amount of memory. (It kept on making new
4082 copies of the header line as it rewrote, and never released the old ones.)
4083 At the expense of a bit more processing, the header rewriting function has
4084 been changed so that it no longer eats memory in this way.
4086 66. The generation of the Received: header has been moved from the time that a
4087 message starts to be received, to the time that it finishes. The timestamp
4088 in the Received: header should now be very close to that of the <= log
4089 line. There are two side-effects of this change:
4091 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4092 logged header lines no longer include the local Received: line, because
4093 it has not yet been created. The same applies to a copy of the message
4094 that is returned to a non-SMTP sender when a message is rejected.
4096 (b) When a filter file is tested using -bf, no additional Received: header
4097 is added to the test message. After some thought, I decided that this
4100 This change does not affect the value of $received_for. It is still set
4101 after address rewriting, but before local_scan() is called.
4103 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4105 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4106 gave an unhelpful panic error message, and a defer error. I have managed to
4107 change this behaviour so that it now rejects any supplied certificate,
4108 which seems right, as the list of acceptable certificates is empty.
4110 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4111 gave an unhelpful defer error. I have not managed to make this reject any
4112 supplied certificates, but the error message it gives is "no certificate
4113 supplied", which is not helpful.
4115 70. exigrep's output now also includes lines that are not associated with any
4116 message, but which match the given pattern. Implemented by a patch from
4117 Martin Sluka, which also tidied up the Perl a bit.
4119 71. Recipient callout verification, like sender verification, was using <> in
4120 the MAIL FROM command. This isn't really the right thing, since the actual
4121 sender may affect whether the remote host accepts the recipient or not. I
4122 have changed it to use the actual sender in the callout; this means that
4123 the cache record is now keyed on a recipient/sender pair, not just the
4124 recipient address. There doesn't seem to be a real danger of callout loops,
4125 since a callout by the remote host to check the sender would use <>.
4126 [SEE ABOVE: changed after hitting problems.]
4128 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4129 temporary errors. However, in the case of such a code being given after
4130 the end of a data transmission (i.e. after ".") Exim was failing to write
4131 a retry record for the message. (Yes, there was some broken host that was
4132 actually sending 8xx at this point.)
4134 73. An unknown lookup type in a host list could cause Exim to panic-die when
4135 the list was checked. (An example that provoked this was putting <; in the
4136 middle of a list instead of at the start.) If this happened during a DATA
4137 ACL check, a -D file could be left lying around. This kind of configuration
4138 error no longer causes Exim to die; instead it causes a defer errror. The
4139 incident is still logged to the main and panic logs.
4141 74. Buglet left over from Exim 3 conversion. The message "too many messages
4142 in one connection" was written to the rejectlog but not the mainlog, except
4143 when address rewriting (yes!) was being logged.
4145 75. Added write_rejectlog option.
4147 76. When a system filter was run not as root (that is, when system_filter_user
4148 was set), the values of the $n variables were not being returned to the
4149 main process; thus, they were not subsequently available in the $sn
4152 77. Added +return_path_on_delivery log selector.
4154 78. A connection timeout was being treated differently from recipients deferred
4155 when testing hosts_max_try with a message that was older than the host's
4156 retry timeout. (The host should not be counted, thus allowing all hosts to
4157 be tried at least once before bouncing.) This may have been the cause of an
4158 occasionally reported bug whereby a message would remain on the queue
4159 longer than the retry timeout, but would be bounced if a delivery was
4160 forced. I say "may" because I never totally pinned down the problem;
4161 setting up timeout/retry tests is difficult. See also the next item.
4163 79. The ultimate address timeout was not being applied to errors that involved
4164 a combination of host plus message (for example, a timeout on a MAIL
4165 command). When an address resolved to a number of possible hosts, and they
4166 were not all tried for each delivery (e.g. because of hosts_max_try), a
4167 message could remain on the queue longer than the retry timeout.
4169 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4172 81. Fixed an obscure SMTP outgoing bug which required at least the following
4173 conditions: (a) there was another message waiting for the same server;
4174 (b) the server returned 5xx to all RCPT commands in the first message so
4175 that the message was not completed; (c) the server dropped the connection
4176 or gave a negative response to the RSET that Exim sends to abort the
4177 transaction. The observed case was a dropped connection after DATA that had
4178 been sent in pipelining mode. That is, the server had advertised PIPELINING
4179 but was not implementing it correctly. The effect of the bug was incorrect
4180 behaviour, such as trying another host, and this could lead to a crash.
4186 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4187 and daemon.c were passed as pointers to ints; they should have been
4188 pointers to socklen_t variables (which are typically unsigned ints).
4190 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4193 3. Fixed a really odd bug that affected only the testing scheme; patching a
4194 certain fixed string in the binary changed the value of another string that
4195 happened to be identical to the end of the original first string.
4197 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4198 name", it returns that address as the IP address. On some operating
4199 systems (e.g. Solaris), it also passes back the IP address string as the
4200 "host name". However, on others (e.g. Linux), it passes back an empty
4201 string. Exim wasn't checking for this, and was changing the host name to an
4202 empty string, assuming it had been canonicized.
4204 5. Although rare, it is permitted to have more than one PTR record for a given
4205 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4206 all the names associated with an address, because they do in Solaris.
4207 However, it seems that they do not in Linux for data that comes from the
4208 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4209 I found this out when I moved to a new Linux workstation and tried to run
4210 the Exim test suite.
4212 To get round this problem I have changed the code so that it now does its
4213 own call to the DNS to look up PTR records when searching for a host name.
4214 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4215 addresses that are only in /etc/hosts are still found.
4217 This behaviour is, however, controlled by an option called host_lookup_
4218 order, which defaults to "bydns:byaddr". If people want to use the other
4219 order, or indeed, just use one or the other means of lookup, they can
4220 specify it in this variable.
4222 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4223 some operating systems, this comes back from gethostbyaddr() as an empty
4224 string, and this is what Exim used to test for. However, it seems that in
4225 other systems, "." is yielded. Exim now tests for this case too.
4227 7. The values of check_spool_space and check_log_space are now held internally
4228 as a number of kilobytes instead of an absolute number of bytes. If a
4229 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4230 kilobyte. This means that much larger values can be stored.
4232 8. Exim monitor: an attempt to get the action menu when not actually pointing
4233 at a message produces an empty menu entitled "No message selected". This
4234 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4235 no entries in it ("Shell widget menu has zero width and/or height"). So I
4236 have added a single, blank menu entry in this case.
4238 9. Added ${quote_local_part.
4240 10. MIME decoding is now applied to the contents of Subject: header lines when
4243 11. Now that a reference to $sender_host_address automatically causes a reverse
4244 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4245 host lookup before query-style lookups in lists that might use this
4246 variable. This has therefore been abolished, and the "net-" prefix is no
4247 longer necessary for query-style lookups.
4249 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4250 have been a typo for LFLAGS, so it has been changed.
4252 13. The install script calls Exim with "-C /dev/null" in order to find the
4253 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4254 to be output. Howeve, since Exim outputs its version number before the
4255 error, it didn't break the script. It just looked ugly. I fixed this by
4256 always allowing "-C /dev/null" if the caller is root.
4258 14. Ignore overlarge ACL variable number when reading spool file - insurance
4259 against a later release with more variables having written the file.
4261 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4262 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4263 this, as well as the form without the "IPv6" on the front.
4265 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4266 OS/Makefile-Darwin file.
4268 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4269 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4271 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4272 variable. However, some OS use ints for inodes. I've added cast to long int
4273 to get rid of the compiler warning.
4275 19. I had forgotten to lock out "/../" in configuration file names when
4276 ALT_CONFIG_PREFIX was set.
4278 20. Routers used for verification do not need to specify transports. However,
4279 if such a router generated a host list, and callout was configured, Exim
4280 crashed, because it could not find a port number from the (non-existent)
4281 transport. It now assumes port 25 in this circumstance.
4283 21. Added the -t option to exigrep.
4285 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4286 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4287 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4288 them is set, LOOKUP_LSEARCH is forced.
4290 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4292 24. Added sender and host information to the "rejected by local_scan()" log
4293 line; previously there was no indication of these.
4295 25. Added .include_if_exists.
4297 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4298 Linux. It turns out that not all file systems support this. Apparently some
4299 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4300 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4301 not supported on the file descriptor, is now ignored when Exim is trying to
4302 sync a directory. This applies only to Linux.
4304 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4306 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4307 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4310 29. Added support for maildirsize files from supplied patch (modified a bit).
4312 30. The use of :fail: followed by an empty string could lead Exim to respond to
4313 sender verification failures with (e.g.):
4315 550 Verification failed for <xxx>
4316 550 Sender verify failed
4318 where the first response line was missing the '-' that indicates it is not
4319 the final line of the response.
4321 31. The loop for finding the name of the user that called Exim had a hardwired
4322 limit of 10; it now uses the value of finduser_retries, which is used for
4323 all other user lookups.
4325 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4328 33. Exim was neglecting to zero errno before one call of strtol() when
4329 expanding a string and expecting an integer value. On some systems this
4330 resulted in spurious "integer overflow" errors. Also, it was casting the
4331 result into an int without checking.
4333 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4334 did not work. The code looks as if it has *never* worked, though it appears
4335 to have been documented since at least releast 1.62. I have made it work.
4337 35. The "timeout_DNS" error in retry rules, also documented since at least
4338 1.62, also never worked. As it isn't clear exactly what this means, and
4339 clearly it isn't a major issue, I have abolished the feature by treating it
4340 as "timeout", and writing a warning to the main and panic logs.
4342 36. The display of retry rules for -brt wasn't always showing the error code
4345 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4346 timeout_connect_A, timeout_connect_MX.
4348 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4349 to the empty sender.
4351 39. The daemon was not analysing the content of -oX till after it had closed
4352 stderr and disconnected from the controlling terminal. This meant that any
4353 syntax errors were only noted on the panic log, and the return code from
4354 the command was 0. By re-arranging the code a little, I've made the
4355 decoding happen first, so such errors now appear on stderr, and the return
4356 code is 1. However, the actual setting up of the sockets still happens in
4357 the disconnected process, so errors there are still only recorded on the
4360 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4361 connections (as happens on some IP stacks) was logged at start up time as
4362 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4363 it from "IPv6 and IPv4", which means that two separate sockets are being
4366 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4367 says whether AF_INET or AF_INET6 was passed as an argument.
4369 42. Exiwhat output was messed up when time zones were included in log
4372 43. Exiwhat now gives more information about the daemon's listening ports,
4373 and whether -tls-on-connect was used.
4375 44. The "port" option of the smtp transport is now expanded.
4377 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4378 silently ignored. Now an error message is written to the main and panic
4381 46. There's a new ACL modifier called "logwrite" which writes to a log file
4382 as soon as it is encountered.
4384 47. Added $local_user_uid and $local_user_gid at routing time.
4386 48. Exim crashed when trying to verify a sender address that was being
4389 49. Exim was recognizing only a space character after ".include". It now also
4390 recognizes a tab character.
4392 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4393 extracting the relevant information from the specification. The man page no
4394 longer contains scrambled data for the -d option, and I've added a section
4395 at the front about calling Exim under different names.
4397 51. Added "extra_headers" argument to the "mail" command in filter files.
4399 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4402 53. Installed eximstats 1.29.
4404 54. Added transport_filter_timeout as a generic transport option.
4406 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4407 Cc: header lines. This was required by RFC 822, but it not required by RFC
4410 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4411 incoming messages that did not have them. Now it does so only if the
4412 message originates locally, that is, if there is no associated remote host
4413 address. When Resent- header lines are present, this applies to the Resent-
4414 lines rather than the non-Resent- lines.
4416 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4417 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4419 58. Messages for configuration errors now include the name of the main
4420 configuration file - useful now that there may be more than one file in a
4421 list (.included file names were always shown).
4423 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4424 for those rare installations that do not start the daemon as root or run it
4425 setuid root. I've cut out the call to initgroups() if the daemon is not
4428 60. The Exim user and group can now be bound into the binary as text strings
4429 that are looked up at the start of Exim's processing.
4431 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4433 62. Added $mailstore_basename variable.
4435 63. Installed patch to sieve.c from Michael Haardt.
4437 64. When Exim failed to open the panic log after failing to open the main log,
4438 the original message it was trying to log was written to stderr and debug
4439 output, but if they were not available (the usual case in production), it
4440 was lost. Now it is written to syslog before the two lines that record the
4441 failures to open the logs.
4443 65. Users' Exim filters run in subprocesses under the user's uid. It is
4444 possible for a "deliver" command or an alias in a "personal" command to
4445 provoke an address rewrite. If logging of address rewriting is configured,
4446 this fails because the process is not running as root or exim. There may be
4447 a better way of dealing with this, but for the moment (because 4.30 needs
4448 to be released), I have disabled address rewrite logging when running a
4449 filter in a non-root, non-exim process.
4455 1. The buildconfig auxiliary program wasn't quoting the value set for
4456 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4457 not defined. This bug was masked in 4.22 by the effect that was fixed in
4460 2. Some messages that were rejected after a message id was allocated were
4461 shown as "incomplete" by exigrep. It no longer does this for messages that
4462 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4464 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4465 have allow_domain_literals set, the ID did not get logged in the <= line.
4466 Domain literals are now always recognized in Message-ID: header lines.
4468 4. The first argument for a ${extract expansion item is the key name or field
4469 number. Leading and trailing spaces in this item were not being ignored,
4470 causing some misleading effects.
4472 5. When deliver_drop_privilege was set, single queue runner processes started
4473 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4474 same command in the process it spins off) were not dropping privilege.
4476 6. When the daemon running as "exim" started a queue runner, it always
4477 re-executed Exim in the spun-off process. This is a waste of effort when
4478 deliver_drop_privilege is set. The new process now just calls the
4479 queue-runner function directly.
4485 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4488 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4489 ignored. Though the use of -oP was forcing the writing of a pid file, it
4490 was always written to the default place.
4492 3. If the message "no IP address found for host xxxx" is generated during
4493 incoming verification, it is now followed by identification of the incoming
4494 connection (so you can more easily find what provoked it).
4496 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4498 5. Added some features to "harden" Exim a bit more against certain attacks:
4500 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4501 be put in Local/Makefile. This is like the never_users runtime option,
4502 but it cannot be overridden. The default setting is "root".
4504 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4505 prefix string with which any file named in a -C command line option
4508 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4509 is retained for -C and -D only if the caller of Exim is root. Without
4510 it, the exim user may also use -C and -D and retain privilege.
4512 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4513 command line option is disabled.
4515 6. Macro names set by the -D option must start with an upper case letter, just
4516 like macro names defined in the configuration file.
4518 7. Added "dereference=" facility to LDAP.
4520 8. Two instances of the typo "uknown" in the source files are fixed.
4522 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4523 the Configure-Makefile script screwed up while processing it.
4525 10. Incorporated PCRE 4.4.
4527 11. The SMTP synchronization check was not operating right at the start of an
4528 SMTP session. For example, it could not catch a HELO sent before the client
4529 waited for the greeting. There is now a check for outstanding input at the
4530 point when the greeting is written. Because of the duplex, asynchronous
4531 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4532 way, but not yet received, when the check is performed.
4534 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4535 on TCP/IP sockets, because this apparently causes some broken clients to
4538 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4539 unchanged) from the Cygwin maintainer.
4541 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4542 was supported instead of testing for mbx. Effectively a typo.
4544 15. The spa authenticator server code was not checking that the input it
4545 received was valid base64.
4547 16. The debug output line for the "set" modifier in ACLs was not showing the
4548 name of the variable that was being set.
4550 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4552 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4553 lookup on its own. Something else was needed to trigger the lookup. For
4554 example, a match in host_lookup or the need for a host name in a host list.
4555 Now, if $sender_host_name is referenced and the host name has not yet been
4556 looked up, a lookup is performed. If the lookup fails, the variable remains
4557 empty, and $host_lookup_failed is set to "1".
4559 19. Added "eqi" as a case-independent comparison operator.
4561 20. The saslauthd authentication condition could segfault if neither service
4562 nor realm was specified.
4564 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4565 error message that was logged was misleading, and incoming SMTP
4566 connections were dropped. The message is now more accurate, and temporary
4567 errors are given to SMTP connections.
4569 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4570 (except RSET) with a 421 error, until QUIT is received. However, it was
4571 failing to send a response to QUIT.
4573 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4574 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4575 ACL is now run after the helo_try_verify_hosts code.
4577 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4578 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4579 case-independent, so other case variants are also recognized). Apparently
4580 some systems use these upper case variants.
4582 25. If more than two messages were waiting for the same host, and a transport
4583 filter was specified for the transport, Exim sent two messages over the
4584 same TCP/IP connection, and then failed with "socket operation on non-
4585 socket" when it tried to send the third.
4587 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4589 27. The extern definition of crypt16() in expand.c was not being excluded when
4590 the OS had its own crypt16() function.
4592 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4593 as a preferred synonym for return_size_limit, both as an option and as an
4596 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4598 30. Changed the default configuration ACL to relax the local part checking rule
4599 for addresses that are not in any local domains. For these addresses,
4600 slashes and pipe symbols are allowed within local parts, but the sequence
4601 /../ is explicitly forbidden.
4603 31. SPA server authentication was not clearing the challenge buffer before
4606 32. log_message in a "warn" ACL statement was writing to the reject log as
4607 well as to the main log, which contradicts the documentation and doesn't
4608 seem right (because no rejection is happening). So I have stopped it.
4610 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4611 However, I am unable to do any testing of this.
4613 34. Fixed an infelicity in the appendfile transport. When checking directories
4614 for a mailbox, to see if any needed to be created, it was accidentally
4615 using path names with one or more superfluous leading slashes; tracing
4616 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4618 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4619 contents are added to the log line that is written for every discarded
4620 recipient. (Previously a log_message setting was ignored.)
4622 36. The ${quote: operator now quotes the string if it is empty.
4624 37. The install script runs exim in order to find its version number. If for
4625 some reason other than non-existence or emptiness, which it checks, it
4626 could not run './exim', it was installing it with an empty version number,
4627 i.e. as "exim-". This error state is now caught, and the installation is
4630 38. An argument was missing from the function that creates an error message
4631 when Exim fails to connect to the socket for saslauthd authentication.
4632 This could cause Exim to crash, or give a corrupted message.
4634 39. Added isip, isip4, and isip6 to ${if conditions.
4636 40. The ACL variables $acl_xx are now saved with the message, and can be
4637 accessed later in routers, transports, and filters.
4639 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4640 strings in the file are not string-expanded.
4642 42. If a MAIL command specified a SIZE value that was too large to fit into an
4643 int variable, the check against message_size_limit failed. Such values are
4644 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4645 day this will have to be increased, but I don't think I want to be around
4646 when emails are that large.
4653 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4654 iconv() is not standard in FreeBSD.
4656 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4657 IPv6 enabled. The observed symptom was a segmentation fault on return from
4658 the function os_common_find_running_interfaces() in src/os.c.
4660 3. In the check_special_case() function in daemon.c I had used "errno" as an
4661 argument name, which causes warnings on some systems. This was basically a
4662 typo, since it was named "eno" in the comments!
4664 4. The code that waits for the clock to tick (at a resolution of some fraction
4665 of a second) so as to ensure message-id uniqueness was always waiting for
4666 at least one whole tick, when it could have waited for less. [This is
4667 almost certainly not relevant at current processor speeds, where it is
4668 unlikely to ever wait at all. But we try to future-proof.]
4670 5. The function that sleeps for a time interval that includes fractions of a
4671 second contained a race. It did not block SIGALRM between setting the
4672 timer, and suspending (a couple of lines later). If the interval was short
4673 and the sigsuspend() was delayed until after it had expired, the suspension
4674 never ended. On busy systems this could lead to processes getting stuck for
4677 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4678 process, before it forks any delivery processes. The open lookup caching
4679 mechanism meant that the open file or database connection was passed into
4680 the delivery process. The problem was that delivery processes always tidy
4681 up cached lookup data. This could cause a problem for the next delivery
4682 process started by the queue runner, because the external queue runner
4683 process does not know about the closure. So the next delivery process
4684 still has data in the lookup cache. In the case of a file lookup, there was
4685 no problem because closing a file descriptor in a subprocess doesn't affect
4686 the parent. However, if the lookup was caching a connection to a database,
4687 the connection was closed, and the second delivery process was likely to
4688 see errors such as "PGSQL: query failed: server closed the connection
4689 unexpectedly". The problem has been fixed by closing all cached lookups
4690 in a queue runner before running a delivery process.
4692 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4693 seem to have the "const" qualifier which it has on other OS. I've
4696 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4697 *of the same type* (client or server) with the same public name that an
4698 error should be diagnosed.
4700 9. When Exim looked up a host name for an IP address, but failed to find the
4701 original IP address when looking up the host name (a safety check), it
4702 output the message "<ip address> does not match any IP for NULL", which was
4703 confusing, to say the least. The bug was that the host name should have
4704 appeared instead of "NULL".
4706 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4707 user that is built into the binary, and the -C or -D options is used, root
4708 privilege is dropped before the configuration file is read. In addition,
4709 logging is switched to stderr instead of the normal log files. If the
4710 configuration then re-defines the Exim user, the unprivileged environment
4711 is probably not what is expected, so Exim logs a panic warning message (but
4714 However, if deliver_drop_privilege is set, the unprivileged state may well
4715 be exactly what is intended, so the warning has been cut out in that case,
4716 and Exim is allowed to try to write to its normal log files.
4722 1. smtp_return_error_details was not giving details for temporary sender
4723 or receiver verification errors.
4725 2. Diagnose a configuration error if two authenticators have the same public
4728 3. Exim used not to create the message log file for a message until the first
4729 delivery attempt. This could be confusing when incoming messages were held
4730 for policy or load reasons. The message log file is now created at the time
4731 the message is received, and an initial "Received" line is written to it.
4733 4. The automatically generated man page for command line options had a minor
4734 bug that caused no ill effects; however, a more serious problem was that
4735 the procedure for building the man page automatically didn't always
4736 operate. Consequently, release 4.20 contains an out-of-date version. This
4737 shouldn't happen again.
4739 5. When building Exim with embedded Perl support, the script that builds the
4740 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4741 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4743 6. The freeze_tell option was not being used for messages that were frozen on
4744 arrival, either by an ACL or by local_scan().
4746 7. Added the smtp_incomplete_transaction log selector.
4748 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4749 was accepting AUTH without a new EHLO.
4751 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4752 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4755 10. Logging of TCP/IP connections (when configured) now happens in the main
4756 daemon process instead of the child process, so that the TCP/IP connection
4757 count is more accurate (but it can never be perfect).
4759 11. The use of "drop" in a nested ACL was not being handled correctly in the
4760 outer ACL. Now, if condition failure induced by the nested "drop" causes
4761 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4762 or "require"), the connection is dropped.
4764 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4765 that yield "discard" can now be used with an "accept" or a "discard" verb,
4766 but an error is generated for any others (because I can't see a useful way
4767 to define what should happen).
4769 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4770 are now processed in the same way as lines in the Exim configuration file.
4771 In particular, continuation lines are supported.
4773 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4775 15. Added -ti meaning -t -i.
4777 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4778 domains, and warn by logging if others are found.
4780 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4781 returned from GIFCONF when Exim is trying to find the list of interfaces on
4782 a host. The code in os.c has been modified to copy each ifreq to an aligned
4783 structure in all cases.
4785 Also, in some cases, the returned ifreq's were being copied to a 'struct
4786 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4787 means the last couple of bytes of an IPv6 address could be chopped if the
4788 ifreq contained only a normal sockaddr (14 bytes storage).
4790 18. Named domain lists were not supported in the hosts_treat_as_local option.
4791 An entry such as +xxxx was not recognized, and was treated as a literal
4794 19. Ensure that header lines added by a DATA ACL are included in the reject log
4795 if the ACL subsequently rejects the message.
4797 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4798 MD5 (which is deprecated).
4800 21. When testing a filter file using -bf, Exim was writing a message when it
4801 took the sender from a "From " line in the message, but it was not doing so
4802 when it took $return_path from a Return-Path: header line. It now does.
4804 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4805 with a valid header line field name (a series of printing characters
4806 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4808 23. Changed "disc" in the source to "disk" to conform to the documentation and
4809 the book and for uniformity.
4811 24. Ignore Sendmail's -Ooption=value command line item.
4813 25. When execve() failed while trying to run a command in a pipe transport,
4814 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4815 could be confused with a return value of 69 from the command itself. This
4816 has been changed to 127, the value the shell returns if it is asked to run
4817 a non-existent command. The wording for the related log line suggests a
4818 non-existent command as the problem.
4820 26. If received_header_text expands to an empty string, do not add a Received:
4821 header line to the message. (Well, it adds a token one on the spool, but
4822 marks it "old" so that it doesn't get used or transmitted.)
4824 27. Installed eximstats 1.28 (addition of -nt option).
4826 28. There was no check for failure on the call to getsockname() in the daemon
4827 code. This can fail if there is a shortage of resources on the system, with
4828 ENOMEM, for example. A temporary error is now given on failure.
4830 29. Contrary to the C standard, it seems that in some environments, the
4831 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4832 program. Exim now does this explicitly; it affects the formatting of
4833 timestamps using strftime().
4835 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4836 complaints. I've now initialized all the variables, to avoid this.
4838 32. Header lines added by a system filter were not being "seen" during
4839 transport-time rewrites.
4841 33. The info_callback() function passed to OpenSSL is set up with type void
4842 (*)(SSL *, int, int), as described somewhere. However, when calling the
4843 function (actually a macro) that sets it up, the type void(*)() is
4844 expected. I've put in a cast to prevent warnings from picky compilers.
4846 34. If a DNS black list lookup found a CNAME record, but there were no A
4847 records associated with the domain it pointed at, Exim crashed.
4849 35. If a DNS black list lookup returned more than one A record, Exim ignored
4850 all but the first. It now scans all returned addresses if a particular IP
4851 value is being sought. In this situation, the contents of the
4852 $dnslist_value variable are a list of all the addresses, separated by a
4855 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4856 to accept a host name and all its aliases if the forward lookup for any of
4857 them yielded the IP address of the incoming connection. Now it accepts only
4858 those names whose forward lookup yields the correct IP address. Any other
4859 names are discarded. This closes a loophole whereby a rogue DNS
4860 administrator could create reverse DNS records to break through a
4861 wildcarded host restriction in an ACL.
4863 37. If a user filter or a system filter that ran in a subprocess used any of
4864 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4865 the wrong values were passed to the pipe command ($thisaddress had the
4866 value of $0, $0 had the value of $1, etc). This bug was introduced by
4867 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4869 38. Improved the line breaking for long SMTP error messages from ACLs.
4870 Previously, if there was no break point between 40 and 75 characters, Exim
4871 left the rest of the message alone. Two changes have been made: (a) I've
4872 reduced the minimum length to 35 characters; (b) if it can't find a break
4873 point between 35 and 75 characters, it looks ahead and uses the first one
4874 that it finds. This may give the occasional overlong line, but at least the
4875 remaining text gets split now.
4877 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4878 file descriptors might be low, and that setting 1000 would always raise it.
4879 It turns out that in some environments, the limit is already over 1000 and
4880 that lowering it causes trouble. So now Exim takes care not to decrease it.
4882 40. When delivering a message, the value of $return_path is set to $sender_
4883 address at the start of routing (routers may change the value). By an
4884 oversight, this default was not being set up when an address was tested by
4885 -bt or -bv, which affected the outcome if any router or filter referred to
4888 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4889 log only when "message" or "log_message" are set. However, if one of the
4890 conditions was an address verification, or a call to a nested ACL, the
4891 messages generated by the underlying test were being passed through. This
4892 no longer happens. The underlying message is available in $acl_verify_
4893 message for both "message" and "log_message" expansions, so it can be
4894 passed through if needed.
4896 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4897 new expansion $bh_ to give the encoded byte string without charset
4898 translation. Translation happens only if iconv() is available; HAVE_ICONV
4899 indicates this at build time. HEADERS_CHARSET gives the charset to
4900 translate to; headers_charset can change it in the configuration, and
4901 "headers charset" can change it in an individual filter file.
4903 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4904 that creates RFC 2047 encoded "words" labels them as that charset instead
4905 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4906 expansion operator; (ii) when Exim creates a From: line for a local
4907 message; (iii) when a header line is rewritten to include a "phrase" part.
4909 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4910 buggy, causing it to skip the first lines of messages whose message ID
4911 ended in 'D'. This would not have bitten before Exim release 4.14, because
4912 message IDs were unlikely to end in 'D' before then. The effect was to have
4913 incorrect size information for certain domains.
4915 45. #include "config.h" was missing at the start of the crypt16.c module. This
4916 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4919 46. If there was a timeout during a "random" callout check, Exim treated it as
4920 a failure of the random address, and carried on sending RSET and the real
4921 address. If the delay was just some slowness somewhere, the response to the
4922 original RCPT would be taken as a response to RSET and so on, causing
4923 mayhem of various kinds.
4925 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4926 when I implemented it. It didn't allow for the fact that some option values
4927 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4928 the right test for positive values.
4930 48. Domain names in DNS records are case-independent. Exim always looks them up
4931 in lower case. Some resolvers return domain names in exactly the case they
4932 appear in the zone file, that is, they may contain uppercase letters. Not
4933 all resolvers do this - some return always lower case. Exim was treating a
4934 change of case by a resolver as a change of domain, similar to a widening
4935 of a domain abbreviation. This triggered its re-routing code and so it was
4936 trying to route what was effectively the same domain again. This normally
4937 caused routing to fail (because the router wouldn't handle the domain
4938 twice). Now Exim checks for this case specially, and just changes the
4939 casing of the domain that it ultimately uses when it transmits the message
4942 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4945 50. If a filter generated a file delivery with a non-absolute name (possible if
4946 no home directory exists for the router), the forbid_file option was not
4949 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4950 the existing equality matching.
4952 52. Exim was using ints instead of ino_t variables in some places where it was
4953 dealing with inode numbers.
4955 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4956 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4957 called TMPDIR, and if it finds it is different, it changes its value.
4959 54. The smtp_printf() function is now made available to local_scan() so
4960 additional output lines can be written before returning. There is also an
4961 smtp_fflush() function to enable the detection of a dropped connection.
4962 The variables smtp_input and smtp_batched_input are exported to
4965 55. Changed the default runtime configuration: the message "Unknown user"
4966 has been removed from the ACL, and instead placed on the localuser router,
4967 using the cannot_route_message feature. This means that any verification
4968 failures that generate their own messages won't get overridden. Similarly,
4969 the "Unrouteable address" message that was in the ACL for unverifiable
4970 relay addresses has also been removed.
4972 56. Added hosts_avoid_esmtp to the smtp transport.
4974 57. The exicyclog script was not checking for the esoteric option
4975 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4976 will work only if exicyclog is run under the appropriate euid.
4978 58. Following a discussion on the list, the rules by which Exim recognises line
4979 endings on incoming messages have been changed. The -dropcr and drop_cr
4980 options are now no-ops, retained only for backwards compatibility. The
4981 following line terminators are recognized: LF CRLF CR. However, special
4982 processing applies to CR:
4984 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4985 nor a local message in the state where . is a terminator.
4987 (ii) If a bare CR is encountered in a header line, an extra space is added
4988 after the line terminator so as not to end the header. The reasoning
4989 behind this is that bare CRs in header lines are most likely either
4990 to be mistakes, or people trying to play silly games.
4992 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4993 was being incorrectly given as 18 bytes larger than it should have been.
4994 This is a VOB (very old bug).
4996 60. This may never have affected anything current, but just in case it has:
4997 When the local host is found other than at the start of a list of hosts,
4998 the local host, those with the same MX, and any that follow, are discarded.
4999 When the list in question was part of a longer list of hosts, the following
5000 hosts (not currently being processed) were also being discarded. This no
5001 longer happens. I'm not sure if this situation could ever has previously
5004 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5007 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5008 header. This is something that is recommended in a new Internet Draft, and
5009 is something that is documented as being done by Sendmail. There are two
5010 possible values. For messages generated by the autoreply transport, Exim
5013 Auto-Submitted: auto-replied
5015 whereas for all other generated messages (e.g. bounces) it adds
5017 Auto-Submitted: auto-generated
5019 63. The "personal" condition in filters now includes a test for the
5020 Auto-Submitted: header. If it contains the string "auto-" the message it
5021 not considered personal.
5023 64. Added rcpt_include_affixes as a generic transport option.
5025 65. Added queue_only_override (default true).
5027 66. Added the syslog_duplication option.
5029 67. If what should have been the first header line of a message consisted of
5030 a space followed by a colon, Exim was mis-interpreting it as a header line.
5031 It isn't of course - it is syntactically invalid and should therefore be
5032 treated as the start of the message body. The misbehaviour could have
5033 caused a number of strange effects, including loss of data in subsequent
5034 header lines, and spool format errors.
5036 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5037 client host had authenticated. This control can now be exercised by an ACL
5038 for more flexibility.
5040 69. By default, callouts do not happen when testing with -bh. There is now a
5041 variant, -bhc, which does actually run the callout code, including
5042 consulting and updating the callout cache.
5044 70. Added support for saslauthd authentication, courtesy of Alexander
5047 71. If statvfs() failed on the spool or log directories while checking their
5048 size for availability, Exim confusingly gave the error "space shortage".
5049 Furthermore, in debugging mode it crashed with a floating point exception.
5050 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5051 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5052 serious problem, Exim now writes to the main and panic logs when this
5053 happens, with details of the failure. It then refuses to accept the
5054 incoming message, giving the message "spool directory problem" or "log
5055 directory problem" with a 421 code for SMTP messages.
5057 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5058 0, 1, and 2 exist, because some OS complain for execs without them (see
5059 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5060 descriptors. However, the code omitted to check that the open succeeded,
5061 causing mysterious errors if for some reason the permissions on /dev/null
5062 got screwed. Now Exim writes a message to the main and panic logs, and
5063 bombs out if it can't open /dev/null.
5065 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5066 interact so that it is all more flexible. It is supposed to remain
5067 backwards compatible. Also added extra_local_interfaces.
5069 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5070 to bomb out with an assertion failure - to the client this appears as a
5071 connection drop. This problem occurs in the part of the code that was taken
5072 from the Samba project. Fortunately, the assertion is in a very simple
5073 function, so I have fixed this by reproducing the function inline in the
5074 one place where it is called, and arranging for authentication to fail
5075 instead of killing the process with assert().
5077 75. The SPA client code was not working when the server requested OEM rather
5078 than Unicode encoding.
5080 76. Added code to make require_files with a specific uid setting more usable in
5081 the case where statting the file as root fails - usually a non-root-mounted
5082 NFS file system. When this happens and the failure is EACCES, Exim now
5083 forks a subprocess and does the per-uid checking as the relevant uid.
5085 77. Added process_log_path.
5087 78. If log_file_path was not explicitly set, a setting of check_log_space or
5088 check_log_inodes was ignored.
5090 79. If a space check for the spool or log partitions fails, the incident is now
5091 logged. Of course, in the latter case the data may get lost...
5093 80. Added the %p formatting code to string_format() so that it can be used to
5094 print addresses in debug_print(). Adjusted all the address printing in the
5095 debugging in store.c to use %p rather than %d.
5097 81. There was a concern that a line of code in smtp_in.c could overflow a
5098 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5099 initially expressed, the concern was not well-founded, because trailing
5100 spaces are removed early. However, if the trailing spaces were followed by
5101 a NULL, they did not get removed, so the overflow was possible. Two fixes
5104 (a) I re-wrote the offending code in a cleaner fashion.
5105 (b) If an incoming SMTP command contains a NULL character, it is rejected
5108 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5109 runs initgroups(), so that if the Exim user is in any additional groups,
5110 they will be used during message reception.
5116 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob