1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.620 2010/06/05 10:34:29 pdp Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
8 PP/01 Date: & Message-Id: revert to normally being appended to a message,
9 only prepend for the Resent-* case. Fixes regression introduced in
10 Exim 4.70 by NM/22 for Bugzilla 607.
12 PP/02 Include check_rfc2047_length in configure.default because we're seeing
13 increasing numbers of administrators be bitten by this.
15 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
17 PP/03 Bugzilla 994: added openssl_options main configuration option.
19 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
21 PP/05 Bugzilla 834: provide a permit_codedump option for pipe transports.
23 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
25 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
26 without a peer certificate (I believe), leading to a segfault because of
27 an assumption that peers always have certificates. Be a little more
28 paranoid. Problem reported by Martin Tscholak.
34 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
35 $data_path, and $header_path variables; fixed documentation bugs and
38 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
39 exipick to access non-standard spools, including the "frozen" queue
42 NM/01 Bugzilla 965: Support mysql stored procedures.
43 Patch from Alain Williams
45 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
47 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
48 Patch from Andreas Metzler
50 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
51 Patch from Kirill Miazine
53 NM/05 Bugzilla 671: Added umask to procmail example.
55 JJ/03 installed exipick 20100323.0, fixing doc bug
57 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
58 directory. Notification and patch from Dan Rosenberg.
60 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
62 TK/02 Improve log output when DKIM signing operation fails.
64 MH/01 Treat the transport option dkim_domain as a colon separated
65 list, not as a single string, and sign the message with each element,
66 omitting multiple occurences of the same signer.
68 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
69 Bugzilla 985, 986. Patch by Simon Arlott
71 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
74 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
75 MBX locking. Notification from Dan Rosenberg.
81 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
83 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
85 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
87 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
89 NM/04 Bugzilla 915: Flag broken perl installation during build.
95 TK/01 Added patch by Johannes Berg that expands the main option
96 "spamd_address" if it starts with a dollar sign.
98 TK/02 Write list of recipients to X-Envelope-Sender header when building
99 the mbox-format spool file for content scanning (suggested by Jakob
102 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
103 (http://www.dcc-servers.net/) support via dccifd. Activated by
104 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
106 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
107 by Mark Daniel Reidel <mr@df.eu>.
109 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
110 When building exim an external PCRE library is now needed -
111 PCRE is a system library on the majority of modern systems.
112 See entry on PCRE_LIBS in EDITME file.
114 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
115 conversation. Added nologin parameter to request.
116 Patch contributed by Kirill Miazine
118 TF/01 Do not log submission mode rewrites if they do not change the address.
120 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
122 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
123 log files in place. Contributed by Roberto Lima
125 NM/04 Bugzilla 667: close socket used by dovecot authenticator
127 TF/03 Bugzilla 615: When checking the local_parts router precondition
128 after a local_part_suffix or local_part_prefix option, Exim now
129 does not use the address's named list lookup cache, since this
130 contains cached lookups for the whole local part.
132 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
133 Robert Millan. Documentation is in experimental-spec.txt
135 TF/04 Bugzilla 668: Fix parallel build (make -j).
137 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
139 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
140 Patch provided by Jan Srzednicki
142 TF/05 Leading white space used to be stripped from $spam_report which
143 wrecked the formatting. Now it is preserved.
145 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
146 that they are available at delivery time.
148 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
150 TF/08 TLS error reporting now respects the incoming_interface and
151 incoming_port log selectors.
153 TF/09 Produce a more useful error message if an SMTP transport's hosts
154 setting expands to an empty string.
156 NM/06 Bugzilla 744: EXPN did not work under TLS.
157 Patch provided by Phil Pennock
159 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
160 Patch provided by Richard Godbee
162 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
163 acl_smtp_notquit, added index entry.
165 NM/09 Bugzilla 787: Potential buffer overflow in string_format
166 Patch provided by Eugene Bujak
168 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
169 Patch provided by Maxim Dounin
171 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
172 Patch provided by Phil Pennock
174 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
176 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
177 Patch provided by Brad "anomie" Jorsch
179 NM/14 Bugzilla 622: Added timeout setting to transport filter.
180 Patch provided by Dean Brooks
182 TK/05 Add native DKIM support (does not depend on external libraries).
184 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
185 Patch provided by Graeme Fowler
187 NM/16 Bugzilla 851: Documentation example syntax fix.
189 NM/17 Changed NOTICE file to remove references to embedded PCRE.
191 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
193 NM/19 Bugzilla 745: TLS version reporting
194 Patch provided by Phil Pennock
196 NM/20 Bugzilla 167: bool: condition support
197 Patch provided by Phil Pennock
199 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
200 Patch provided by Phil Pennock
202 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
203 Patch provided by Brad "anomie" Jorsch
205 NM/23 Bugzilla 687: Fix misparses in eximstats
206 Patch provided by Heiko Schlittermann
208 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
209 Patch provided by Heiko Schlittermann
211 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
212 plus update to original patch.
214 NM/26 Bugzilla 799: Documentation correction for ratelimit
216 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
217 Patch provided by David Brownlee
219 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
221 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
223 NM/30 Bugzilla 888: TLS documentation bugfixes
225 NM/31 Bugzilla 896: Dovecot buffer overrun fix
227 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
228 Unlike the original bugzilla I have changed all shell scripts in src tree
230 NM/33 Bugzilla 898: Transport filter timeout fix
231 Patch by Todd Rinaldo
233 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
234 Patch by Serge Demonchaux
236 NM/35 Bugzilla 39: Base64 decode bug fixes
237 Patch by Jakob Hirsch
239 NM/36 Bugzilla 909: Correct connect() call in dcc code
241 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
243 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
245 NM/39 Bugzilla 911: Fixed MakeLinks build script
251 TK/01 Add preliminary DKIM support. Currently requires a forked version of
252 ALT-N's libdkim that I have put here:
253 http://duncanthrax.net/exim-experimental/
255 Note to Michael Haardt: I had to rename some vars in sieve.c. They
256 were called 'true' and it seems that C99 defines that as a reserved
257 keyword to be used with 'bool' variable types. That means you could
258 not include C99-style headers which use bools without triggering
259 build errors in sieve.c.
261 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
262 as mailq or other aliases. Changed the --help handling significantly
263 to do whats expected. exim_usage() emits usage/help information.
265 SC/01 Added the -bylocaldomain option to eximstats.
267 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
269 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
271 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
277 PH/01 Another patch from the Sieve maintainer.
279 PH/02 When an IPv6 address is converted to a string for single-key lookup
280 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
281 dots are used instead of colons so that keys in lsearch files need not
282 contain colons. This was done some time before quoting was made available
283 in lsearch files. However, iplsearch files do require colons in IPv6 keys
284 (notated using the quote facility) so as to distinguish them from IPv4
285 keys. This meant that lookups for IP addresses in host lists did not work
286 for iplsearch lookups.
288 This has been fixed by arranging for IPv6 addresses to be expressed with
289 colons if the lookup type is iplsearch. This is not incompatible, because
290 previously such lookups could never work.
292 The situation is now rather anomolous, since one *can* have colons in
293 ordinary lsearch keys. However, making the change in all cases is
294 incompatible and would probably break a number of configurations.
296 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
299 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
300 conversion specification without a maximum field width, thereby enabling
301 a rogue spamd server to cause a buffer overflow. While nobody in their
302 right mind would setup Exim to query an untrusted spamd server, an
303 attacker that gains access to a server running spamd could potentially
304 exploit this vulnerability to run arbitrary code as the Exim user.
306 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
307 $primary_hostname instead of what libspf2 thinks the hosts name is.
309 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
310 a directory entry by the name of the lookup key. Previously, if a
311 symlink pointed to a non-existing file or a file in a directory that
312 Exim lacked permissions to read, a lookup for a key matching that
313 symlink would fail. Now it is enough that a matching directory entry
314 exists, symlink or not. (Bugzilla 503.)
316 PH/03 The body_linecount and body_zerocount variables are now exported in the
319 PH/04 Added the $dnslist_matched variable.
321 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
322 This means they are set thereafter only if the connection becomes
325 PH/06 Added the client_condition to authenticators so that some can be skipped
326 by clients under certain conditions.
328 PH/07 The error message for a badly-placed control=no_multiline_responses left
329 "_responses" off the end of the name.
331 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
333 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
334 (without spaces) instead of just copying the configuration text.
336 PH/10 Added the /noupdate option to the ratelimit ACL condition.
338 PH/11 Added $max_received_linelength.
340 PH/12 Added +ignore_defer and +include_defer to host lists.
342 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
343 way in which PCRE > 7.0 is built.
345 PH/14 Implemented queue_only_load_latch.
347 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
348 MAIL command. The effect was to mangle the value on 64-bit systems.
350 PH/16 Another patch from the Sieve maintainer.
352 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
354 PH/18 If a system quota error occurred while trying to create the file for
355 a maildir delivery, the message "Mailbox is full" was not appended to the
356 bounce if the delivery eventually timed out. Change 4.67/27 below applied
357 only to a quota excession during the actual writing of the file.
359 PH/19 It seems that peer DN values may contain newlines (and other non-printing
360 characters?) which causes problems in log lines. The DN values are now
361 passed through string_printing() before being added to log lines.
363 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
364 and InterBase are left for another time.)
366 PH/21 Added message_body_newlines option.
368 PH/22 Guard against possible overflow in moan_check_errorcopy().
370 PH/23 POSIX allows open() to be a macro; guard against that.
372 PH/24 If the recipient of an error message contained an @ in the local part
373 (suitably quoted, of course), incorrect values were put in $domain and
374 $local_part during the evaluation of errors_copy.
380 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
381 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
384 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
385 issue a MAIL command.
387 PH/02 In an ACL statement such as
389 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
391 if a client was not listed at all, or was listed with a value other than
392 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
393 the condition was not true (as it should be), so access was not denied.
394 The bug was that the ! inversion was incorrectly passed on to the second
395 item. This has been fixed.
397 PH/03 Added additional dnslists conditions == and =& which are different from
398 = and & when the dns lookup returns more than one IP address.
400 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
401 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
403 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
404 FSYNC, which compiles an option called disable_fsync that allows for
405 bypassing fsync(). The documentation is heavily laced with warnings.
407 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
409 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
410 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
411 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
412 including adding "make clean"; (3) Added -fPIC when compiling the test
413 dynamically loaded module, to get rid of a warning.
415 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
416 message fails, move_frozen_messages = true and ignore_bounce_errors_after
417 = 0s. The bug is otherwise harmless.
419 PH/07 There was a bug in the dovecot authenticator such that the value of
420 $auth1 could be overwritten, and so not correctly preserved, after a
421 successful authentication. This usually meant that the value preserved by
422 the server_setid option was incorrect.
424 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
426 PH/09 Installed PCRE release 7.0.
428 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
429 run for batched SMTP input. It is now run at the start of every message
430 in the batch. While fixing this I discovered that the process information
431 (output by running exiwhat) was not always getting set for -bs and -bS
432 input. This is fixed, and it now also says "batched" for BSMTP.
434 PH/11 Added control=no_pipelining.
436 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
437 patch, slightly modified), and move the expansion of helo_data till after
438 the connection is made in the smtp transport (so it can use these
441 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
443 PH/14 Added log_selector = +pid.
445 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
447 PH/16 Add ${if forany and ${if forall.
449 PH/17 Added dsn_from option to vary the From: line in DSNs.
451 PH/18 Flush SMTP output before performing a callout, unless control =
452 no_callout_flush is set.
454 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
455 was true (the default) a successful delivery failed to delete the retry
456 item, thus causing premature timeout of the address. The bug is now
459 PH/20 Added hosts_avoid_pipelining to the smtp transport.
461 PH/21 Long custom messages for fakedefer and fakereject are now split up
462 into multiline reponses in the same way that messages for "deny" and
463 other ACL rejections are.
465 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
466 with slight modification.
468 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
469 draft, changing the syntax and factoring some duplicate code".
471 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
472 for deliveries of the second and subsequent messages over the same SMTP
475 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
476 ${reduce, with only minor "tidies".
478 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
480 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
481 expansion side effects.
483 PH/27 When a message times out after an over-quota error from an Exim-imposed
484 quota, the bounce message says "mailbox is full". This message was not
485 being given when it was a system quota that was exceeded. It now should
488 MH/03 Made $recipients available in local_scan(). local_scan() already has
489 better access to the recipient list through recipients_list[], but
490 $recipients can be useful in postmaster-provided expansion strings.
492 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
493 in the case of a MAIL command with additional options following the
494 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
495 were accidentally chopped off.
497 PH/29 SMTP synchronization checks are implemented when a command is read -
498 there is a check that no more input is waiting when there shouldn't be
499 any. However, for some commands, a delay in an ACL can mean that it is
500 some time before the response is written. In this time, more input might
501 arrive, invalidly. So now there are extra checks after an ACL has run for
502 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
503 pipelining has not been advertised.
505 PH/30 MH's patch to allow iscntrl() characters to be list separators.
507 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
508 returned in the SMTP response when smtp_return_error_details was false.
511 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
512 instead of the C I/O that was originally supplied, because problems were
515 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
516 Exim which did not show up earlier: it was assuming that a call to
517 SSL_CTX_set_info_callback() might give an error value. In fact, there is
518 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
519 was a macro that became an assignment, so it seemed to work. This has
520 changed to a proper function call with a void return, hence the compile
521 error. Exim's code has been fixed.
523 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
526 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
528 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
530 PH/37 If a message is not accepted after it has had an id assigned (e.g.
531 because it turns out to be too big or there is a timeout) there is no
532 "Completed" line in the log. When some messages of this type were
533 selected by exigrep, they were listed as "not completed". Others were
534 picked up by some special patterns. I have improved the selection
535 criteria to be more general.
537 PH/38 The host_find_failed option in the manualroute router can now be set
538 to "ignore", to completely ignore a host whose IP address cannot be
539 found. If all hosts are ignored, the behaviour is controlled by the new
540 host_all_ignored option.
542 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
543 homing or because of multiple MX records with /mx) generated more than
544 one IP address, and the following item turned out to be the local host,
545 all the secondary addresses of the first item were incorrectly removed
546 from the list, along with the local host and any following hosts (which
547 is what is supposed to happen).
549 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
550 whoever called Exim into the -H file. In the case of the daemon it was
551 behaving confusingly. When first started, it used values for whoever
552 started the daemon, but after a SIGHUP it used the Exim user (because it
553 calls itself on a restart). I have changed the code so that it now always
556 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
557 message are rejected with the same error (e.g. no authentication or bad
558 sender address), and a DATA command is nevertheless sent (as can happen
559 with PIPELINING or a stupid MUA), the error message that was given to the
560 RCPT commands is included in the rejection of the DATA command. This is
561 intended to be helpful for MUAs that show only the final error to their
564 PH/42 Another patch from the Sieve maintainer.
566 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
567 Eximstats - Fixed some broken HTML links and added missing column headers
569 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
570 columns for Rejects, Temp Rejects, Ham, and Spam rows.
572 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
574 PH/43 Yet another patch from the Sieve maintainer.
576 PH/44 I found a way to check for a TCP/IP connection going away before sending
577 the response to the final '.' that terminates a message, but only in the
578 case where the client has not sent further data following the '.'
579 (unfortunately, this is allowed). However, in many cases there won't be
580 any further data because there won't be any more messages to send. A call
581 to select() can be used: if it shows that the input is "ready", there is
582 either input waiting, or the socket has been closed. An attempt to read
583 the next input character can distinguish the two cases. Previously, Exim
584 would have sent an OK response which the client would never have see.
585 This could lead to message repetition. This fix should cure that, at
586 least in a lot of common cases.
588 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
589 advertised in response to EHLO.
595 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
596 fixed by 4.65/MH/01 (is this a record?) are fixed:
598 (i) An empty string was always treated as zero by the numeric comparison
599 operators. This behaviour has been restored.
601 (ii) It is documented that the numeric comparison operators always treat
602 their arguments as decimal numbers. This was broken in that numbers
603 starting with 0 were being interpreted as octal.
605 While fixing these problems I realized that there was another issue that
606 hadn't been noticed. Values of message_size_limit (both the global option
607 and the transport option) were treated as octal if they started with 0.
608 The documentation was vague. These values are now always treated as
609 decimal, and I will make that clear in the documentation.
615 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
616 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
619 MH/01 Don't check that the operands of numeric comparison operators are
620 integers when their expansion is in "skipping" mode (fixes bug
621 introduced by 4.64-PH/07).
623 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
624 child addresses, Exim now panics and dies. Previously, because the count
625 is held in a short int, deliveries were likely to be lost. As such a
626 large number of recipients for a single message is ridiculous
627 (performance will be very, very poor), I have chosen to impose a limit
628 rather than extend the field.
634 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
635 leftover -K file (the existence of which was triggered by #402).
636 While we were at it, introduced process PID as part of the -K
637 filename. This should rule out race conditions when creating
640 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
641 processing considerably. Previous code took too long for large mails,
642 triggering a timeout which in turn triggers #401.
644 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
645 in the DK code in transports.c. sendfile() is not really portable,
646 hence the _LINUX specificness.
648 TF/01 In the add_headers option to the mail command in an Exim filter,
649 there was a bug that Exim would claim a syntax error in any
650 header after the first one which had an odd number of characters
653 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
654 callout verification, Exim cached a "reject" for the entire domain. This
655 is correct for most verifications, but it is not correct for a recipient
656 verification with use_sender or use_postmaster set, because in that case
657 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
658 case of MAIL FROM:<> rejection from other early rejections (e.g.
659 rejection of HELO). When verifying a recipient using a non-null MAIL
660 address, the cache is ignored if it shows MAIL FROM:<> rejection.
661 Whatever the result of the callout, the value of the domain cache is
662 left unchanged (for any other kind of callout, getting as far as trying
663 RCPT means that the domain itself is ok).
665 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
668 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
669 manifest itself as EPIPE rather than ECONNECT. When tidying away a
670 session, the daemon ignores ECONNECT errors and logs others; it now
671 ignores EPIPE as well.
673 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
674 (quoted-printable decoding).
676 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
677 later the small subsequent patch to fix an introduced bug.
679 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
681 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
683 PH/08 An error is now given if message_size_limit is specified negative.
685 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
686 to be given (somewhat) arbitrary names.
688 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
691 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
692 miscellaneous code fixes
694 PH/10 Added the log_reject_target ACL modifier to specify where to log
697 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
698 hostname. This is wrong, because it relates to the incoming message (and
699 probably the interface on which it is arriving) and not to the outgoing
700 callout (which could be using a different interface). This has been
701 changed to use the value of the helo_data option from the smtp transport
702 instead - this is what is used when a message is actually being sent. If
703 there is no remote transport (possible with a router that sets up host
704 addresses), $smtp_active_hostname is used.
706 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
707 tweaks were necessary in order to get it to work (see also 21 below):
708 (a) The code assumed that strncpy() returns a negative number on buffer
709 overflow, which isn't the case. Replaced with Exim's string_format()
711 (b) There were several signed/unsigned issues. I just did the minimum
712 hacking in of casts. There is scope for a larger refactoring.
713 (c) The code used strcasecmp() which is not a standard C function.
714 Replaced with Exim's strcmpic() function.
715 (d) The code set only $1; it now sets $auth1 as well.
716 (e) A simple test gave the error "authentication client didn't specify
717 service in request". It would seem that Dovecot has changed its
718 interface. Fortunately there's a specification; I followed it and
719 changed what the client sends and it appears to be working now.
721 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
724 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
725 address A is aliased to B and C, where B exists and C does not. Without
726 -v the output is "A verified" because verification stops after a
727 successful redirection if more than one address is generated. However,
728 with -v the child addresses are also verified. Exim was outputting "A
729 failed to verify" and then showing the successful verification for C,
730 with its parentage. It now outputs "B failed to verify", showing B's
731 parentage before showing the successful verification of C.
733 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
734 look up a TXT record in a specific list after matching in a combined
737 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
738 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
739 they consult the DNS. I had assumed they would set it the way they
740 wanted; and indeed my experiments on Linux seem to show that in some
741 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
742 To be on the safe side, however, I have now made the interface to
743 host_find_byname() similar to host_find_bydns(), with an argument
744 containing the DNS resolver options. The host_find_byname() function now
745 sets these options at its start, just as host_find_bydns() does. The smtp
746 transport options dns_qualify_single and dns_search_parents are passed to
747 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
748 of host_find_byname() use the default settings of RES_DEFNAMES
749 (qualify_single) but not RES_DNSRCH (search_parents).
751 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
752 spool_read_header() do less string testing, by means of a preliminary
753 switch on the second character of optional "-foo" lines. (This is
754 overdue, caused by the large number of possibilities that now exist.
755 Originally there were few.) While I was there, I also converted the
756 str(n)cmp tests so they don't re-test the leading "-" and the first
757 character, in the hope this might squeeze out yet more improvement.
759 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
760 flag allowing group syntax was set by the header_syntax check but not
761 turned off, possible causing trouble later; (2) The flag was not being
762 set at all for the header_verify test, causing "group"-style headers to
763 be rejected. I have now set it in this case, and also caused header_
764 verify to ignore an empty address taken from a group. While doing this, I
765 came across some other cases where the code for allowing group syntax
766 while scanning a header line wasn't quite right (mostly, not resetting
767 the flag correctly in the right place). These bugs could have caused
768 trouble for malformed header lines. I hope it is now all correct.
770 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
771 with the "reply" argument non-NULL. The code, however (which originally
772 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
773 but it didn't always do it. This confused somebody who was copying the
774 code for some other use. I have removed all the tests.
776 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
777 feature that was used to support insecure browsers during the U.S. crypto
778 embargo. It requires special client support, and Exim is probably the
779 only MTA that supported it -- and would never use it because real RSA is
780 always available. This code has been removed, because it had the bad
781 effect of slowing Exim down by computing (never used) parameters for the
782 RSA_EXPORT functionality.
784 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
785 authenticator to fail if there's a tab character in the incoming data
786 (there should never be unless someone is messing about, as it's supposed
787 to be base64-encoded). Also added, on Timo's advice, the "secured" option
788 if the connection is using TLS or if the remote IP is the same as the
789 local IP, and the "valid-client-cert option" if a client certificate has
792 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
793 authenticators. This can be used for authorization after authentication
794 succeeds. (In the case of plaintext, it servers for both authentication
797 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
798 if any retry times were supplied.
800 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
801 connection, where there is no client IP address to check. In this
802 situation, the verify now always succeeds.
804 PH/25 Applied John Jetmore's -Mset patch.
806 PH/26 Added -bem to be like -Mset, but loading a message from a file.
808 PH/27 In a string expansion for a processed (not raw) header when multiple
809 headers of the same name were present, leading whitespace was being
810 removed from all of them, but trailing whitespace was being removed only
811 from the last one. Now trailing whitespace is removed from each header
812 before concatenation. Completely empty headers in a concatenation (as
815 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
816 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
818 PH/29 [Removed. This was a change that I later backed out, and forgot to
819 correct the ChangeLog entry (that I had efficiently created) before
820 committing the later change.]
822 PH/30 Exim was sometimes attempting to deliver messages that had suffered
823 address errors (4xx response to RCPT) over the same connection as other
824 messages routed to the same hosts. Such deliveries are always "forced",
825 so retry times are not inspected. This resulted in far too many retries
826 for the affected addresses. The effect occurred only when there were more
827 hosts than the hosts_max_try setting in the smtp transport when it had
828 the 4xx errors. Those hosts that it had tried were not added to the list
829 of hosts for which the message was waiting, so if all were tried, there
830 was no problem. Two fixes have been applied:
832 (i) If there are any address or message errors in an SMTP delivery, none
833 of the hosts (tried or untried) are now added to the list of hosts
834 for which the message is waiting, so the message should not be a
835 candidate for sending over the same connection that was used for a
836 successful delivery of some other message. This seems entirely
837 reasonable: after all the message is NOT "waiting for some host".
838 This is so "obvious" that I'm not sure why it wasn't done
839 previously. Hope I haven't missed anything, but it can't do any
840 harm, as the worst effect is to miss an optimization.
842 (ii) If, despite (i), such a delivery is accidentally attempted, the
843 routing retry time is respected, so at least it doesn't keep
844 hammering the server.
846 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
847 in ${readsocket because some servers need this prod.
849 PH/32 Added some extra debug output when updating a wait-xxx database.
851 PH/33 The hint "could be header name not terminated by colon", which has been
852 given for certain expansion errors for a long time, was not being given
853 for the ${if def:h_colon_omitted{... case.
855 PH/34 The spec says: "With one important exception, whenever a domain list is
856 being scanned, $domain contains the subject domain." There was at least
857 one case where this was not true.
859 PH/35 The error "getsockname() failed: connection reset by peer" was being
860 written to the panic log as well as the main log, but it isn't really
861 panic-worthy as it just means the connection died rather early on. I have
862 removed the panic log writing for the ECONNRESET error when getsockname()
865 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
866 runs only) independently of the message's sender address. This meant
867 that, if the 4xx error was in fact related to the sender, a different
868 message to the same recipient with a different sender could confuse
869 things. In particualar, this can happen when sending to a greylisting
870 server, but other circumstances could also provoke similar problems.
871 I have changed the default so that the retry time for these errors is now
872 based a combination of the sender and recipient addresses. This change
873 can be overridden by setting address_retry_include_sender=false in the
876 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
877 remote server are returned as part of bounce messages. This was not
878 happening for LMTP over a pipe (the lmtp transport), but now it is the
879 same for both kinds of LMTP.
881 PH/38 Despite being documented as not happening, Exim was rewriting addresses
882 in header lines that were in fact CNAMEs. This is no longer the case.
884 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
885 and queue runs started by the daemon processed all messages. This has
886 been fixed so that -R and -S can now usefully be given with -q<time>.
888 PH/40 Import PCRE release 6.7 (fixes some bugs).
890 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
892 PH/42 Give an error if -q is specified more than once.
894 PH/43 Renamed the variables $interface_address and $interface_port as
895 $received_ip_address and $received_port, to make it clear that these
896 values apply to message reception, and not to the outgoing interface when
897 a message is delivered. (The old names remain recognized, of course.)
899 PH/44 There was no timeout on the connect() call when using a Unix domain
900 socket in the ${readsocket expansion. There now is.
902 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
903 be meaningful with "accept".
905 SC/01 Eximstats V1.43
906 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
908 SC/02 Eximstats V1.44
909 Use a glob alias rather than an array ref in the generated
910 parser. This improves both readability and performance.
912 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
913 Collect SpamAssassin and rejection statistics.
914 Don't display local sender or destination tables unless
915 there is data to show.
916 Added average volumes into the top table text output.
918 SC/04 Eximstats V1.46
919 Collect data on the number of addresses (recipients)
920 as well as the number of messages.
922 SC/05 Eximstats V1.47
923 Added 'Message too big' to the list of mail rejection
924 reasons (thanks to Marco Gaiarin).
926 SC/06 Eximstats V1.48
927 Mainlog lines which have GMT offsets and are too short to
928 have a flag are now skipped.
930 SC/07 Eximstats V1.49 (Alain Williams)
931 Added the -emptyok flag.
933 SC/08 Eximstats V1.50
934 Fixes for obtaining the IP address from reject messages.
936 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
937 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
938 whitesspace changes from 4.64-PH/27
940 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
943 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
944 are found, allow negative numbers in numeric criteria)
946 JJ/06 exipick.20061117.2, added new $message_body_missing variable
948 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
949 to match changes made in 4.64-PH/43
951 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
953 PH/47 Put in an explicit test for a DNS lookup of an address record where the
954 "domain" is actually an IP address, and force a failure. This locks out
955 those revolvers/nameservers that support "A-for-A" lookups, in
956 contravention of the specifications.
958 PH/48 When a host name was looked up from an IP address, and the subsequent
959 forward lookup of the name timed out, the host name was left in
960 $sender_host_name, contrary to the specification.
962 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
963 restricted to single-key lookups, Exim was not diagnosing an error if
964 * or *@ was used with a query-style lookup.
966 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
968 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
969 long ago, but noone interested enough thought of it. Let's just say that
970 the "1.1" means that there are some new functions that weren't there at
971 some point in the past.
973 PH/51 Error processing for expansion failure of helo_data from an smtp
974 transport during callout processing was broken.
976 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
977 tested/used via the -bh/-bhc/-bs options.
979 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
980 bug, fixed in subsequent PCRE releases).
982 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
983 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
985 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
991 SC/01 Use a glob alias rather than an array ref in eximstats generated
992 parser. This improves both readability and performance.
994 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
995 Don't display local sender or destination tables in eximstats unless
996 there is data to show.
997 Added average volumes into the eximstats top table text output.
999 SC/03 Collect data on the number of addresses (recipients) as well
1000 as the number of messages in eximstats.
1002 TF/01 Correct an error in the documentation for the redirect router. Exim
1003 does (usually) call initgroups() when daemonizing.
1005 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1006 with consistent privilege compared to when running as a daemon.
1008 TF/03 Note in the spec that $authenticated_id is not set for local
1009 submissions from trusted users.
1011 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1012 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1014 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1015 by adding some example configuration directives to the default
1016 configuration file. A little bit of work is required to uncomment the
1017 directives and define how usernames and passwords are checked, but
1018 there is now a framework to start from.
1020 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1021 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1022 without this. I don't know how relevant this is to other LDAP libraries.
1024 PH/02 Add the verb name to the "unknown ACL verb" error.
1026 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1028 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1030 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1031 directory not expanded when it should be if an expanded home directory
1032 was set for the address (which is overridden by the transport).
1034 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1037 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1038 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1039 because it is too late at that time, and has no effect.
1041 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1042 security issue with \' (bugzilla #107). I could not use the
1043 PQescapeStringConn() function, because it needs a PGconn value as one of
1046 PH/08 When testing addresses using -bt, indicate those final addresses that
1047 are duplicates that would not cause an additional delivery. At least one
1048 person was confused, thinking that -bt output corresponded to deliveries.
1049 (Suppressing duplicates isn't a good idea as you lose the information
1050 about possibly different redirections that led to the duplicates.)
1052 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1053 systems where poll() doesn't work, in particular OS X.
1055 PH/10 Added more information to debugging output for retry time not reached.
1057 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1058 operations in malware.c.
1060 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1063 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1064 syslog_duplication set false, log lines that would normally be written
1065 both the the main log and to the reject log were not written to syslog at
1068 PH/14 In the default configuration, change the use of "message" in ACL warn
1069 statements to "add_header".
1071 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1072 not followed by a command (e.g. "seen endif").
1074 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1075 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1078 PH/17 Added extra conditions to the default value of delay_warning_condition
1082 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1083 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1084 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1087 The Auto-Submitted: and various List- headers are standardised, whereas I
1088 don't think Precedence: ever was.
1090 PH/18 Refactored debugging code in route_finduser() to show more information,
1091 in particular, the error code if getpwnam() issues one.
1093 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1094 This is apparently needed in addition to the PH/07 change above to avoid
1095 any possible encoding problems.
1097 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1098 but not after initializing Perl.
1100 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1101 output them only if debugging. By default they are written stderr,
1102 apparently, which is not desirable.
1104 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1107 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1110 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1112 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1113 authenticated or an ident call has been made. Suppress the default
1114 values for $authenticated_id and $authenticated_sender (but permit -oMai
1115 and -oMas) when testing with -bh.
1117 PH/24 Re-jigged the order of the tests in the default configuration so that the
1118 tests for valid domains and recipients precede the DNS black list and CSA
1119 tests, on the grounds that those ones are more expensive.
1121 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1122 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1123 This bug exists in every version of Exim that I still have, right back to
1126 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1127 However, an attempt to turn on case-sensitivity in a regex key by
1128 including (?-i) didn't work because the subject string was already
1129 lowercased, and the effects were non-intuitive. It turns out that a
1130 one-line patch can be used to allow (?-i) to work as expected.
1136 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1137 other effects) broke the use of negated acl sub-conditions.
1139 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1142 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1143 "Deny" causes Exim to reject the incoming connection with a 554 error.
1144 Unfortunately, if there is a major crisis, such as a disk failure,
1145 tcp-wrappers gives "deny", whereas what one would like would be some
1146 kind of temporary error. A kludge has been added to help with this.
1147 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1148 554 error is used if errno is still zero or contains ENOENT (which occurs
1149 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1152 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1154 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1155 errors. Otherwise a message that provokes a temporary error (when other
1156 messages do not) can cause a whole host to time out.
1158 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1159 addresses were routed directly to files or pipes from a redirect router.
1160 File deliveries just didn't batch; pipe deliveries might have suffered
1163 PH/06 A failure to get a lock for a hints database would erroneously always say
1164 "Failed to get write lock", even when it was really a read lock.
1166 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1167 of 0600. This has been changed to use the value of the lockfile_mode
1168 option (which defaults to 0600).
1170 PH/08 Applied small patch from the Sieve maintainer.
1172 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1173 folder from quota calculations, a direct delivery into this folder messed
1174 up the contents of the maildirsize file. This was because the regex was
1175 used only to exclude .Trash (or whatever) when the size of the mailbox
1176 was calculated. There was no check that a delivery was happening into an
1177 excluded directory. This bug has been fixed by ignoring all quota
1178 processing for deliveries into excluded directories.
1180 PH/10 Added the maildirfolder_create_regex option to appendfile.
1186 PH/01 The code for finding all the local interface addresses on a FreeBSD
1187 system running IPv6 was broken. This may well have applied to all BSD
1188 systems, as well as to others that have similar system calls. The broken
1189 code found IPv4 interfaces correctly, but gave incorrect values for the
1190 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1191 that it would not match correctly against @[] and not recognize the IPv6
1194 PH/02 The ipliteral router was not recognizing addresses of the form user@
1195 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1197 PH/03 Added disable_ipv6.
1199 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1200 decoded form, because it is most often used to construct To: headers
1201 lines in autoreplies, and the decoded form may well be syntactically
1202 invalid. However, $reply_address has leading white space removed, and all
1203 newlines turned into spaces so that the autoreply transport does not
1206 PH/05 If group was specified without a user on a router, and no group or user
1207 was specified on a transport, the group from the router was ignored.
1209 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1210 for visible compile-time settings that can be used to change these
1211 numbers, for those that want even more. Backwards compatibility with old
1212 spool files has been maintained. However, going back to a previous Exim
1213 release will lost any variables that are in spool files.
1215 PH/07 Two small changes when running in the test harness: increase delay when
1216 passing a TCP/IP connection to a new process, in case the original
1217 process has to generate a bounce, and remove special handling of
1218 127.0.0.2 (sic), which is no longer necessary.
1220 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1221 be the same on different OS.
1223 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1226 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1227 whether --show-vars was specified or not
1229 JJ/02 exipick: Added support for new ACL variable spool format introduced
1232 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1233 syntactically invalid From: or Reply-to: line, and a filter used this to
1234 generate an autoreply, and therefore failed to obtain an address for the
1235 autoreply, Exim could try to deliver to a non-existent relative file
1236 name, causing unrelated and misleading errors. What now happens is that
1237 it logs this as a hard delivery error, but does not attempt to create a
1240 PH/11 The exinext utility has a -C option for testing purposes, but although
1241 the given file was scanned by exinext itself; it wasn't being passed on
1242 when Exim was called.
1244 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1245 an end-of-file indication when reading a command response.
1247 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1248 compiled. In many other places in Exim, IPv6 addresses are always
1249 recognized, so I have changed this. It also means that IPv4 domain
1250 literals of the form [IPV4:n.n.n.n] are now always recognized.
1252 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1253 used if the router is not running as root, for example, when verifying at
1254 ACL time, or when using -bh. The debugging output from this situation was
1255 non-existent - all you got was a failure to exec. I have made two
1258 (a) Failures to set uid/gid, the current directory, or a process leader
1259 in a subprocess such as that created by queryprogram now generate
1260 suitable debugging ouput when -d is set.
1262 (b) The queryprogram router detects when it is not running as root,
1263 outputs suitable debugging information if -d is set, and then runs
1264 the subprocess without attempting to change uid/gid.
1266 PH/15 Minor change to Makefile for building test_host (undocumented testing
1269 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1270 additional section of a DNS packet that returns MX or SRV records.
1271 Instead, it always explicitly searches for A/AAAA records. This avoids
1272 major problems that occur when a DNS server includes only records of one
1273 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1274 fixed another bug: if SRV records were looked up and the corresponding
1275 address records were *not* found in the additional section, the port
1276 values from the SRV records were lost.
1278 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1279 using the correct key (the original address) when searching the retry
1280 rules in order to find which one to use for generating the retry hint.
1282 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1283 adding the default one. Similarly, if it contains a Reply-To: header, the
1284 errors_reply_to option, if set, is not used.
1286 PH/19 When calculating a retry time, Exim used to measure the "time since
1287 failure" by looking at the "first failed" field in the retry record. Now
1288 it does not use this if it is later than than the arrival time of the
1289 message. Instead it uses the arrival time. This makes for better
1290 behaviour in cases where some deliveries succeed, thus re-setting the
1291 "first failed" field. An example is a quota failure for a huge message
1292 when small messages continue to be delivered. Without this change, the
1293 "time since failure" will always be short, possible causing more frequent
1294 delivery attempts for the huge message than are intended.
1295 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1297 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1298 $1, $2, $3) because the numerical variables can be reset during some
1299 expansion items (e.g. "match"), thereby losing the authentication data.
1301 PH/21 Make -bV show the size of off_t variables so that the test suite can
1302 decide whether to run tests for quotas > 2G.
1304 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1305 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1306 filecount value is greater than 2G or if a quota value is greater than 2G
1307 on a system where the size of off_t is not greater than 4, a panic error
1310 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1311 never match. The debug and -bh output now contains an explicit error
1312 message indicating a malformed IPv4 address or mask.
1314 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1315 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1316 PH/23 above applies.
1318 PH/25 Do not write to syslog when running in the test harness. The only
1319 occasion when this arises is a failure to open the main or panic logs
1320 (for which there is an explicit test).
1322 PH/26 Added the /no_tell option to "control=freeze".
1324 PH/27 If a host name lookup failed very early in a connection, for example, if
1325 the IP address matched host_lookup and the reverse lookup yielded a name
1326 that did not have a forward lookup, an error message of the form "no IP
1327 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1328 could be logged. Now it outputs the IP address instead of "NULL".
1330 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1331 allows the sender and the authenticated sender to be set when
1332 submitting a message from within Exim. Since child_open_exim() is
1333 documented for local_scan(), the new function should be too.
1335 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1336 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1337 results in an empty string is now treated as unset.
1339 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1341 PH/31 Added sender_verify_fail logging option.
1343 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1344 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1345 tidied the source and removed it altogether.
1347 PH/33 When a queue run was abandoned because the load average was too high, a
1348 log line was always written; now it is written only if the queue_run log
1349 selector is set. In addition, the log line for abandonment now contains
1350 information about the queue run such as the pid. This is always present
1351 in "start" and "stop" lines but was omitted from the "abandon" line.
1353 PH/34 Omit spaces between a header name and the colon in the error message that
1354 is given when verify = headers_syntax fails (if there are lots of them,
1355 the message gets confusing).
1357 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1358 names, as there are now some PTR records that contain slashes. This check
1359 is only to protect against broken name servers that fall over on strange
1360 characters, so the fact that it applies to all lookups doesn't matter.
1362 PH/36 Now that the new test suite is complete, we can remove some of the
1363 special code in Exim that was needed for the old test suite. For example,
1364 sorting DNS records because real resolvers return them in an arbitrary
1365 order. The new test suite's fake resolver always returns records in the
1368 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1369 bounces) except when queue_only is set, to avoid logging races between
1370 the different processes.
1372 PH/38 Panic-die if .include specifies a non-absolute path.
1374 PH/39 A tweak to the "H" retry rule from its user.
1376 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1377 a label. They prevented compilation on older perls.
1379 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1380 a warning to be raised on newish perls.
1382 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1383 on queue. Changes to match documented behaviour of showing count of
1384 messages matching specified criteria.
1386 PH/40 Changed the default ident timeout from 30s to 5s.
1388 PH/41 Added support for the use of login_cap features, on those BSD systems
1389 that have them, for controlling the resources used by pipe deliveries.
1391 PH/42 The content-scanning code uses fopen() to create files in which to put
1392 message data. Previously it was not paying any attention to the mode of
1393 the files. Exim runs with umask(0) because the rest of the code creates
1394 files with open(), and sets the required mode explicitly. Thus, these
1395 files were ending up world-writeable. This was not a big issue, because,
1396 being within the spool directory, they were not world-accessible. I have
1397 created a function called modefopen, which takes an additional mode
1398 argument. It sets umask(777), creates the file, chmods it to the required
1399 mode, then resets the umask. All the relevant calls to fopen() in the
1400 content scanning code have been changed to use this function.
1402 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1403 to 24 hours. This avoids potential overflow problems when processing G
1404 and H retry rules. I suspect nobody ever tinkers with this value.
1406 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1408 PH/45 When the plaintext authenticator is running as a client, the server's
1409 challenges are checked to ensure they are valid base64 strings. By
1410 default, the authentication attempt is cancelled if an invalid string is
1411 received. Setting client_ignore_invalid_base64 true ignores these errors.
1412 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1413 they are received. Thus, the responses can be made to depend on the
1414 challenges. If an invalid string is ignored, an empty string is placed in
1417 PH/46 Messages that are created by the autoreply transport now contains a
1418 References: header, in accordance with RFCs 2822 and 3834.
1420 PH/47 Added authenticated_sender_force to the smtp transport.
1422 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1424 PH/49 Installed latest patch from the Sieve maintainer.
1426 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1427 was also set, the appendfile transport was unnecessarily scanning a
1428 directory of message files (e.g. for maildir delivery) to find the count
1429 of files (along with the size), even though it did not need this
1430 information. It now does the scan only if it needs to find either the
1431 size of the count of files.
1433 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1435 PH/52 Two bugs concerned with error handling when the smtp transport is
1438 (i) Exim was not creating retry information for temporary errors given
1439 for individual recipients after the DATA command when the smtp transport
1440 was used in LMTP mode. This meant that they could be retried too
1441 frequently, and not timed out correctly.
1443 (ii) Exim was setting the flag that allows error details to be returned
1444 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1445 recipients that were returned after the DATA command.
1447 PH/53 This is related to PH/52, but is more general: for any failing address,
1448 when detailed error information was permitted to be returned to the
1449 sender, but the error was temporary, then after the final timeout, only
1450 "retry timeout exceeded" was returned. Now it returns the full error as
1451 well as "retry timeout exceeded".
1453 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1454 do this, and (what is worse) MTAs that accept it.
1456 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1457 will now be deprecated.
1459 PH/56 New os.c-cygwin from the Cygwin maintainer.
1461 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1462 formats (previously only available in exim formats via -bpr, -bpru,
1463 and -bpra. Now also available in native and exiqgrep formats)
1465 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1466 with very large, slow to parse queues
1468 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1470 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1472 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1473 responses to authentication challenges, though it was showing the
1474 challenges; (ii) I've removed the CR characters from the debug output for
1477 PH/58 Allow for the insertion of a newline as well as a space when a string
1478 is turned into more than one encoded-word during RFC 2047 encoding. The
1479 Sieve code now uses this.
1481 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1482 data_4xx, lost_connection, tls_required.
1484 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1485 message was being sent as an SMTP response.
1487 PH/61 Add -l and -k options to exicyclog.
1489 PH/62 When verifying, if an address was redirected to one new address, so that
1490 verification continued, and the new address failed or deferred after
1491 having set something in $address_data, the value of $address_data was not
1492 passed back to the ACL. This was different to the case when no
1493 redirection occurred. The value is now passed back in both cases.
1495 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1496 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1497 use login_cap.h, so on its own it isn't the distinguishing feature. The
1498 new name refers directly to the setclassresources() function.
1500 PH/65 Added configuration files for NetBSD3.
1502 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1504 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1505 is preferred over IPv4.
1507 PH/68 The bounce_return_message and bounce_return_body options were not being
1508 honoured for bounces generated during the reception of non-SMTP messages.
1509 In particular, this applied to messages rejected by the ACL. This bug has
1510 been fixed. However, if bounce_return_message is true and bounce_return_
1511 body is false, the headers that are returned for a non-SMTP message
1512 include only those that have been read before the error was detected.
1513 (In the case of an ACL rejection, they have all been read.)
1515 PH/69 The HTML version of the specification is now built in a directory called
1516 spec_html instead of spec.html, because the latter looks like a path with
1517 a MIME-type, and this confuses some software.
1519 PH/70 Catch two compiler warnings in sieve.c.
1521 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1522 function verify_get_ident() calls ip_connect() to connect a socket, but
1523 if the "connect()" function timed out, ip_connect() used to close the
1524 socket. However, verify_get_ident() also closes the socket later, and in
1525 between Exim writes to the log, which may get opened at this point. When
1526 the socket was closed in ip_connect(), the log could get the same file
1527 descriptor number as the socket. This naturally causes chaos. The fix is
1528 not to close the socket in ip_connect(); the socket should be closed by
1529 the function that creates it. There was only one place in the code where
1530 this was missing, in the iplookup router, which I don't think anybody now
1531 uses, but I've fixed it anyway.
1533 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1534 well as to direct DNS lookups. Otherwise the handling of names in host
1535 lists is inconsistent and therefore confusing.
1541 PH/01 Two changes to the default runtime configuration:
1543 (1) Move the checks for relay_from_hosts and authenticated clients from
1544 after to before the (commented out) DNS black list checks.
1546 (2) Add control=submission to the relay_from_hosts and authenticated
1547 clients checks, on the grounds that messages accepted by these
1548 statements are most likely to be submissions.
1550 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1552 (1) Generate an error if the third argument for the ${prvs expansion is
1555 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1558 (3) Reset the variables that are obtained from the first argument of
1559 ${prvscheck and used in the second argument before leaving the code,
1560 because their memory is reclaimed, so using them afterwards may do
1563 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1564 one (it's much easier than Tom thought :-).
1566 (5) Because of (4), we can now allow for the use of $prvscheck_result
1567 inside the third argument.
1569 PH/03 For some reason, the default setting of PATH when running a command from
1570 a pipe transport was just "/usr/bin". I have changed it to
1573 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1574 anything to be listed in the output from -bV.
1576 PH/05 When a filter generated an autoreply, the entire To: header line was
1577 quoted in the delivery log line, like this:
1579 => >A.N.Other <ano@some.domain> <original@ddress> ...
1581 This has been changed so that it extracts the operative address. There
1582 may be more than one such address. If so, they are comma-separated, like
1585 => >ano@some.domain,ona@other.domain <original@ddress> ...
1587 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1588 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1589 not being looked up in the rDNS to get a host name, Exim was showing the
1590 IP address twice in Received: lines, even though the IP addresses were
1591 identical. For example:
1593 Received: from [1.2.3.4] (helo=[1.2.3.4])
1595 However, if the real host name was known, it was omitting the HELO data
1596 if it matched the actual IP address. This has been tidied up so that it
1597 doesn't show the same IP address twice.
1599 PH/07 When both +timestamp and +memory debugging was on, the value given by
1600 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1601 called by the debug printing, thereby overwriting the timestamp buffer.
1602 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1604 PH/08 When the original message was included in an autoreply transport, it
1605 always said "this is a copy of the message, including all the headers",
1606 even if body_only or headers_only was set. It now gives an appropriate
1609 PH/09 Applied a patch from the Sieve maintainer which:
1611 o fixes some comments
1612 o adds the (disabled) notify extension core
1613 o adds some debug output for the result of if/elsif tests
1614 o points to the current vacation draft in the documentation
1615 and documents the missing references header update
1619 o fixes a bug in processing the envelope test (when testing
1620 multiple envelope elements, the last element determinted the
1623 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1624 Electronic Mail") by including:
1626 Auto-submitted: auto-generated
1628 in the messages that it generates (bounce messages and others, such as
1629 warnings). In the case of bounce messages for non-SMTP mesages, there was
1630 also a typo: it was using "Auto_submitted" (underscore instead of
1631 hyphen). Since every message generated by Exim is necessarily in response
1632 to another message, thes have all been changed to:
1634 Auto-Submitted: auto-replied
1636 in accordance with these statements in the RFC:
1638 The auto-replied keyword:
1640 - SHOULD be used on messages sent in direct response to another
1641 message by an automatic process,
1643 - MUST NOT be used on manually-generated messages,
1645 - MAY be used on Delivery Status Notifications (DSNs) and Message
1646 Disposition Notifications (MDNs),
1648 - MUST NOT be used on messages generated by automatic or periodic
1649 processes, except for messages which are automatic responses to
1652 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1653 to the default Received: header definition.
1655 PH/12 Added log selector acl_warn_skipped (default on).
1657 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1658 variables because (a) they are in the wrong storage pool and (b) even if
1659 they were copied, it wouldn't work properly because of the caching.
1661 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1662 checking when decoding. Apparently there are clients that generate
1663 overlong encoded strings. Why am I not surprised?
1665 PH/15 If the first argument of "${if match_address" was not empty, but did not
1666 contain an "@" character, Exim crashed. Now it writes a panic log message
1667 and treats the condition as false.
1669 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1671 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1672 extension "envelope-auth". The code is finished and in agreement with
1673 other implementations, but there is no documentation so far and in fact,
1674 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1675 not changing the active code.
1677 Print executed "if" and "elsif" statements when debugging is used. This
1678 helps a great deal to understand what a filter does.
1680 Document more things not specified clearly in RFC3028. I had all this
1681 sorted out, when out of a sudden new issues came to my mind. Oops."
1683 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1686 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1687 canonical form (as documented). However, after a host name lookup from
1688 the IP address, check_host() was doing a simple string comparison with
1689 addresses acquired from the DNS when checking that the found name did
1690 have the original IP as one of its addresses. Since any found IPv6
1691 addresses are likely to be in abbreviated form, the comparison could
1692 fail. Luckily, there already exists a function for doing the comparison
1693 by converting both addresses to binary, so now that is used instead of
1694 the text comparison.
1696 PH/20 There was another similar case to PH/19, when a complete host name was
1697 given in a host list; looking up its IP address could give an abbreviated
1698 form, whereas the current host's name might or might not be abbreviated.
1699 The same fix has been applied.
1705 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1706 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1709 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1710 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1712 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1714 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1715 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1716 Also, for "Auto-Submitted", treat anything other than "no" as
1717 non-personal, in accordance with RFC 3834. (Previously it treated
1718 anything starting "auto-" as non-personal.)
1720 TF/01 The control=submission/name=... option had a problem with syntax
1721 errors if the name included a slash character. The /name= option
1722 now slurps the rest of the string, so it can include any characters
1723 but it must come last in the list of options (after /sender_retain
1726 PH/05 Some modifications to the interface to the fake nameserver for the new
1734 TK/01 Added the "success_on_redirect" address verification option. See
1735 NewStuff for rationale and an example.
1737 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1739 PH/02 Patch to exigrep to allow it to work on syslog lines.
1741 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1742 fread() to skip over the body file's header line, because in Cygwin the
1743 header line is locked and is inaccessible.
1745 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1746 co-exist for some time) to make it clear that it is the Exim ID that is
1747 referenced, not the Message-ID: header line.
1749 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1750 string_format() function, because snprintf() does not exist on all
1753 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1754 ${stat: expansion item.
1756 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1757 protocol synchronization error", to keep the pedants happy.
1759 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1760 well as for IRIX systems, when gcc is being used. See the host.c source
1763 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1765 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1768 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1769 transport and to the smtp transport in LMTP mode.
1771 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1773 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1774 run a filter in a subprocess. This could lead to confusion in subsequent
1775 lookups in the parent process. There should also be a search_tidyup() at
1776 the end of the subprocess.
1778 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1779 only if the host matched helo_try_verify_hosts, which caused the
1780 verification to occur when the EHLO/HELO command was issued. The ACL just
1781 tested the remembered result. Now, if a previous verification attempt has
1782 not happened, "verify = helo" does it there and then.
1784 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1786 TK/03 Fix log output including CR from clamd.
1788 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1789 exist provoked a memory error which could cause a segfault.
1791 PH/15 Installed PCRE 6.2
1793 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1795 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1796 of the problem. Specifically, suggested +O2 rather than +O1 for the
1799 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1801 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1802 record, the port was not correctly logged when the outgoing_port log
1803 selector was set (it logged the transort's default port).
1805 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1806 fallback_hosts, and "hosts" in the smtp transport.
1808 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1809 host errors such as "Connection refused".
1811 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1812 authentication with radiusclient 0.4.9:
1814 - Error returned from rc_read_config was caught wrongly
1815 - Username/password not passed on to radius server due to wrong length.
1817 The presumption is that some radiusclient API changes for 4.51/PH/17
1818 were not taken care of correctly. The code is still untested by me (my
1819 Linux distribution still has 0.3.2 of radiusclient), but it was
1820 contributed by a Radius user.
1822 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1823 expanding the "port" option of the smtp transport.
1825 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1826 while reading a MIME header. Thanks to Tom Hughes for a patch.
1828 PH/24 Include config.h inside local_scan.h so that configuration settings are
1831 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1832 that in an ACL for RCPT (for example), you can examine exactly what was
1835 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1836 commands, but it was not correctly comparing the address with the actual
1837 client host address. Thus, it would show the EHLO address in Received:
1838 header lines when this was not necessary.
1840 PH/27 Added the % operator to ${eval:}.
1842 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1843 it should be ignoring failures (because with -C, for example, it has lost
1844 privilege). It wasn't ignoring creation failures other than "already
1847 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1850 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1851 because an input line was too long, either on its own, or by virtue of
1852 too many continuations, the temporary file was not being removed, and the
1853 return code was incorrect.
1855 PH/31 Missing "BOOL" in function definition in filtertest.c.
1857 PH/32 Applied Sieve patches from the maintainer.
1859 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1861 PH/33 Added "verify = not_blind".
1863 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1864 Local/Makefile (with some defaults set). These are used in built scripts
1865 such as exicyclog, but they have never been used in the exim_install
1866 script (though there are many overriding facilities there). I have
1867 arranged that the exim_install script now takes note of these two
1870 PH/35 Installed configuration files for Dragonfly.
1872 PH/36 When a locally submitted message by a trusted user did not contain a
1873 From: header, and the sender address was obtained from -f or from an SMTP
1874 MAIL command, and the trusted user did not use -F to supply a sender
1875 name, $originator_name was incorrectly used when constructing a From:
1876 header. Furthermore, $originator_name was used for submission mode
1877 messages from external hosts without From: headers in a similar way,
1878 which is clearly wrong.
1880 PH/37 Added control=suppress_local_fixups.
1882 PH/38 When log_selector = +received_sender was set, and the addition of the
1883 sender made the log line's construction buffer exactly full, or one byte
1884 less than full, an overflow happened when the terminating "\n" was
1887 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1888 when the result of a list match is failure because a DNS lookup failed.
1890 PH/40 RM_COMMAND is now used in the building process.
1892 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1893 the "build-* directories that it finds.
1895 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1896 address in a domain literal was a prefix of an interface address.
1898 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1899 when verifying a sender address, unless rewrite_headers is false.
1901 PH/44 Wrote a long comment about why errors_to addresses are verified as
1902 recipients, not senders.
1904 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1905 the ratelimit ACL was added.
1907 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1909 PH/46 Added extra information about PostgreSQL errors to the error string.
1911 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1912 avoiding the need to install special zones in a real server. This is
1913 backwards compatible; if it can't find the fake resolver, it drops back.
1914 Thus, both old and new test suites can be run.
1916 TF/02 Added util/ratelimit.pl
1918 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1919 clock is set back in time.
1921 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1922 Candler <B.Candler@pobox.com>.
1924 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1925 OK for addresses that are the result of redirections.
1927 PH/48 A number of further additions for the benefit of the new test suite,
1928 including a fake gethostbyname() that interfaces to the fake DNS resolver
1931 TF/06 The fix for widen_domains has also been applied to qualify_single and
1932 search_parents which are the other dnslookup options that can cause
1935 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1938 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1940 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1941 service" if it encountered an MX record with an empty target hostname.
1942 The message is now "an MX or SRV record indicated no SMTP service".
1944 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1945 if the DNS of the sending site is misconfigured. This is quite a
1946 common situation. This change restores the behaviour of treating a
1947 helo verification defer as a failure.
1949 PH/51 If self=fail was set on a router, the bounce message did not include the
1950 actual error message.
1956 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1958 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1959 command itself ended in error, the underlying message about the transport
1960 filter timeout was being overwritten with the pipe command error. Now the
1961 underlying error message should be appended to the second error message.
1963 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1965 PH/02 Reduce the amount of output that "make" produces by default. Full output
1966 can still be requested.
1968 PH/03 The warning log line about a condition test deferring for a "warn" verb
1969 was being output only once per connection, rather than after each
1970 occurrence (because it was using the same function as for successful
1971 "warn" verbs). This seems wrong, so I have changed it.
1973 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1974 it should not have, which might have caused a crash in the right
1975 circumstances, but probably never did.
1977 PH/04 Installed a modified version of Tony Finch's patch to make submission
1978 mode fix the return path as well as the Sender: header line, and to
1979 add a /name= option so that you can make the user's friendly name appear
1982 TF/03 Added the control = fakedefer ACL modifier.
1984 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1985 Mark Lowes for thorough testing.
1987 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1989 TK/03 Merged latest SRS patch from Miles Wilton.
1991 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1992 with the definition in sysexits.h (which is #included earlier).
1993 Fortunately, Exim does not actually use EX_OK. The code used to try to
1994 preserve the sysexits.h value, by assumimg that macro definitions were
1995 scanned for macro replacements. I have been disabused of this notion,
1996 so now the code just undefines EX_OK before #including unistd.h.
1998 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1999 in the smtp transport. When a block could not be written in a single
2000 write() function, the timeout was being re-applied to each part-write.
2001 This seems wrong - if the receiver was accepting one byte at a time it
2002 would take for ever. The timeout is now adjusted when this happens. It
2003 doesn't have to be particularly precise.
2005 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2006 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2008 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2009 without a domain if the check to <postmaster@domain> fails.
2011 SC/01 Eximstats: added -xls and the ability to specify output files
2012 (patch written by Frank Heydlauf).
2014 SC/02 Eximstats: use FileHandles for outputing results.
2016 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2018 SC/04 Eximstats: fixed display of large numbers with -nvr option
2020 SC/05 Eximstats: fixed merging of reports with empty tables.
2022 SC/06 Eximstats: added the -include_original_destination flag
2024 SC/07 Eximstats: removed tabs and trailing whitespace.
2026 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2028 TK/06 MBOX spool code: Add real "From " MBOX separator line
2029 so the .eml file is really in mbox format (even though
2030 most programs do not really care). Patch from Alex Miller.
2032 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2033 The latter is generated from $received_to and is only set if the
2034 message has one envelope recipient. SA can use these headers,
2035 obviously out-of-the-box. Patch from Alex Miller.
2037 PH/08 The ${def test on a variable was returning false if the variable's
2038 value was "0", contrary to what the specification has always said!
2039 The result should be true unless the variable is empty.
2041 PH/09 The syntax error of a character other than { following "${if
2042 def:variable_name" (after optional whitespace) was not being diagnosed.
2043 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2044 accidental colon was present, for example, could give incorrect results.
2046 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2047 result is used (not including appendfile, where other changes are about
2050 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2051 This involved changing a lot of size variables from int to off_t. It
2052 should work with maildirs and everything.
2054 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2055 spamd dying while we are connected to it.
2057 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2060 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2061 function for BDB 4.3.
2063 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2065 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2066 the sources for PCRE have been re-arranged and more files are now
2069 PH/15 The code I had for printing potentially long long variables in PH/11
2070 above was not the best (it lost precision). The length of off_t variables
2071 is now inspected at build time, and an appropriate printing format (%ld
2072 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2073 to be "long long int" or "long int". This is needed for the internal
2074 formatting function string_vformat().
2076 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2077 the configuration file to be ":syslog", then the script "guesses" where
2078 the logs files are, rather than using the compiled in default. In our
2079 case the guess is not the same as the compiled default, so the script
2080 suddenly stopped working when I started to use syslog. The patch checks
2081 to see if log_file_path is "". If so, it attempts to read it from exim
2082 with no configuration file to get the compiled in version, before it
2083 falls back to the previous guessing code."
2085 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2086 implementing BATV in an Exim configuration. See NewStuff for the gory
2089 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2090 Makefile that are specific to HP-UX.
2092 PH/18 If the "use_postmaster" option was set for a recipient callout together
2093 with the "random" option, the postmaster address was used as the MAIL
2094 FROM address for the random test, but not for the subsequent recipient
2095 test. It is now used for both.
2097 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2098 patch removes a few documentation additions to RFC 3028, because the
2099 latest draft now contains them. It adds the new en;ascii-case comparator
2100 and a new error check for 8bit text in MIME parts. Comparator and
2101 require names are now matched exactly. I enabled the subaddress
2102 extension, but it is not well tested yet (read: it works for me)."
2104 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2105 rework some of the code of TK/09 above to avoid the hardwired use of
2106 "%lld" and "long long". Replaced the call to snprintf() with a call to
2109 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2110 records point to non-existent hosts", "retry timeout exceeded", and
2111 "retry time not reached for any host after a long failure period".
2113 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2114 experimental DomainKeys support:
2116 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2117 (2) On an error such as an illegally used "control", the wrong name for
2118 the control was given.
2120 These problems did NOT occur unless DomainKeys support was compiled.
2122 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2124 PH/24 Added ${if match_ip condition.
2126 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2127 databases so that it will be absolutely obvious if a crash occurs in the
2128 DB library. This is a regular occurrence (often caused by mis-matched
2131 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2132 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2133 picked up on a user's system that detects such things. There doesn't seem
2134 to be a gcc warning option for this - only an attribute that has to be
2135 put on the function's prototype. It seems that in Fedora Core 4 they have
2136 set this on a number of new functions. No doubt there will be more in due
2139 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2140 specify a transport. However, if an address that was verified by such a
2141 router was the subject of a callout, Exim crashed because it tried to
2142 read the rcpt_include_affixes from the non-existent transport. Now it
2143 just assumes that the setting of that option is false. This bug was
2144 introduced by 4.51/PH/31.
2146 PH/28 Changed -d+all to exclude +memory, because that information is very
2147 rarely of interest, but it makes the output a lot bigger. People tend to
2148 do -d+all out of habit.
2150 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2151 code in os-type was giving problems when libc.so lives in lib64, like on
2154 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2155 aren't the modern standard, and it seems that some systems' include files
2156 don't always have them. Exim was already checking for some of the newer
2157 ones like T_AAAA, and defining it itself. I've added checks for all the
2158 record types that Exim uses.
2160 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2161 not automatically generating a new one, as it is supposed to. This
2162 prevented TLS from working. If the file did exist, but contained invalid
2163 data, a new version was generated, as expected. It was only the case of a
2164 non-existent file that was broken.
2166 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2167 with a change in libdomainkeys > 0.64.
2169 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2170 from DNS. If the selector record carries the flag, it now has
2171 precedence over the domain-wide flag.
2173 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2175 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2176 the use of a port name that isn't defined in /etc/services) occurred, the
2177 message was deferred as in a normal delivery, and thus remained on the
2178 spool, instead of being failed because of the mua_wrapper setting. This
2179 is now fixed, and I tidied up some of the mua_wrapper messages at the
2182 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2183 the messages in a hash of arrays rather than using individual hashes.
2184 This is a bit cleaner and results in dramatic memory savings, albeit
2185 at a slight CPU cost.
2187 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2188 as requested by Marc Sherman.
2190 SC/10 Eximstats: added histograms for user specified patterns as requested
2193 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2195 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2196 fopen() in the content-scanning modules that did not already have it.
2202 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2203 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2205 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2207 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2209 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2210 to test Sieve filters that use "vacation".
2212 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2213 that changes the way the GnuTLS parameters are stored in the cache file.
2214 The new format can be generated externally. For backward compatibility,
2215 if the data in the cache doesn't make sense, Exim assumes it has read an
2216 old-format file, and it generates new data and writes a new file. This
2217 means that you can't go back to an older release without removing the
2220 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2221 work if there are any delivery delays because "one_time" forces the
2222 parent to be marked "delivered", so its unseen clone is never tried
2223 again. For this reason, Exim now forbids the simultaneous setting of
2226 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2227 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2228 entry if you want to know the details. The fix, however, neglected to
2229 consider the case when local delivery batching is involved. The test for
2230 "previously delivered" was not happening when checking to see if an
2231 address could be batched with a previous (undelivered) one; under
2232 certain circumstances this could lead to multiple deliveries to the same
2235 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2236 in its include files, and this causes problems building Exim.
2238 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2239 header_syntax) but Exim was just ignoring anything given after a slash.
2240 In particular, this caused confusion with an attempt to use "verify =
2241 reverse_host_lookup/defer_ok". An error is now given when options are
2242 supplied for verify items that do not have them. (Maybe reverse_host_
2243 lookup should have a defer_ok option, but that's a different point.)
2245 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2246 defined by RFC 821) to 2048, because there were problems with some AUTH
2247 commands, and RFC 1869 says the size should be increased for extended
2248 SMTP commands that take arguments.
2250 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2253 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2254 "unknown" error; now it says that the functionality isn't in the binary.
2256 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2257 an address' error message when a string expansion fails (syntax or
2258 whatever). Otherwise the password may appear in the log. Following change
2259 PH/42 below, there is no longer a chance of it appearing in a bounce
2262 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2264 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2265 of its addresses was ever tried. (Bugzilla bug #2.)
2267 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2268 the result incorrectly in the debug output. (It correctly added a newline
2269 to what was transported.)
2271 TF/01 Added $received_time.
2273 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2274 commented out examples of how to interface to a virus scanner and to
2275 SpamAssassin. Also added commented examples of av_scanner and
2276 spamd_address settings.
2278 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2279 and controls are allowed in which ACLs. There were a couple of minor
2280 errors. Some of the entries in the conditions table (which is a table of
2281 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2282 negation of where the condition IS allowed.
2284 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2286 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2287 header file does not have a version number, so I've had to invent a new
2288 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2289 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2290 radiusclient), but it was contributed by a Radius user.
2292 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2293 files or directories, for OpenSSL.
2295 PH/19 When an Exim process that is running as root has to create an Exim log
2296 file, it does so in a subprocess that runs as exim:exim so as to get the
2297 ownership right at creation (otherwise, other Exim processes might see
2298 the file with the wrong ownership). There was no test for failure of this
2299 fork() call, which would lead to the process getting stuck as it waited
2300 for a non-existent subprocess. Forks do occasionally fail when resources
2301 run out. I reviewed all the other calls to fork(); they all seem to check
2304 PH/20 When checking for unexpected SMTP input at connect time (before writing
2305 the banner), Exim was not dealing correctly with a non-positive return
2306 from the read() function. If the client had disconnected by this time,
2307 the result was a log entry for a synchronization error with an empty
2308 string after "input=" when read() returned zero. If read() returned -1
2309 (an event I could not check), uninitialized data bytes were printed.
2310 There were reports of junk text (parts of files, etc) appearing after
2313 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2315 PH/22 Added support for macro redefinition, and (re)definition in between
2316 driver and ACL definitions.
2318 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2319 forgetting to use the resulting value; it was using the unexpanded value.
2321 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2322 hadn't been configured. The fix is from Juergen Kreileder, who
2323 understands it better than I do:
2325 "Here's what I see happening with three configured cyrus_sasl
2326 authenticators configured (plain, login, cram-md5):
2328 On startup auth_cyrus_sasl_init() gets called for each of these.
2329 This means three calls to sasl_listmech() without a specified mech_list.
2330 => SASL tests which mechs of all available mechs actually work
2331 => three warnings about OTP not working
2332 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2334 With the patch, sasl_listmech() also gets called three times. But now
2335 SASL's mech_list option is set to the server_mech specified in the the
2336 authenticator. Or in other words, the answer from sasl_listmech()
2337 gets limited to just the mech you're testing for (which is different
2339 => the return list contains just 'plain' or 'login', 'cram-md5' or
2340 nothing depending on the value of ob->server_mech.
2342 I've just tested the patch: Authentication still works fine,
2343 unavailable mechs specified in the exim configuration are still
2344 caught, and the auth.log warnings about OTP are gone."
2346 PH/25 When debugging is enabled, the contents of the command line are added
2347 to the debugging output, even when log_selector=+arguments is not
2350 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2351 answer is "GNU", and only if the return is "GNU/something" is the answer
2354 PH/27 $acl_verify_message is now set immediately after the failure of a
2355 verification in an ACL, and so is available in subsequent modifiers. In
2356 particular, the message can be preserved by coding like this:
2358 warn !verify = sender
2359 set acl_m0 = $acl_verify_message
2361 Previously, $acl_verify_message was set only while expanding "message"
2362 and "log_message" when a very denied access.
2364 PH/28 Modified OS/os.c-Linux with
2366 -#ifndef OS_LOAD_AVERAGE
2367 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2369 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2370 nomenclature these days.)
2372 PH/29 Installed patch from the Sieve maintainer that adds the options
2373 sieve_useraddress and sieve_subaddress to the redirect router.
2375 PH/30 In these circumstances:
2376 . Two addresses routed to the same list of hosts;
2377 . First host does not offer TLS;
2378 . First host accepts first address;
2379 . First host gives temporary error to second address;
2380 . Second host offers TLS and a TLS session is established;
2381 . Second host accepts second address.
2382 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2383 and peerdn, if requested) that were in fact used only for the second
2386 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2387 attention to any local part prefix or suffix that was matched by the
2388 router that accepted the address. It now behaves in the same way as it
2389 does for delivery: the affixes are removed from the local part unless
2390 rcpt_include_affixes is set on the transport.
2392 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2393 timeout during the DATA phase of an incoming message.
2395 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2396 applied a patch sanctioned by the Sieve maintainer.
2398 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2399 the uid or gid is negative. A case of a negative gid caused this to be
2400 noticed. The fix allows for either to be negative.
2402 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2403 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2406 PH/36 The change PH/12 above was broken. Fixed it.
2408 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2409 the grounds that routing the same address twice would always produce the
2410 same answer. This might have been true once, but it is certainly no
2411 longer true now. Routing a child address may depend on the previous
2412 routing that produced that child. Some complicated redirection strategies
2413 went wrong when messages had multiple recipients, and made Exim's
2414 behaviour dependent on the order in which the addresses were given.
2416 I have moved the duplicate checking until after the routing is complete.
2417 Exim scans the addresses that are assigned to local and remote
2418 transports, and removes any duplicates. This means that more work will be
2419 done, as duplicates will always all be routed, but duplicates are
2420 presumably rare, so I don't expect this is of any significance.
2422 For deliveries to pipes, files, and autoreplies, the duplicate checking
2423 still happens during the routing process, since they are not going to be
2426 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2427 It corrects a timeout issue with spamd. This is Ian's comment: "The
2428 background is that sometimes spamd either never reads data from a
2429 connection it has accepted, or it never writes response data. The exiscan
2430 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2431 blindly assumes that writes won't block so it may never time out."
2433 PH/39 Allow G after quota size as well as K and M.
2435 PH/40 The value set for $authenticated_id in an authenticator may not contain
2436 binary zeroes or newlines because the value is written to log lines and
2437 to spool files. There was no check on this. Now the value is run through
2438 the string_printing() function so that such characters are converted to
2439 printable escape sequences.
2441 PH/41 $message_linecount is a new variable that contains the total number of
2442 lines in the message. Compare $body_linecount, which is the count for the
2445 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2446 bounce and delay warning messages, except in certain special cases, which
2449 (a) An SMTP error message from a remote host;
2450 (b) A message specified in a :fail: redirection;
2451 (c) A message specified in a "fail" command in a system filter;
2452 (d) A message specified in a FAIL return from the queryprogram router;
2453 (e) A message specified by the cannot_route_message router option.
2455 In these cases only, Exim does include the error details in bounce and
2456 warning messages. There are also a few cases where bland messages such
2457 as "unrouteable address" or "local delivery error" are given.
2459 PH/43 $value is now also set for the "else" part of a ${run expansion.
2461 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2462 being worked on, but at least Exim now implements the latest version to
2465 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2466 process to complete was treated as a delivery failure, a timeout while
2467 writing the message to the pipe was logged, but erroneously treated as a
2468 successful delivery. Such timeouts include transport filter timeouts. For
2469 consistency with the overall process timeout, these timeouts are now
2470 treated as errors, giving rise to delivery failures by default. However,
2471 there is now a new Boolean option for the pipe transport called
2472 timeout_defer, which, if set TRUE, converts the failures into defers for
2473 both kinds of timeout. A transport filter timeout is now identified in
2476 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2477 systems where "make" and "gmake" are different, calling "gmake" at top
2478 level broke things. I've arranged for the value of $(MAKE) to be passed
2479 from the Makefile to this script so that it can call the same version of
2483 A note about Exim versions 4.44 and 4.50
2484 ----------------------------------------
2486 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2487 changes of various kinds. As a consequence, a big documentation update was
2488 needed. This delayed the release for rather longer than seemed good, especially
2489 in the light of a couple of (minor) security issues. Therefore, the changes
2490 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2491 release. So 4.44 and 4.50 are in effect two different branches that both start
2494 I have left the 4.50 change log unchanged; it contains all the changes since
2495 4.43. The change log for 4.44 is below; many of its items are identical to
2496 those for 4.50. This seems to be the most sensible way to preserve the
2497 historical information.
2503 1. Minor wording change to the doc/README.SIEVE file.
2505 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2506 computation of the current number of files was incorrect.
2508 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2509 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2512 4. Give more explanation in the error message when the command for a transport
2513 filter fails to execute.
2515 5. There are several places where Exim runs a non-Exim command in a
2516 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2517 was being done only for the command run by the queryprogram router. It is
2518 now done for all such subprocesses. The other cases are: ${run, transport
2519 filters, and the commands run by the lmtp and pipe transports.
2521 6. Added CONFIGURE_GROUP build-time option.
2523 7. Some older OS have a limit of 256 on the maximum number of file
2524 descriptors. Exim was using setrlimit() to set 1000 as a large value
2525 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2526 systems. I've change it so that if it can't get 1000, it tries for 256.
2528 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2529 was an oversight, and furthermore, ever since the addition of extra
2530 controls (e.g. 4.43/32), the checks on when to allow different forms of
2531 "control" were broken. There should now be diagnostics for all cases when a
2532 control that does not make sense is encountered.
2534 9. Added the /retain_sender option to "control=submission".
2536 10. $recipients is now available in the predata ACL (oversight).
2538 11. Tidy the search cache before the fork to do a delivery from a message
2539 received from the command line. Otherwise the child will trigger a lookup
2540 failure and thereby defer the delivery if it tries to use (for example) a
2541 cached ldap connection that the parent has called unbind on.
2543 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2544 of $address_data from the recipient verification was clobbered by the
2545 sender verification.
2547 13. The value of address_data from a sender verification is now available in
2548 $sender_address_data in subsequent conditions in the ACL statement.
2550 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2552 15. Added a new option "connect=<time>" to callout options, to set a different
2555 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2556 was its contents. (It was OK if the option was not defined at all.)
2558 17. A "Completed" log line is now written for messages that are removed from
2559 the spool by the -Mrm option.
2561 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2562 information about exactly what failed.
2564 19. Added -dd to debug only the daemon process.
2566 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2567 handles timeouts, both on the server side and network timeouts. Renamed the
2568 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2570 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2571 It is now set to "smtps".
2573 22. $host_address is now set to the target address during the checking of
2574 ignore_target_hosts.
2576 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2577 being passed; this would have caused $sender_host_name to have been used if
2578 matching the list had actually called for a host name (not very likely,
2579 since this list is usually IP addresses). A host name is now passed as
2582 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2583 code that specifies a non-restarting handler (typically sigaction() in
2584 modern systems) in an attempt to fix a rare and obscure crash bug.
2586 25. Narrowed the window for a race in the daemon that could cause it to ignore
2587 SIGCHLD signals. This is not a major problem, because they are used only to
2588 wake it up if nothing else does.
2590 26. A malformed maildirsize file could cause Exim to calculate negative values
2591 for the mailbox size or file count. Odd effects could occur as a result.
2592 The maildirsize information is now recalculated if the size or filecount
2595 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2596 support for a long time. Removed HAVE_SYS_VFS_H.
2598 28. Installed the latest version of exipick from John Jetmore.
2600 29. In an address list, if the pattern was not a regular expression, an empty
2601 subject address (from a bounce message) matched only if the pattern was an
2602 empty string. Non-empty patterns were not even tested. This was the wrong
2603 because it is perfectly reasonable to use an empty address as part of a
2604 database query. An empty address is now tested by patterns that are
2605 lookups. However, all the other forms of pattern expect the subject to
2606 contain a local part and a domain, and therefore, for them, an empty
2607 address still always fails if the pattern is not itself empty.
2609 30. Exim went into a mad DNS loop when attempting to do a callout where the
2610 host was specified on an smtp transport, and looking it up yielded more
2611 than one IP address.
2613 31. Re-factored the code for checking spool and log partition space into a
2614 function that finds that data and another that does the check. The former
2615 is then used to implement four new variables: $spool_space, $log_space,
2616 $spool_inodes, and $log_inodes.
2618 32. The RFC2047 encoding function was originally intended for short strings
2619 such as real names; it was not keeping to the 75-character limit for
2620 encoded words that the RFC imposes. It now respects the limit, and
2621 generates multiple encoded words if necessary. To be on the safe side, I
2622 have increased the buffer size for the ${rfc2047: expansion operator from
2625 33. It is now permitted to omit both strings after an "if" condition; if the
2626 condition is true, the result is "true". As before, when the second string
2627 is omitted, a false condition yields an empty string. This makes it less
2628 cumbersome to write custom ACL and router conditions.
2630 34. Failure to deliver a bounce message always caused it to be frozen, even if
2631 there was an errors_to setting on the router. The errors_to setting is now
2634 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2635 canonical form (fully expanded) before being placed in
2636 $sender_host_address.
2638 36. The table in the code that translates DNS record types into text (T_A to
2639 "A" for instance) was missing entries for NS and CNAME. It is just possible
2640 that this could have caused confusion if both these types were looked up
2641 for the same domain, because the text type is used as part of Exim's
2642 per-process caching. But the chance of anyone hitting this buglet seems
2645 37. The dnsdb lookup has been extended in a number of ways.
2647 (1) There is a new type, "zns", which walks up the domain tree until it
2648 finds some nameserver records. It should be used with care.
2650 (2) There is a new type, "mxh", which is like "mx" except that it returns
2651 just the host names, not the priorities.
2653 (3) It is now possible to give a list of domains (or IP addresses) to be
2654 looked up. The behaviour when one of the lookups defers can be
2655 controlled by a keyword.
2657 (4) It is now possible to specify the separator character for use when
2658 multiple records are returned.
2660 38. The dnslists ACL condition has been extended: it is now possible to supply
2661 a list of IP addresses and/or domains to be looked up in a particular DNS
2664 39. Added log_selector=+queue_time_overall.
2666 40. When running the queue in the test harness, wait just a tad after forking a
2667 delivery process, to get repeatability of debugging output.
2669 41. Include certificate and key file names in error message when GnuTLS fails
2670 to set them up, because the GnuTLS error message doesn't include the name
2671 of the failing file when there is a problem reading it.
2673 42. Allow both -bf and -bF in the same test run.
2675 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2677 44. The "Exiscan patch" is now merged into the mainline Exim source.
2679 45. Sometimes the final signoff response after QUIT could fail to get
2680 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2681 < 0 before doing a fflush(). This bug looks as though it goes back to the
2682 introduction of TLS in release 3.20, but "sometimes" must have been rare
2683 because the tests only now provoked it.
2685 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2686 (this can affect the format of dates).
2688 47. exim_tidydb, when checking for the continued existence of a message for
2689 which it has found a message-specific retry record, was not finding
2690 messages that were in split spool directories. Consequently, it was
2691 deleting retry records that should have stayed in existence.
2693 48. Steve fixed some bugs in eximstats.
2695 49. The SPA authentication driver was not abandoning authentication and moving
2696 on to the next authenticator when an expansion was forced to fail,
2697 contradicting the general specification for all authenticators. Instead it
2698 was generating a temporary error. It now behaves as specified.
2700 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2701 (the order specifies the preference for clients). The order is now AES256,
2702 AES128, 3DES, ARCFOUR128.
2704 51. Small patch to Sieve code - explicitly set From: when generating an
2707 52. Exim crashed if a remote delivery caused a very long error message to be
2708 recorded - for instance if somebody sent an entire SpamAssassin report back
2709 as a large number of 550 error lines. This bug was coincidentally fixed by
2710 increasing the size of one of Exim's internal buffers (big_buffer) that
2711 happened as part of the Exiscan merge. However, to be on the safe side, I
2712 have made the code more robust (and fixed the comments that describe what
2715 53. Now that there can be additional text after "Completed" in log lines (if
2716 the queue_time_overall log selector is set), a one-byte patch to exigrep
2717 was needed to allow it to recognize "Completed" as not the last thing in
2720 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2721 patch that reportedly fixes this has been added. I am not expert enough to
2722 create a test for it. This is what the patch creator wrote:
2724 "I found a little strange behaviour of ldap code when working with
2725 Windows 2003 AD Domain, where users was placed in more than one
2726 Organization Units. When I tried to give exim partial DN, the exit code
2727 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2728 But simultaneously result of request was absolutely normal ldap result,
2729 so I produce this patch..."
2731 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2732 REFERENCE, so I have modified the code to exclude the patch when that macro
2735 55. Some experimental protocols are using DNS PTR records for new purposes. The
2736 keys for these records are domain names, not reversed IP addresses. The
2737 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2738 leaves it alone. Component reversal etc. now happens only for IP addresses.
2741 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2743 57. Double the size of the debug message buffer (to 2048) so that more of very
2744 long debug lines gets shown.
2746 58. The exicyclog utility now does better if the number of log files to keep
2747 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2749 59. Two changes related to the smtp_active_hostname option:
2751 (1) $smtp_active_hostname is now available as a variable.
2752 (2) The default for smtp_banner uses $smtp_active_hostname instead
2753 of $primary_hostname.
2755 60. The host_aton() function is supposed to be passed a string that is known
2756 to be a valid IP address. However, in the case of IPv6 addresses, it was
2757 not checking this. This is a hostage to fortune. Exim now panics and dies
2758 if the condition is not met. A case was found where this could be provoked
2759 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2760 components; fortuitously, this particular loophole had already been fixed
2761 by change 4.50/55 above.
2763 If there are any other similar loopholes, the new check in host_aton()
2764 itself should stop them being exploited. The report I received stated that
2765 data on the command line could provoke the exploit when Exim was running as
2766 exim, but did not say which command line option was involved. All I could
2767 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2768 running as the user.
2771 61. There was a buffer overflow vulnerability in the SPA authentication code
2772 (which came originally from the Samba project). I have added a test to the
2773 spa_base64_to_bits() function which I hope fixes it.
2776 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2777 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2779 63. The daemon start-up calls getloadavg() while still root for those OS that
2780 need the first call to be done as root, but it missed one case: when
2781 deliver_queue_load_max is set with deliver_drop_privilege. This is
2782 necessary for the benefit of the queue runner, because there is no re-exec
2783 when deliver_drop_privilege is set.
2785 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2786 This has been fixed.
2788 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2789 was in use, was not putting the data itself into the right store pool;
2790 consequently, it could be overwritten for a subsequent message in the same
2791 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2794 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2796 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2797 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2798 the function were treating the return as a boolean value, which happened to
2799 work because 0=false and not-0=true, but is not correct code.
2801 68. The host_aton() function was not handling scoped IPv6 addresses (those
2802 with, for example, "%eth0" on the end) correctly.
2804 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2805 negated items (that is, ~something) in unsigned ints. Some compilers
2806 apparently mutter when there is no cast.
2808 70. If an address verification called from an ACL failed, and did not produce a
2809 user-specific message (i.e. there was only a "system" message), nothing was
2810 put in $acl_verify_message. In this situation, it now puts the system
2813 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2814 session; change 4.31/43 added the unwanted input to the log line - except
2815 that it did not do this in the start of session case. It now does.
2817 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2818 This is wrong and can cause the other end to generate a synchronization
2819 error if it is another Exim or anything else that does the synchronization
2820 check. A QUIT command is no longer sent after a timeout.
2822 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2823 during host lookups.
2825 74. The defer_ok option of callout verification was not working if it was used
2826 when verifying addresses in header lines, that is, for this case:
2828 verify = header_sender/callout=defer_ok
2830 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2831 those file descriptors could be used for SMTP connections. If anything
2832 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2833 could be sent to the SMTP client, causing chaos. The daemon now opens
2834 stdin, stdout, and stderr to /dev/null when it puts itself into the
2837 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2838 log by default. The user can override this with suitable Perl magic.
2840 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2841 the log message when discard triggers, was not working for the DATA ACL or
2842 for the non-SMTP ACL.
2844 78. Error message wording change in sieve.c.
2846 79. If smtp_accept_max_per_host was set, the number of connections could be
2847 restricted to fewer than expected, because the daemon was trying to set up
2848 a new connection before checking whether the processes handling previous
2849 connections had finished. The check for completed processes is now done
2850 earlier. On busy systems, this bug wouldn't be noticed because something
2851 else would have woken the daemon, and it would have reaped the completed
2854 80. If a message was submitted locally by a user whose login name contained one
2855 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2856 It caused a spool format error. I have fixed the spool reading code. A
2857 related problem was that the "from" clause in the Received: line became
2858 illegal because of the space(s). It is now covered by ${quote_local_part.
2860 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2863 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2865 83. Patch from Sieve maintainer to fix unterminated string problem in
2866 "vacation" handling.
2868 84. Some minor changes to the Linux configuration files to help with other
2869 OS variants using glibc.
2871 85. One more patch for Sieve to update vacation handling to latest spec.
2874 ----------------------------------------------------
2875 See the note above about the 4.44 and 4.50 releases.
2876 ----------------------------------------------------
2882 1. Change 4.43/35 introduced a bug that caused file counts to be
2883 incorrectly computed when quota_filecount was set in an appendfile
2886 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2887 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2890 3. Give more explanation in the error message when the command for a transport
2891 filter fails to execute.
2893 4. There are several places where Exim runs a non-Exim command in a
2894 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2895 was being done only for the command run by the queryprogram router. It is
2896 now done for all such subprocesses. The other cases are: ${run, transport
2897 filters, and the commands run by the lmtp and pipe transports.
2899 5. Some older OS have a limit of 256 on the maximum number of file
2900 descriptors. Exim was using setrlimit() to set 1000 as a large value
2901 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2902 systems. I've change it so that if it can't get 1000, it tries for 256.
2904 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2905 was an oversight, and furthermore, ever since the addition of extra
2906 controls (e.g. 4.43/32), the checks on when to allow different forms of
2907 "control" were broken. There should now be diagnostics for all cases when a
2908 control that does not make sense is encountered.
2910 7. $recipients is now available in the predata ACL (oversight).
2912 8. Tidy the search cache before the fork to do a delivery from a message
2913 received from the command line. Otherwise the child will trigger a lookup
2914 failure and thereby defer the delivery if it tries to use (for example) a
2915 cached ldap connection that the parent has called unbind on.
2917 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2918 of $address_data from the recipient verification was clobbered by the
2919 sender verification.
2921 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2922 was its contents. (It was OK if the option was not defined at all.)
2924 11. A "Completed" log line is now written for messages that are removed from
2925 the spool by the -Mrm option.
2927 12. $host_address is now set to the target address during the checking of
2928 ignore_target_hosts.
2930 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2931 being passed; this would have caused $sender_host_name to have been used if
2932 matching the list had actually called for a host name (not very likely,
2933 since this list is usually IP addresses). A host name is now passed as
2936 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2937 code that specifies a non-restarting handler (typically sigaction() in
2938 modern systems) in an attempt to fix a rare and obscure crash bug.
2940 15. Narrowed the window for a race in the daemon that could cause it to ignore
2941 SIGCHLD signals. This is not a major problem, because they are used only to
2942 wake it up if nothing else does.
2944 16. A malformed maildirsize file could cause Exim to calculate negative values
2945 for the mailbox size or file count. Odd effects could occur as a result.
2946 The maildirsize information is now recalculated if the size or filecount
2949 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2950 support for a long time. Removed HAVE_SYS_VFS_H.
2952 18. Updated exipick to current release from John Jetmore.
2954 19. Allow an empty sender to be matched against a lookup in an address list.
2955 Previously the only cases considered were a regular expression, or an
2958 20. Exim went into a mad DNS lookup loop when doing a callout where the
2959 host was specified on the transport, if the DNS lookup yielded more than
2962 21. The RFC2047 encoding function was originally intended for short strings
2963 such as real names; it was not keeping to the 75-character limit for
2964 encoded words that the RFC imposes. It now respects the limit, and
2965 generates multiple encoded words if necessary. To be on the safe side, I
2966 have increased the buffer size for the ${rfc2047: expansion operator from
2969 22. Failure to deliver a bounce message always caused it to be frozen, even if
2970 there was an errors_to setting on the router. The errors_to setting is now
2973 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2974 canonical form (fully expanded) before being placed in
2975 $sender_host_address.
2977 24. Updated eximstats to version 1.33
2979 25. Include certificate and key file names in error message when GnuTLS fails
2980 to set them up, because the GnuTLS error message doesn't include the name
2981 of the failing file when there is a problem reading it.
2983 26. Expand error message when OpenSSL has problems setting up cert/key files.
2986 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2987 (this can affect the format of dates).
2989 28. exim_tidydb, when checking for the continued existence of a message for
2990 which it has found a message-specific retry record, was not finding
2991 messages that were in split spool directories. Consequently, it was
2992 deleting retry records that should have stayed in existence.
2994 29. eximstats updated to version 1.35
2995 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2996 1.35 - bugfix such that pie charts by volume are generated correctly
2998 30. The SPA authentication driver was not abandoning authentication and moving
2999 on to the next authenticator when an expansion was forced to fail,
3000 contradicting the general specification for all authenticators. Instead it
3001 was generating a temporary error. It now behaves as specified.
3003 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3004 (the order specifies the preference for clients). The order is now AES256,
3005 AES128, 3DES, ARCFOUR128.
3007 31. Small patch to Sieve code - explicitly set From: when generating an
3010 32. Exim crashed if a remote delivery caused a very long error message to be
3011 recorded - for instance if somebody sent an entire SpamAssassin report back
3012 as a large number of 550 error lines. This bug was coincidentally fixed by
3013 increasing the size of one of Exim's internal buffers (big_buffer) that
3014 happened as part of the Exiscan merge. However, to be on the safe side, I
3015 have made the code more robust (and fixed the comments that describe what
3018 33. Some experimental protocols are using DNS PTR records for new purposes. The
3019 keys for these records are domain names, not reversed IP addresses. The
3020 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3021 leaves it alone. Component reversal etc. now happens only for IP addresses.
3024 34. The host_aton() function is supposed to be passed a string that is known
3025 to be a valid IP address. However, in the case of IPv6 addresses, it was
3026 not checking this. This is a hostage to fortune. Exim now panics and dies
3027 if the condition is not met. A case was found where this could be provoked
3028 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3029 components; fortuitously, this particular loophole had already been fixed
3030 by change 4.50/55 or 4.44/33 above.
3032 If there are any other similar loopholes, the new check in host_aton()
3033 itself should stop them being exploited. The report I received stated that
3034 data on the command line could provoke the exploit when Exim was running as
3035 exim, but did not say which command line option was involved. All I could
3036 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3037 running as the user.
3040 35. There was a buffer overflow vulnerability in the SPA authentication code
3041 (which came originally from the Samba project). I have added a test to the
3042 spa_base64_to_bits() function which I hope fixes it.
3045 36. The daemon start-up calls getloadavg() while still root for those OS that
3046 need the first call to be done as root, but it missed one case: when
3047 deliver_queue_load_max is set with deliver_drop_privilege. This is
3048 necessary for the benefit of the queue runner, because there is no re-exec
3049 when deliver_drop_privilege is set.
3051 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3052 was in use, was not putting the data itself into the right store pool;
3053 consequently, it could be overwritten for a subsequent message in the same
3054 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3057 38. Sometimes the final signoff response after QUIT could fail to get
3058 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3059 < 0 before doing a fflush(). This bug looks as though it goes back to the
3060 introduction of TLS in release 3.20, but "sometimes" must have been rare
3061 because the tests only now provoked it.
3067 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3068 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3069 Now it returns an integer. A number of calls were still expecting a T/F
3070 return. Fortuitously, in all cases, the tests worked in OK situations,
3071 which is the norm. However, things would have gone wrong on any write
3072 failures on the smtp file descriptor. This function is used when sending
3073 messages over SMTP and also when doing verify callouts.
3075 2. When Exim is called to do synchronous delivery of a locally submitted
3076 message (the -odf or -odi options), it no longer closes stderr before doing
3079 3. Implemented the mua_wrapper option.
3081 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3083 5. Implemented the functions header_remove(), header_testname(),
3084 header_add_at_position(), and receive_remove_recipient(), and exported them
3087 6. If an ACL "warn" statement specified the addition of headers, Exim already
3088 inserted X-ACL-Warn: at the start if there was no header name. However, it
3089 was not making this test for the second and subsequent header lines if
3090 there were newlines in the string. This meant that an invalid header could
3091 be inserted if Exim was badly configured.
3093 7. Allow an ACL "warn" statement to add header lines at the start or after all
3094 the Received: headers, as well as at the end.
3096 8. Added the rcpt_4xx retry error code.
3098 9. Added postmaster_mailfrom=xxx to callout verification option.
3100 10. Added mailfrom=xxxx to the callout verification option, for verify=
3103 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3104 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3106 12. Inserted some casts to stop certain compilers warning when using pointer
3107 differences as field lengths or precisions in printf-type calls (mostly
3108 affecting debugging statements).
3110 13. Added optional readline() support for -be (dynamically loaded).
3112 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3113 same clock tick as a message's arrival, so that its received time was the
3114 same as the "first fail" time on the retry record, and that message
3115 remained on the queue past the ultimate address timeout, every queue runner
3116 would try a delivery (because it was past the ultimate address timeout) but
3117 after another failure, the ultimate address timeout, which should have then
3118 bounced the address, did not kick in. This was a "< instead of <=" error;
3119 in most cases the first failure would have been in the next clock tick
3120 after the received time, and all would be well.
3122 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3123 being recognized when the domain list was tested by the match_domain
3124 condition in an expansion string.
3126 16. Added the ${str2b64: operator.
3128 17. Exim was always calling setrlimit() to set a large limit for the number of
3129 processes, without checking whether the existing limit was already
3130 adequate. (It did check for the limit on file descriptors.) Furthermore,
3131 errors from getrlimit() and setrlimit() were being ignored. Now they are
3132 logged to the main and panic logs, but Exim does carry on, to try to do its
3133 job under whatever limits there are.
3135 18. Imported PCRE 5.0.
3137 19. Trivial typo in log message " temporarily refused connection" (the leading
3140 20. If the log selector return_path_on_delivery was set and an address was
3141 redirected to /dev/null, the delivery process crashed because it assumed
3142 that a return path would always be set for a "successful" delivery. In this
3143 case, the whole delivery is bypassed as an optimization, and therefore no
3146 21. Internal re-arrangement: the function for sending a challenge and reading
3147 a response while authentication was assuming a zero-terminated challenge
3148 string. It's now changed to take a pointer and a length, to allow for
3149 binary data in such strings.
3151 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3153 23. Exim was not respecting finduser_retries when seeking the login of the
3154 uid under which it was called; it was always trying 10 times. (The default
3155 setting of finduser_retries is zero.) Also, it was sleeping after the final
3156 failure, which is pointless.
3158 24. Implemented tls_on_connect_ports.
3160 25. Implemented acl_smtp_predata.
3162 26. If the domain in control=submission is set empty, Exim assumes that the
3163 authenticated id is a complete email address when it generates From: or
3164 Sender: header lines.
3166 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3167 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3168 chown and chgrp in /bin and hostname in /usr/bin.
3170 28. Exim was keeping the "process log" file open after each use, just as it
3171 does for the main log. This opens the possibility of it remaining open for
3172 long periods when the USR1 signal hits a daemon. Occasional processlog
3173 errors were reported, that could have been caused by this. Anyway, it seems
3174 much more sensible not to leave this file open at all, so that is what now
3177 29. The long-running daemon process does not normally write to the log once it
3178 has entered its main loop, and it closes the log before doing so. This is
3179 so that log files can straightforwardly be renamed and moved. However,
3180 there are a couple of unusual error situations where the daemon does write
3181 log entries, and I had neglected to close the log afterwards.
3183 30. The text of an SMTP error response that was received during a remote
3184 delivery was being truncated at 512 bytes. This is too short for some of
3185 the long messages that one sometimes sees. I've increased the limit to
3188 31. It is now possible to make retry rules that apply only when a message has a
3189 specific sender, in particular, an empty sender.
3191 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3192 it possible to be selective about when SMTP synchronization is enforced.
3194 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3196 32. Implemented hosts_connection_nolog.
3198 33. Added an ACL for QUIT.
3200 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3203 35. Added mailbox_size and mailbox_filecount to appendfile.
3205 36. Added control = no_multiline_responses to ACLs.
3207 37. There was a bug in the logic of the code that waits for the clock to tick
3208 in the case where the clock went backwards by a substantial amount such
3209 that the microsecond fraction of "now" was more than the microsecond
3210 fraction of "then" (but the whole seconds number was less).
3212 38. Added support for the libradius Radius client library this is found on
3213 FreeBSD (previously only the radiusclient library was supported).
3219 1. When certain lookups returned multiple values in the form name=value, the
3220 quoting of the values was not always being done properly. Specifically:
3221 (a) If the value started with a double quote, but contained no whitespace,
3223 (b) If the value contained whitespace other than a space character (i.e.
3224 tabs or newlines or carriage returns) it was not quoted.
3225 This fix has been applied to the mysql and pgsql lookups by writing a
3226 separate quoting function and calling it from the lookup code. The fix
3227 should probably also be applied to nisplus, ibase and oracle lookups, but
3228 since I cannot test any of those, I have not disturbed their existing code.
3230 2. A hit in the callout cache for a specific address caused a log line with no
3231 reason for rejecting RCPT. Now it says "Previous (cached) callout
3232 verification failure".
3234 3. There was an off-by-one bug in the queryprogram router. An over-long
3235 return line was truncated at 256 instead of 255 characters, thereby
3236 overflowing its buffer with the terminating zero. As well as fixing this, I
3237 have increased the buffer size to 1024 (and made a note to document this).
3239 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3240 when Exim is waiting for an SMTP response from a remote server, Exim
3241 restarts its select() call on the socket, thereby resetting its timeout.
3242 This is not a problem when such interrupts are rare. Somebody set up a cron
3243 job to run exiwhat every 2 minutes, which is less than the normal select()
3244 timeout (5 or 10 minutes). This meant that the select() timeout never
3245 kicked in because it was always reset. I have fixed this by comparing the
3246 time when an interrupt arrives with the time at the start of the first call
3247 to select(). If more time than the timeout has elapsed, the interrupt is
3248 treated as a timeout.
3250 5. Some internal re-factoring in preparation for the addition of Sieve
3251 extensions (by MH). In particular, the "personal" test is moved to a
3252 separate function, and given an option for scanning Cc: and Bcc: (which is
3253 not set for Exim filters).
3255 6. When Exim created an email address using the login of the caller as the
3256 local part (e.g. when creating a From: or Sender: header line), it was not
3257 quoting the local part when it contained special characters such as @.
3259 7. Installed new OpenBSD configuration files.
3261 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3262 try to make them clearer.
3264 9. Callout options, other than the timeout value, were being ignored when
3265 verifying sender addresses in header lines. For example, when using
3267 verify = header_sender/callout=no_cache
3269 the cache was (incorrectly) being used.
3271 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3272 only the Cygwin environment.
3274 11. When return_path_on_delivery was set as a log selector, if different remote
3275 addresses in the same message used different return paths and parallel
3276 remote delivery occurred, the wrong values would sometimes be logged.
3277 (Whenever a remote delivery process finished, the return path value from
3278 the most recently started remote delivery process was logged.)
3280 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3281 lines when AUTH and/or TLS are in use. This is the "received protocol"
3282 field. Exim used to use "asmtp" for authenticated SMTP, without any
3283 indication (in the protocol name) for TLS use. Now it follows the RFC and
3284 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3285 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3286 names appear in log lines as well as in Received: header lines.
3288 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3289 extensions, and comparison tests, and to fix some bugs.
3291 14. Changes to the "personal" filter test:
3293 (1) The test was buggy in that it was just doing the equivalent of
3294 "contains" tests on header lines. For example, if a user's address was
3295 anne@some.where, the "personal" test would incorrectly be true for
3297 To: susanne@some.where
3299 This test is now done by extracting each address from the header in turn,
3300 and checking the entire address. Other tests that are part of "personal"
3301 are now done using regular expressions (for example, to check local parts
3302 of addresses in From: header lines).
3304 (2) The list of non-personal local parts in From: addresses has been
3305 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3306 taken from the Sieve specification recommendations.
3308 (3) If the message contains any header line starting with "List-" it is
3309 treated as non-personal.
3311 (4) The test for "circular" in the Subject: header line has been removed
3312 because it now seems ill-conceived.
3314 15. Minor typos in src/EDITME comments corrected.
3316 16. Installed latest exipick from John Jetmore.
3318 17. If headers_add on a router specified a text string that was too long for
3319 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3320 of string_sprintf() is now avoided.
3322 18. $message_body_size was not set (it was always zero) when running the DATA
3323 ACL and the local_scan() function.
3325 19. For the "mail" command in an Exim filter, no default was being set for
3326 the once_repeat time, causing a random time value to be used if "once" was
3327 specified. (If the value happened to be <= 0, no repeat happened.) The
3328 default is now 0s, meaning "never repeat". The "vacation" command was OK
3329 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3330 (I found it when inspecting the code).
3332 20. There is now an overall timeout for performing a callout verification. It
3333 defaults to 4 times the callout timeout, which applies to individual SMTP
3334 commands during the callout. The overall timeout applies when there is more
3335 than one host that can be tried. The timeout is checked before trying the
3336 next host. This prevents very long delays if there are a large number of
3337 hosts and all are timing out (e.g. when the network connections are timing
3338 out). The value of the overall timeout can be changed by specifying an
3339 additional sub-option for "callout", called "maxwait". For example:
3341 verify = sender/callout=5s,maxwait=20s
3343 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3344 to the end before writing, but this should make it even safer).
3346 22. Exim was forgetting that it had advertised PIPELINING for the second and
3347 subsequent messages on an SMTP connection. It was also not resetting its
3348 memory on STARTTLS and an internal HELO.
3350 23. When Exim logs an SMTP synchronization error within a session, it now
3351 records whether PIPELINING has been advertised or not.
3353 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3354 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3355 rather than long int.
3357 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3359 26. Added the never_mail option to autoreply.
3365 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3366 crash if the getsockname() call failed; this can happen if a connection is
3367 closed very soon after it is established. The problem was simply in the
3368 order in which certain operations were done, causing Exim to try to write
3369 to the SMTP stream before it had set up the file descriptor. The bug has
3370 been fixed by making things happen in the correct order.
3376 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3377 before the connection was closed, thus losing the rejection response.
3379 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3380 some early Solaris releases, but causes trouble in current releases where
3381 socklen_t is defined.
3383 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3386 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3389 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3390 that releasing the top of it at the end releases what was used for sub-
3391 expansions (unless the block got too big). However, discard this block if
3392 the first thing is a variable or header, so that we can use its block when
3393 it is dynamic (useful for very large $message_headers, for example).
3395 6. Lookups now cache *every* query, not just the most recent. A new, separate
3396 store pool is used for this. It can be recovered when all lookup caches are
3397 flushed. Lookups now release memory at the end of their result strings.
3398 This has involved some general refactoring of the lookup sources.
3400 7. Some code has been added to the store_xxx() functions to reduce the amount
3401 of flapping under certain conditions.
3403 8. log_incoming_interface used to affect only the <= reception log lines. Now
3404 it causes the local interface and port to be added to several more SMTP log
3405 lines, for example "SMTP connection from", and rejection lines.
3407 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3409 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3411 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3412 could be overwritten at the end of the current message (or the start of a
3413 new message if it was set in a HELO ACL). The value is now preserved for
3414 the duration of the SMTP connection.
3416 12. If a transport had a headers_rewrite setting, and a matching header line
3417 contained an unqualified address, that address was qualified, even if it
3418 did not match any rewriting rules. The underlying bug was that the values
3419 of the flags that permit the existence of unqualified sender and recipient
3420 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3421 non-local messages, and by -bnq for local messages) were not being
3422 preserved with the message after it was received.
3424 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3425 "next input=" as part of the text comprising the host identity instead of
3426 the correct text. The code was using the same buffer for two different
3427 strings. However, depending on which order the printing function evaluated
3428 its arguments, the bug did not always show up. Under Linux, for example, my
3429 test suite worked just fine.
3431 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3432 doesn't work with some older versions of Perl. It has been changed to "my",
3433 which in any case is probably the better facility to use.
3435 15. A really picky compiler found some instances of statements for creating
3436 error messages that either had too many or two few arguments for the format
3439 16. The size of the buffer for calls to the DNS resolver has been increased
3440 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3441 for addresses that have a lot of PTR records. This alleviates a problem; it
3442 does not fully solve it.
3444 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3445 buffer now truncates the list and logs the incident, which is the same
3446 action as happens when Exim is looking up a host name and its aliases.
3447 Previously in this situation something unpredictable would happen;
3448 sometimes it was "internal error: store_reset failed".
3450 18. If a server dropped the connection unexpectedly when an Exim client was
3451 using GnuTLS and trying to read a response, the client delivery process
3452 crashed while trying to generate an error log message.
3454 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3455 string, for example:
3457 warn message = H1: something\nH2: something
3459 the text was added as a single header line from Exim's point of view
3460 though it ended up OK in the delivered message. However, searching for the
3461 second and subsequent header lines using $h_h2: did not work. This has been
3462 fixed. Similarly, if a system filter added multiple headers in this way,
3463 the routers could not see them.
3465 20. Expanded the error message when iplsearch is called with an invalid key to
3466 suggest using net-iplsearch in a host list.
3468 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3469 ACLs are no longer actually imposed (and a message to that effect is
3472 22. If a "gecos" field in a passwd entry contained escaped characters, in
3473 particular, if it contained a \" sequence, Exim got it wrong when building
3474 a From: or a Sender: header from that name. A second bug also caused
3475 incorrect handling when an unquoted " was present following a character
3476 that needed quoting.
3478 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3479 was not being matched caselessly.
3481 24. Arranged for all hyphens in the exim.8 source to be escaped with
3484 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3485 properly. Recipient callout cache records were still being keyed to include
3486 the sender, even when use_sender was set false. This led to far more
3487 callouts that were necessary. The sender is no longer included in the key
3488 when use_sender is false.
3490 26. Added "control = submission" modifier to ACLs.
3492 27. Added the ${base62d: operator to decode base 62 numbers.
3494 28. dnsdb lookups can now access SRV records.
3496 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3497 the configuration file.
3499 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3500 (-v) mode. This makes the output for a verbose queue run more intelligible.
3502 31. Added a use_postmaster feature to recipient callouts.
3504 32. Added the $body_zerocount variable, containing the number of binary zero
3505 bytes in the message body.
3507 33. The time of last modification of the "new" subdirectory is now used as the
3508 "mailbox time last read" when there is a quota error for a maildir
3511 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3513 35. Added +ignore_unknown as a special item in host lists.
3515 36. Code for decoding IPv6 addresses in host lists is now included, even if
3516 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3517 address was recognized as an IP address, but was then not correctly decoded
3518 into binary, causing unexpected and incorrect effects when compared with
3525 1. Very minor rewording of debugging text in manualroute to say "list of
3526 hosts" instead of "hostlist".
3528 2. If verify=header_syntax was set, and a header line with an unqualified
3529 address (no domain) and a large number of spaces between the end of the
3530 name and the colon was received, the reception process suffered a buffer
3531 overflow, and (when I tested it) crashed. This was caused by some obsolete
3532 code that should have been removed. The fix is to remove it!
3534 3. When running in the test harness, delay a bit after writing a bounce
3535 message to get a bit more predictability in the log output.
3537 4. Added a call to search_tidyup() just before forking a reception process. In
3538 theory, someone could use a lookup in the expansion of smtp_accept_max_
3539 per_host which, without the tidyup, could leave open a database connection.
3541 5. Added the variables $recipient_data and $sender_data which get set from a
3542 lookup success in an ACL "recipients" or "senders" condition, or a router
3543 "senders" option, similar to $domain_data and $local_part_data.
3545 6. Moved the writing of debug_print from before to after the "senders" test
3548 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3549 problems for message scanning, either using a data ACL, or using
3550 local_scan() because the Received: header was not generated till after they
3551 were called (in order to set the time as the time of reception completion).
3552 I have revised the way this works. The header is now generated after the
3553 body is received, but before the ACL or local_scan() are called. After they
3554 are run, the timestamp in the header is updated.
3560 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3561 before starting a queue runner without re-exec. This happened only when
3562 deliver_drop_privilege was set or when the Exim user was set to root. The
3563 effect of the bug was that timeouts during subsequent deliveries caused
3564 crashes instead of being properly handled. The handler is now left at its
3565 default (and expected) setting.
3567 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3568 message, again when deliver_drop_privilege is set or Exim is run as root.
3569 The bug described in (1) was not present in this case, but the tidying up
3570 of the other signals was missing. I have made the two cases consistent.
3572 3. The ignore_target_hosts setting on a manualroute router was being ignored
3573 for hosts that were looked up using the /MX notation.
3575 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3578 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3579 operated on the sender address. After changing the $sender_address to <>
3580 for the sender address verify, Exim was re-instated it as the original
3581 (before rewriting) address, but remembering that it had rewritten it, so it
3582 wasn't rewriting it again. This bug also had the effect of breaking the
3583 sender address verification caching when the sender address was rewritten.
3585 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3586 This has been changed so that if the ip literal address matches
3587 ignore_target_hosts, the router declines.
3589 7. Added expansion conditions match_domain, match_address, and match_local_
3590 part (NOT match_host).
3592 8. The placeholder for the Received: header didn't have a length field set.
3594 9. Added code to Exim itself and to exim_lock to test for a specific race
3595 condition that could lead to file corruption when using MBX delivery. The
3596 issue is with the lockfile that is created in /tmp. If this file is removed
3597 after a process has opened it but before that process has acquired a lock,
3598 there is the potential for a second process to recreate the file and also
3599 acquire a lock. This could lead to two Exim processes writing to the file
3600 at the same time. The added code performs the same test as UW imapd; it
3601 checks after acquiring the lock that its file descriptor still refers to
3602 the same named file.
3604 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3605 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3608 11. Added the smtp_active_hostname option. If used, this will typically be made
3609 to depend on the incoming interface address. Because $interface_address is
3610 not set up until the daemon has forked a reception process, error responses
3611 that can happen earlier (such as "too many connections") no longer contain
3614 12. If an expansion in a condition on a "warn" statement fails because a lookup
3615 defers, the "warn" statement is abandoned, and the next ACL statement is
3616 processed. Previously this caused the whole ACL to be aborted.
3618 13. Added the iplsearch lookup type.
3620 14. Added ident_timeout as a log selector.
3622 15. Added tls_certificate_verified as a log selector.
3624 16. Added a global option tls_require_ciphers (compare the smtp transport
3625 option of the same name). This controls incoming TLS connections.
3627 17. I finally figured out how to make tls_require_ciphers do a similar thing
3628 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3629 before starting the TLS session.
3631 18. Tabs are now shown as \t in -bP output.
3633 19. If the log selector return_path_on_delivery was set, Exim crashed when
3634 bouncing a message because it had too many Received: header lines.
3636 20. If two routers both had headers_remove settings, and the first one included
3637 a superfluous trailing colon, the final name in the first list and the
3638 first name in the second list were incorrectly joined into one item (with a
3639 colon in the middle).
3645 1. Added -C and -D options to the exinext utility, mainly to make it easier
3646 to include in the automated testing, but these could be helpful when
3647 multiple configurations are in use.
3649 2. The exinext utility was not formatting the output nicely when there was
3650 an alternate port involved in the retry record key, nor when there was a
3651 message id as well (for retries that were specific to a specific message
3652 and a specific host). It was also confused by IPv6 addresses, because of
3653 the additional colons they contain. I have fixed the IPv4 problem, and
3654 patched it up to do a reasonable job for IPv6.
3656 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3657 delivery, the log line now contains "pipelined" if PIPELINING was used.
3659 4. An SMTP transport process used to panic and die if the bind() call to set
3660 an explicit outgoing interface failed. This has been changed; it is now
3661 treated in the same way as a connect() failure.
3663 5. A reference to $sender_host_name in the part of a conditional expansion
3664 that was being skipped was still causing a DNS lookup. This no longer
3667 6. The def: expansion condition was not recognizing references to header lines
3668 that used bh_ and bheader_.
3670 7. Added the _cache feature to named lists.
3672 8. The code for checking quota_filecount in the appendfile transport was
3673 allowing one more file than it should have been.
3675 9. For compatibility with Sendmail, the command line option
3683 and sets the incoming protocol and host name (for trusted callers). The
3684 host name and its colon can be omitted when only the protocol is to be set.
3685 Note the Exim already has two private options, -pd and -ps, that refer to
3686 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3687 "s", but I don't think that's a major issue.
3689 10. A number of refactoring changes to the code, none of which should affect
3692 (a) The number of logging options was getting close to filling up the
3693 32-bit word that was used as a bit map. I have split them into two classes:
3694 those that are passed in the argument to log_write(), and those that are
3695 only ever tested independently outside of that function. These are now in
3696 separate 32-bit words, so there is plenty of room for expansion again.
3697 There is no change in the user interface or the logging behaviour.
3699 (b) When building, for example, log lines, the code previously used a
3700 macro that called string_cat() twice, in order to add two strings. This is
3701 not really sufficiently general. Furthermore, there was one instance where
3702 it was actually wrong because one of the argument was used twice, and in
3703 one call a function was used. (As it happened, calling the function twice
3704 did not affect the overall behaviour.) The macro has been replaced by a
3705 function that can join an arbitrary number of extra strings onto a growing
3708 (c) The code for expansion conditions now uses a table and a binary chop
3709 instead of a serial search (which was left over from when there were very
3710 few conditions). Also, it now recognizes conditions like "pam" even when
3711 the relevant support is not compiled in: a suitably worded error message is
3712 given if an attempt is made to use such a condition.
3714 11. Added ${time_interval:xxxxx}.
3716 12. A bug was causing one of the ddress fields not to be passed back correctly
3717 from remote delivery subprocesses. The field in question was not being
3718 subsequently used, so this caused to problems in practice.
3720 13. Added new log selectors queue_time and deliver_time.
3722 14. Might have fixed a bug in maildirsizefile handling that threw up
3723 "unexpected character" debug warnings, and recalculated the data
3724 unnecessarily. In any case, I expanded the warning message to give more
3727 15. Added the message "Restricted characters in address" to the statements in
3728 the default ACL that block characters like @ and % in local parts.
3730 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3731 Three changes have been made:
3733 (a) There was a serious bug; a negative response to MAIL caused the whole
3734 recipient domain to be cached as invalid, thereby blocking all messages
3735 to all local parts at the same domain, from all senders. This bug has
3736 been fixed. The domain is no longer cached after a negative response to
3737 MAIL if the sender used is not empty.
3739 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3742 (c) A new callout option, "use_sender" has been added for people who want
3743 the modified behaviour.
3749 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3752 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3753 indeed breaks things for older releases.
3755 3. Added additional logging to the case where there is a problem reading data
3756 from a filter that is running in a subprocess using a pipe, in order to
3757 try to track down a specific problem.
3759 4. Testing facility fudge: when running in the test harness and attempting
3760 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3761 getting "No route to host". Convert this to a timeout.
3763 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3766 6. Some OS don't have socklen_t but use size_t instead. This affects the
3767 fifth argument of getsockopt() amongst other things. This is now
3768 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3769 can be set for individual OS. I have set it for SunOS5, OSF1, and
3770 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3771 some earlier ones do not.
3773 7. Change 4.30/15 was not doing the test caselessly.
3775 8. The standard form for an IPv6 address literal was being rejected by address
3776 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3777 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3778 this, as well as the form without the "IPv6" on the front (but only when
3779 address literals are enabled, of course).
3781 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3783 10. Exim crashed if a message with an empty sender address specified by -f
3784 encountered a router with an errors_to setting. This could be provoked only
3785 by a command such as
3789 where an empty string was supplied; "<>" did not hit this bug.
3791 11. Installed PCRE release 4.5.
3793 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3794 remained set. It is now erased.
3796 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3797 times from message ids (which are base 36 rather than the normal 62).
3799 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3800 were being counted as actual protocol errors, and logged if the log
3801 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3802 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3803 and DATA following a set of rejected RCPTs do not count as protocol errors.
3804 In other words, Exim assumes they were pipelined, though this may not
3805 actually be the case. Of course, in all cases the client gets an
3806 appropriate error code.
3808 15. If a lookup fails in an ACL condition, a message about the failure may
3809 be available; it is used if testing the ACL cannot continue, because most
3810 such messages specify what the cause of the deferral is. However, some
3811 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3812 that caused an old message to be retained and used if a later statement
3813 caused a defer, replacing the real cause of the deferral.
3815 16. If an IP address had so many PTR records that the DNS lookup buffer
3816 was not large enough to hold them, Exim could crash while trying to process
3817 the truncated data. It now detects and logs this case.
3819 17. Further to 4.21/58, another change has been made: if (and only if) the
3820 first line of a message (the first header line) ends with CRLF, a bare LF
3821 in a subsequent header line has a space inserted after it, so as not to
3822 terminate the header.
3824 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3825 unnecessarily, used atoi() instead of strtol(), and didn't check the
3826 termination when getting file sizes from file names by regex.
3828 19. Completely re-implemented the support for maildirsize files, in the light
3829 of a number of problems with the previous contributed implementation
3830 (4.30/29). In particular:
3832 . If the quota is zero, the maildirsize file is maintained, but no quota is
3835 . If the maildir directory does not exist, it is created before any attempt
3836 to write a maildirsize file.
3838 . The quota value in the file is just a cache; if the quota is changed in
3839 the transport, the new value overrides.
3841 . A regular expression is available for excluding directories from the
3844 20. The autoreply transport checks the characters in options that define the
3845 message's headers; it allows continued headers, but it was checking with
3846 isspace() after an embedded newline instead of explicitly looking for a
3849 21. If all the "regular" hosts to which an address was routed had passed their
3850 expiry times, and had not reached their retry times, the address was
3851 bounced, even if fallback hosts were defined. Now Exim should go on to try
3854 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3855 equivalent code in the SMTP transport. Some hosts send humungous responses
3856 to HELO/EHLO, more than 1024 it seems.
3858 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3859 gives compiler warnings in some environments. I've now done it "properly",
3862 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3863 (because of problems with the built-in one) was declared to return uschar *
3864 instead of char *, causing compiler failure.
3866 25. Fixed a file descriptor leak when processing alias/forward files.
3868 26. Fixed a minor format string issue in dbfn.c.
3870 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3872 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3873 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3876 29. When a sender address is verified, it is cached, to save repeating the test
3877 when there is more than one recipient in a message. However, when the
3878 verification involves a callout, it is possible for different callout
3879 options to be set for different recipients. It is too complicated to keep
3880 track of this in the cache, so now Exim always runs a verification when a
3881 callout is required, relying on the callout cache for the optimization.
3882 The overhead is duplication of the address routing, but this should not be
3885 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3886 to be verified with callout=postmaster, and the main callout worked but the
3887 postmaster check failed, the verification correctly failed. However, if a
3888 subsequent RCPT command asked for sender verification *without* the
3889 postmaster check, incorrect caching caused this verification also to fail,
3892 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3893 it was not caching the DNS options (qualify_single, search_parents) that
3894 were used when the lookup failed. A subsequent lookup with different
3895 options therefore always gave the same answer, though there were cases
3896 where it should not have. (Example: a "domains = !$mx_any" option on a
3897 dnslookup router: the "domains" option is always processed without any
3898 widening, but the router might have qualify_single set.) Now Exim uses the
3899 cached value only when the same options are set.
3901 32. Added John Jetmore's "exipick" utility to the distribution.
3903 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3904 than a timeout (e.g. a certificate is required, and is not provided), an
3905 Exim server now closes the connection immediately. Previously it waited for
3906 the client to close - but if the client is SSL, it seems that they each
3907 wait for each other, leading to a delay before one of them times out.
3909 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3910 maintained 0.8.x compatibility because I don't think many are using it, and
3911 it is clearly obsolete.
3913 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3916 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3917 client certificate was expired. A simple patch fixes this, though I don't
3918 understand the full logic of why the verify callback is called multiple
3921 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3922 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3923 which causes problems with some clients (such as the Certicom SSL Plus
3924 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3925 disables the coutermeasure allowing Eudora to connect."
3927 38. Exim was not checking that a write() to a log file succeeded. This could
3928 lead to Bad Things if a log got too big, in particular if it hit a file
3929 size limit. Exim now panics and dies if it cannot write to a log file, just
3930 as it does if it cannot open a log file.
3932 39. Modified OS/Makefile-Linux so that it now contains
3934 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3936 The two -D definitions ensure that Exim is compiled with large file
3937 support, which makes it possible to handle log files that are bigger than
3940 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3941 instance) a domain was checked against a named list that involved a lookup,
3942 causing $domain_data to be set, then another domain was checked against the
3943 same list, then the first domain was re-checked, the value of $domain_data
3944 after the final check could be wrong. In particular, if the second check
3945 failed, it could be set empty. This bug probably also applied to
3948 41. The strip_trailing_dot option was not being applied to the address given
3949 with the -f command-line option.
3951 42. The code for reading a message's header from the spool was incrementing
3952 $received_count, but never initializing it. This meant that the value was
3953 incorrect (doubled) while delivering a message in the same process in which
3954 it was received. In the most common configuration of Exim, this never
3955 happens - a fresh exec is done - but it can happen when
3956 deliver_drop_privilege is set.
3958 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3959 it now includes up to 150 characters of the unexpected data in the log
3962 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3963 and building data strings. The size of both of these buffers was 10 000
3964 bytes - far larger than anybody would *ever* want, thought I. Needless to
3965 say, somebody hit the limit. I have increased the maximum line length to
3966 20 000 and the maximum data length of concatenated lines to 100 000. I have
3967 also fixed two bugs, because there was no checking on these buffers. Tsk,
3968 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3969 buffer is too small.
3971 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3972 lsearch lookups. Now it does.
3974 46. When parsing a route_list item in a manualroute router, a fixed-length
3975 buffer was used for the list of hosts. I made this 1024 bytes long,
3976 thinking that nobody would ever have a list of hosts that long. Wrong.
3977 Somebody had a whole pile of complicated expansion conditions, and the
3978 string was silently truncated, leading to an expansion error. It turns out
3979 that it is easier to change to an unlimited length (owing to other changes
3980 that have happened since this code was originally written) than to build
3981 structure for giving a limitation error. The length of the item that
3982 expands into the list of hosts is now unlimited.
3984 47. The lsearch lookup could not handle data where the length of text line was
3985 more than 4095 characters. Such lines were truncated, leading to shortened
3986 data being returned. It should now handle lines of any length.
3988 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3989 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3992 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3993 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3994 no longer happens in comment lines. A list of replacements is now placed
3995 at the head of all of the source files, except those whose only change is
3996 to replace PERL_COMMAND in the very first #! line.
3998 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3999 messages on the queue, with a bottom-up merge sort, using code contributed
4000 by Michael Haardt. This should make operations like -bp somewhat faster on
4001 large queues. It won't affect queue runners, except when queue_run_in_order
4004 51. Installed eximstats 1.31 in the distribution.
4006 52. Added support for SRV lookups to the dnslookup router.
4008 53. If an ACL referred to $message_body or $message_body_end, the value was not
4009 reset for any messages that followed in the same SMTP session.
4011 54. The store-handling optimization for building very long strings was not
4012 differentiating between the different store pools. I don't think this
4013 actually made any difference in practice, but I've tidied it.
4015 55. While running the routers to verify a sender address, $sender_address
4016 was still set to the sender address. This is wrong, because when routing to
4017 send a bounce to the sender, it would be empty. Therefore, I have changed
4018 it so that, while verifying a sender address, $sender_address is set to <>.
4019 (There is no change to what happens when verifying a recipient address.)
4021 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4022 target A or AAAA records (if not already returned) without resetting the
4023 qualify_single or search_parents options of the DNS resolver. These are
4024 inappropriate in this case because the targets of MX and SRV records must
4025 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4026 target that, when qualified, matched something in the local domain. These
4027 two options are now turned off when doing these lookups.
4029 57. It seems that at least some releases of Reiserfs (which does not have the
4030 concept of a fixed number of inodes) returns zero and not -1 for the
4031 number of available inodes. This interacted badly with check_spool_inodes,
4032 which assumed that -1 was the "no such thing" setting. What I have done is
4033 to check that the total number of inodes is greater than zero before doing
4034 the test of how many are available.
4036 58. When a "warn" ACL statement has a log_message modifier, the message is
4037 remembered, and not repeated. This is to avoid a lot of repetition when a
4038 message has many recipients that cause the same warning to be written.
4039 Howewer, Exim was preserving the list of already written lines for an
4040 entire SMTP session, which doesn't seem right. The memory is now reset if a
4041 new message is started.
4043 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4044 debugging output unless log_rewrite was also set.
4046 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4047 of (int)(handle) when we know that handle contains (void *)(-1).
4049 61. The Exim daemon panic-logs an error return when it closes the incoming
4050 connection. However "connection reset by peer" seems to be common, and
4051 isn't really an error worthy of noting specially, so that particular error
4054 62. When Exim is trying to find all the local interfaces, it used to panic and
4055 die if the ioctl to get the interface flags failed. However, it seems that
4056 on at least one OS (Solaris 9) it is possible to have an interface that is
4057 included in the list of interfaces, but for which you get a failure error
4058 for this call. This happens when the interface is not "plumbed" into a
4059 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4060 failure of the "get flags" call assumes that the interface is down.
4062 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4063 makes life easier for people who are doing arithmetic on fields extracted
4064 from dates, where you often get leading zeros that should not be
4065 interpreted as octal.
4067 64. Added qualify_domain to the redirect router, to override the global
4070 65. If a pathologically long header line contained very many addresses (the
4071 report of this problem mentioned 10 000) and each of them was rewritten,
4072 Exim could use up a very large amount of memory. (It kept on making new
4073 copies of the header line as it rewrote, and never released the old ones.)
4074 At the expense of a bit more processing, the header rewriting function has
4075 been changed so that it no longer eats memory in this way.
4077 66. The generation of the Received: header has been moved from the time that a
4078 message starts to be received, to the time that it finishes. The timestamp
4079 in the Received: header should now be very close to that of the <= log
4080 line. There are two side-effects of this change:
4082 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4083 logged header lines no longer include the local Received: line, because
4084 it has not yet been created. The same applies to a copy of the message
4085 that is returned to a non-SMTP sender when a message is rejected.
4087 (b) When a filter file is tested using -bf, no additional Received: header
4088 is added to the test message. After some thought, I decided that this
4091 This change does not affect the value of $received_for. It is still set
4092 after address rewriting, but before local_scan() is called.
4094 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4096 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4097 gave an unhelpful panic error message, and a defer error. I have managed to
4098 change this behaviour so that it now rejects any supplied certificate,
4099 which seems right, as the list of acceptable certificates is empty.
4101 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4102 gave an unhelpful defer error. I have not managed to make this reject any
4103 supplied certificates, but the error message it gives is "no certificate
4104 supplied", which is not helpful.
4106 70. exigrep's output now also includes lines that are not associated with any
4107 message, but which match the given pattern. Implemented by a patch from
4108 Martin Sluka, which also tidied up the Perl a bit.
4110 71. Recipient callout verification, like sender verification, was using <> in
4111 the MAIL FROM command. This isn't really the right thing, since the actual
4112 sender may affect whether the remote host accepts the recipient or not. I
4113 have changed it to use the actual sender in the callout; this means that
4114 the cache record is now keyed on a recipient/sender pair, not just the
4115 recipient address. There doesn't seem to be a real danger of callout loops,
4116 since a callout by the remote host to check the sender would use <>.
4117 [SEE ABOVE: changed after hitting problems.]
4119 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4120 temporary errors. However, in the case of such a code being given after
4121 the end of a data transmission (i.e. after ".") Exim was failing to write
4122 a retry record for the message. (Yes, there was some broken host that was
4123 actually sending 8xx at this point.)
4125 73. An unknown lookup type in a host list could cause Exim to panic-die when
4126 the list was checked. (An example that provoked this was putting <; in the
4127 middle of a list instead of at the start.) If this happened during a DATA
4128 ACL check, a -D file could be left lying around. This kind of configuration
4129 error no longer causes Exim to die; instead it causes a defer errror. The
4130 incident is still logged to the main and panic logs.
4132 74. Buglet left over from Exim 3 conversion. The message "too many messages
4133 in one connection" was written to the rejectlog but not the mainlog, except
4134 when address rewriting (yes!) was being logged.
4136 75. Added write_rejectlog option.
4138 76. When a system filter was run not as root (that is, when system_filter_user
4139 was set), the values of the $n variables were not being returned to the
4140 main process; thus, they were not subsequently available in the $sn
4143 77. Added +return_path_on_delivery log selector.
4145 78. A connection timeout was being treated differently from recipients deferred
4146 when testing hosts_max_try with a message that was older than the host's
4147 retry timeout. (The host should not be counted, thus allowing all hosts to
4148 be tried at least once before bouncing.) This may have been the cause of an
4149 occasionally reported bug whereby a message would remain on the queue
4150 longer than the retry timeout, but would be bounced if a delivery was
4151 forced. I say "may" because I never totally pinned down the problem;
4152 setting up timeout/retry tests is difficult. See also the next item.
4154 79. The ultimate address timeout was not being applied to errors that involved
4155 a combination of host plus message (for example, a timeout on a MAIL
4156 command). When an address resolved to a number of possible hosts, and they
4157 were not all tried for each delivery (e.g. because of hosts_max_try), a
4158 message could remain on the queue longer than the retry timeout.
4160 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4163 81. Fixed an obscure SMTP outgoing bug which required at least the following
4164 conditions: (a) there was another message waiting for the same server;
4165 (b) the server returned 5xx to all RCPT commands in the first message so
4166 that the message was not completed; (c) the server dropped the connection
4167 or gave a negative response to the RSET that Exim sends to abort the
4168 transaction. The observed case was a dropped connection after DATA that had
4169 been sent in pipelining mode. That is, the server had advertised PIPELINING
4170 but was not implementing it correctly. The effect of the bug was incorrect
4171 behaviour, such as trying another host, and this could lead to a crash.
4177 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4178 and daemon.c were passed as pointers to ints; they should have been
4179 pointers to socklen_t variables (which are typically unsigned ints).
4181 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4184 3. Fixed a really odd bug that affected only the testing scheme; patching a
4185 certain fixed string in the binary changed the value of another string that
4186 happened to be identical to the end of the original first string.
4188 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4189 name", it returns that address as the IP address. On some operating
4190 systems (e.g. Solaris), it also passes back the IP address string as the
4191 "host name". However, on others (e.g. Linux), it passes back an empty
4192 string. Exim wasn't checking for this, and was changing the host name to an
4193 empty string, assuming it had been canonicized.
4195 5. Although rare, it is permitted to have more than one PTR record for a given
4196 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4197 all the names associated with an address, because they do in Solaris.
4198 However, it seems that they do not in Linux for data that comes from the
4199 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4200 I found this out when I moved to a new Linux workstation and tried to run
4201 the Exim test suite.
4203 To get round this problem I have changed the code so that it now does its
4204 own call to the DNS to look up PTR records when searching for a host name.
4205 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4206 addresses that are only in /etc/hosts are still found.
4208 This behaviour is, however, controlled by an option called host_lookup_
4209 order, which defaults to "bydns:byaddr". If people want to use the other
4210 order, or indeed, just use one or the other means of lookup, they can
4211 specify it in this variable.
4213 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4214 some operating systems, this comes back from gethostbyaddr() as an empty
4215 string, and this is what Exim used to test for. However, it seems that in
4216 other systems, "." is yielded. Exim now tests for this case too.
4218 7. The values of check_spool_space and check_log_space are now held internally
4219 as a number of kilobytes instead of an absolute number of bytes. If a
4220 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4221 kilobyte. This means that much larger values can be stored.
4223 8. Exim monitor: an attempt to get the action menu when not actually pointing
4224 at a message produces an empty menu entitled "No message selected". This
4225 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4226 no entries in it ("Shell widget menu has zero width and/or height"). So I
4227 have added a single, blank menu entry in this case.
4229 9. Added ${quote_local_part.
4231 10. MIME decoding is now applied to the contents of Subject: header lines when
4234 11. Now that a reference to $sender_host_address automatically causes a reverse
4235 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4236 host lookup before query-style lookups in lists that might use this
4237 variable. This has therefore been abolished, and the "net-" prefix is no
4238 longer necessary for query-style lookups.
4240 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4241 have been a typo for LFLAGS, so it has been changed.
4243 13. The install script calls Exim with "-C /dev/null" in order to find the
4244 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4245 to be output. Howeve, since Exim outputs its version number before the
4246 error, it didn't break the script. It just looked ugly. I fixed this by
4247 always allowing "-C /dev/null" if the caller is root.
4249 14. Ignore overlarge ACL variable number when reading spool file - insurance
4250 against a later release with more variables having written the file.
4252 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4253 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4254 this, as well as the form without the "IPv6" on the front.
4256 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4257 OS/Makefile-Darwin file.
4259 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4260 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4262 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4263 variable. However, some OS use ints for inodes. I've added cast to long int
4264 to get rid of the compiler warning.
4266 19. I had forgotten to lock out "/../" in configuration file names when
4267 ALT_CONFIG_PREFIX was set.
4269 20. Routers used for verification do not need to specify transports. However,
4270 if such a router generated a host list, and callout was configured, Exim
4271 crashed, because it could not find a port number from the (non-existent)
4272 transport. It now assumes port 25 in this circumstance.
4274 21. Added the -t option to exigrep.
4276 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4277 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4278 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4279 them is set, LOOKUP_LSEARCH is forced.
4281 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4283 24. Added sender and host information to the "rejected by local_scan()" log
4284 line; previously there was no indication of these.
4286 25. Added .include_if_exists.
4288 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4289 Linux. It turns out that not all file systems support this. Apparently some
4290 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4291 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4292 not supported on the file descriptor, is now ignored when Exim is trying to
4293 sync a directory. This applies only to Linux.
4295 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4297 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4298 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4301 29. Added support for maildirsize files from supplied patch (modified a bit).
4303 30. The use of :fail: followed by an empty string could lead Exim to respond to
4304 sender verification failures with (e.g.):
4306 550 Verification failed for <xxx>
4307 550 Sender verify failed
4309 where the first response line was missing the '-' that indicates it is not
4310 the final line of the response.
4312 31. The loop for finding the name of the user that called Exim had a hardwired
4313 limit of 10; it now uses the value of finduser_retries, which is used for
4314 all other user lookups.
4316 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4319 33. Exim was neglecting to zero errno before one call of strtol() when
4320 expanding a string and expecting an integer value. On some systems this
4321 resulted in spurious "integer overflow" errors. Also, it was casting the
4322 result into an int without checking.
4324 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4325 did not work. The code looks as if it has *never* worked, though it appears
4326 to have been documented since at least releast 1.62. I have made it work.
4328 35. The "timeout_DNS" error in retry rules, also documented since at least
4329 1.62, also never worked. As it isn't clear exactly what this means, and
4330 clearly it isn't a major issue, I have abolished the feature by treating it
4331 as "timeout", and writing a warning to the main and panic logs.
4333 36. The display of retry rules for -brt wasn't always showing the error code
4336 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4337 timeout_connect_A, timeout_connect_MX.
4339 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4340 to the empty sender.
4342 39. The daemon was not analysing the content of -oX till after it had closed
4343 stderr and disconnected from the controlling terminal. This meant that any
4344 syntax errors were only noted on the panic log, and the return code from
4345 the command was 0. By re-arranging the code a little, I've made the
4346 decoding happen first, so such errors now appear on stderr, and the return
4347 code is 1. However, the actual setting up of the sockets still happens in
4348 the disconnected process, so errors there are still only recorded on the
4351 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4352 connections (as happens on some IP stacks) was logged at start up time as
4353 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4354 it from "IPv6 and IPv4", which means that two separate sockets are being
4357 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4358 says whether AF_INET or AF_INET6 was passed as an argument.
4360 42. Exiwhat output was messed up when time zones were included in log
4363 43. Exiwhat now gives more information about the daemon's listening ports,
4364 and whether -tls-on-connect was used.
4366 44. The "port" option of the smtp transport is now expanded.
4368 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4369 silently ignored. Now an error message is written to the main and panic
4372 46. There's a new ACL modifier called "logwrite" which writes to a log file
4373 as soon as it is encountered.
4375 47. Added $local_user_uid and $local_user_gid at routing time.
4377 48. Exim crashed when trying to verify a sender address that was being
4380 49. Exim was recognizing only a space character after ".include". It now also
4381 recognizes a tab character.
4383 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4384 extracting the relevant information from the specification. The man page no
4385 longer contains scrambled data for the -d option, and I've added a section
4386 at the front about calling Exim under different names.
4388 51. Added "extra_headers" argument to the "mail" command in filter files.
4390 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4393 53. Installed eximstats 1.29.
4395 54. Added transport_filter_timeout as a generic transport option.
4397 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4398 Cc: header lines. This was required by RFC 822, but it not required by RFC
4401 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4402 incoming messages that did not have them. Now it does so only if the
4403 message originates locally, that is, if there is no associated remote host
4404 address. When Resent- header lines are present, this applies to the Resent-
4405 lines rather than the non-Resent- lines.
4407 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4408 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4410 58. Messages for configuration errors now include the name of the main
4411 configuration file - useful now that there may be more than one file in a
4412 list (.included file names were always shown).
4414 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4415 for those rare installations that do not start the daemon as root or run it
4416 setuid root. I've cut out the call to initgroups() if the daemon is not
4419 60. The Exim user and group can now be bound into the binary as text strings
4420 that are looked up at the start of Exim's processing.
4422 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4424 62. Added $mailstore_basename variable.
4426 63. Installed patch to sieve.c from Michael Haardt.
4428 64. When Exim failed to open the panic log after failing to open the main log,
4429 the original message it was trying to log was written to stderr and debug
4430 output, but if they were not available (the usual case in production), it
4431 was lost. Now it is written to syslog before the two lines that record the
4432 failures to open the logs.
4434 65. Users' Exim filters run in subprocesses under the user's uid. It is
4435 possible for a "deliver" command or an alias in a "personal" command to
4436 provoke an address rewrite. If logging of address rewriting is configured,
4437 this fails because the process is not running as root or exim. There may be
4438 a better way of dealing with this, but for the moment (because 4.30 needs
4439 to be released), I have disabled address rewrite logging when running a
4440 filter in a non-root, non-exim process.
4446 1. The buildconfig auxiliary program wasn't quoting the value set for
4447 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4448 not defined. This bug was masked in 4.22 by the effect that was fixed in
4451 2. Some messages that were rejected after a message id was allocated were
4452 shown as "incomplete" by exigrep. It no longer does this for messages that
4453 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4455 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4456 have allow_domain_literals set, the ID did not get logged in the <= line.
4457 Domain literals are now always recognized in Message-ID: header lines.
4459 4. The first argument for a ${extract expansion item is the key name or field
4460 number. Leading and trailing spaces in this item were not being ignored,
4461 causing some misleading effects.
4463 5. When deliver_drop_privilege was set, single queue runner processes started
4464 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4465 same command in the process it spins off) were not dropping privilege.
4467 6. When the daemon running as "exim" started a queue runner, it always
4468 re-executed Exim in the spun-off process. This is a waste of effort when
4469 deliver_drop_privilege is set. The new process now just calls the
4470 queue-runner function directly.
4476 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4479 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4480 ignored. Though the use of -oP was forcing the writing of a pid file, it
4481 was always written to the default place.
4483 3. If the message "no IP address found for host xxxx" is generated during
4484 incoming verification, it is now followed by identification of the incoming
4485 connection (so you can more easily find what provoked it).
4487 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4489 5. Added some features to "harden" Exim a bit more against certain attacks:
4491 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4492 be put in Local/Makefile. This is like the never_users runtime option,
4493 but it cannot be overridden. The default setting is "root".
4495 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4496 prefix string with which any file named in a -C command line option
4499 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4500 is retained for -C and -D only if the caller of Exim is root. Without
4501 it, the exim user may also use -C and -D and retain privilege.
4503 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4504 command line option is disabled.
4506 6. Macro names set by the -D option must start with an upper case letter, just
4507 like macro names defined in the configuration file.
4509 7. Added "dereference=" facility to LDAP.
4511 8. Two instances of the typo "uknown" in the source files are fixed.
4513 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4514 the Configure-Makefile script screwed up while processing it.
4516 10. Incorporated PCRE 4.4.
4518 11. The SMTP synchronization check was not operating right at the start of an
4519 SMTP session. For example, it could not catch a HELO sent before the client
4520 waited for the greeting. There is now a check for outstanding input at the
4521 point when the greeting is written. Because of the duplex, asynchronous
4522 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4523 way, but not yet received, when the check is performed.
4525 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4526 on TCP/IP sockets, because this apparently causes some broken clients to
4529 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4530 unchanged) from the Cygwin maintainer.
4532 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4533 was supported instead of testing for mbx. Effectively a typo.
4535 15. The spa authenticator server code was not checking that the input it
4536 received was valid base64.
4538 16. The debug output line for the "set" modifier in ACLs was not showing the
4539 name of the variable that was being set.
4541 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4543 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4544 lookup on its own. Something else was needed to trigger the lookup. For
4545 example, a match in host_lookup or the need for a host name in a host list.
4546 Now, if $sender_host_name is referenced and the host name has not yet been
4547 looked up, a lookup is performed. If the lookup fails, the variable remains
4548 empty, and $host_lookup_failed is set to "1".
4550 19. Added "eqi" as a case-independent comparison operator.
4552 20. The saslauthd authentication condition could segfault if neither service
4553 nor realm was specified.
4555 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4556 error message that was logged was misleading, and incoming SMTP
4557 connections were dropped. The message is now more accurate, and temporary
4558 errors are given to SMTP connections.
4560 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4561 (except RSET) with a 421 error, until QUIT is received. However, it was
4562 failing to send a response to QUIT.
4564 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4565 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4566 ACL is now run after the helo_try_verify_hosts code.
4568 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4569 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4570 case-independent, so other case variants are also recognized). Apparently
4571 some systems use these upper case variants.
4573 25. If more than two messages were waiting for the same host, and a transport
4574 filter was specified for the transport, Exim sent two messages over the
4575 same TCP/IP connection, and then failed with "socket operation on non-
4576 socket" when it tried to send the third.
4578 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4580 27. The extern definition of crypt16() in expand.c was not being excluded when
4581 the OS had its own crypt16() function.
4583 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4584 as a preferred synonym for return_size_limit, both as an option and as an
4587 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4589 30. Changed the default configuration ACL to relax the local part checking rule
4590 for addresses that are not in any local domains. For these addresses,
4591 slashes and pipe symbols are allowed within local parts, but the sequence
4592 /../ is explicitly forbidden.
4594 31. SPA server authentication was not clearing the challenge buffer before
4597 32. log_message in a "warn" ACL statement was writing to the reject log as
4598 well as to the main log, which contradicts the documentation and doesn't
4599 seem right (because no rejection is happening). So I have stopped it.
4601 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4602 However, I am unable to do any testing of this.
4604 34. Fixed an infelicity in the appendfile transport. When checking directories
4605 for a mailbox, to see if any needed to be created, it was accidentally
4606 using path names with one or more superfluous leading slashes; tracing
4607 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4609 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4610 contents are added to the log line that is written for every discarded
4611 recipient. (Previously a log_message setting was ignored.)
4613 36. The ${quote: operator now quotes the string if it is empty.
4615 37. The install script runs exim in order to find its version number. If for
4616 some reason other than non-existence or emptiness, which it checks, it
4617 could not run './exim', it was installing it with an empty version number,
4618 i.e. as "exim-". This error state is now caught, and the installation is
4621 38. An argument was missing from the function that creates an error message
4622 when Exim fails to connect to the socket for saslauthd authentication.
4623 This could cause Exim to crash, or give a corrupted message.
4625 39. Added isip, isip4, and isip6 to ${if conditions.
4627 40. The ACL variables $acl_xx are now saved with the message, and can be
4628 accessed later in routers, transports, and filters.
4630 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4631 strings in the file are not string-expanded.
4633 42. If a MAIL command specified a SIZE value that was too large to fit into an
4634 int variable, the check against message_size_limit failed. Such values are
4635 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4636 day this will have to be increased, but I don't think I want to be around
4637 when emails are that large.
4644 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4645 iconv() is not standard in FreeBSD.
4647 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4648 IPv6 enabled. The observed symptom was a segmentation fault on return from
4649 the function os_common_find_running_interfaces() in src/os.c.
4651 3. In the check_special_case() function in daemon.c I had used "errno" as an
4652 argument name, which causes warnings on some systems. This was basically a
4653 typo, since it was named "eno" in the comments!
4655 4. The code that waits for the clock to tick (at a resolution of some fraction
4656 of a second) so as to ensure message-id uniqueness was always waiting for
4657 at least one whole tick, when it could have waited for less. [This is
4658 almost certainly not relevant at current processor speeds, where it is
4659 unlikely to ever wait at all. But we try to future-proof.]
4661 5. The function that sleeps for a time interval that includes fractions of a
4662 second contained a race. It did not block SIGALRM between setting the
4663 timer, and suspending (a couple of lines later). If the interval was short
4664 and the sigsuspend() was delayed until after it had expired, the suspension
4665 never ended. On busy systems this could lead to processes getting stuck for
4668 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4669 process, before it forks any delivery processes. The open lookup caching
4670 mechanism meant that the open file or database connection was passed into
4671 the delivery process. The problem was that delivery processes always tidy
4672 up cached lookup data. This could cause a problem for the next delivery
4673 process started by the queue runner, because the external queue runner
4674 process does not know about the closure. So the next delivery process
4675 still has data in the lookup cache. In the case of a file lookup, there was
4676 no problem because closing a file descriptor in a subprocess doesn't affect
4677 the parent. However, if the lookup was caching a connection to a database,
4678 the connection was closed, and the second delivery process was likely to
4679 see errors such as "PGSQL: query failed: server closed the connection
4680 unexpectedly". The problem has been fixed by closing all cached lookups
4681 in a queue runner before running a delivery process.
4683 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4684 seem to have the "const" qualifier which it has on other OS. I've
4687 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4688 *of the same type* (client or server) with the same public name that an
4689 error should be diagnosed.
4691 9. When Exim looked up a host name for an IP address, but failed to find the
4692 original IP address when looking up the host name (a safety check), it
4693 output the message "<ip address> does not match any IP for NULL", which was
4694 confusing, to say the least. The bug was that the host name should have
4695 appeared instead of "NULL".
4697 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4698 user that is built into the binary, and the -C or -D options is used, root
4699 privilege is dropped before the configuration file is read. In addition,
4700 logging is switched to stderr instead of the normal log files. If the
4701 configuration then re-defines the Exim user, the unprivileged environment
4702 is probably not what is expected, so Exim logs a panic warning message (but
4705 However, if deliver_drop_privilege is set, the unprivileged state may well
4706 be exactly what is intended, so the warning has been cut out in that case,
4707 and Exim is allowed to try to write to its normal log files.
4713 1. smtp_return_error_details was not giving details for temporary sender
4714 or receiver verification errors.
4716 2. Diagnose a configuration error if two authenticators have the same public
4719 3. Exim used not to create the message log file for a message until the first
4720 delivery attempt. This could be confusing when incoming messages were held
4721 for policy or load reasons. The message log file is now created at the time
4722 the message is received, and an initial "Received" line is written to it.
4724 4. The automatically generated man page for command line options had a minor
4725 bug that caused no ill effects; however, a more serious problem was that
4726 the procedure for building the man page automatically didn't always
4727 operate. Consequently, release 4.20 contains an out-of-date version. This
4728 shouldn't happen again.
4730 5. When building Exim with embedded Perl support, the script that builds the
4731 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4732 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4734 6. The freeze_tell option was not being used for messages that were frozen on
4735 arrival, either by an ACL or by local_scan().
4737 7. Added the smtp_incomplete_transaction log selector.
4739 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4740 was accepting AUTH without a new EHLO.
4742 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4743 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4746 10. Logging of TCP/IP connections (when configured) now happens in the main
4747 daemon process instead of the child process, so that the TCP/IP connection
4748 count is more accurate (but it can never be perfect).
4750 11. The use of "drop" in a nested ACL was not being handled correctly in the
4751 outer ACL. Now, if condition failure induced by the nested "drop" causes
4752 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4753 or "require"), the connection is dropped.
4755 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4756 that yield "discard" can now be used with an "accept" or a "discard" verb,
4757 but an error is generated for any others (because I can't see a useful way
4758 to define what should happen).
4760 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4761 are now processed in the same way as lines in the Exim configuration file.
4762 In particular, continuation lines are supported.
4764 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4766 15. Added -ti meaning -t -i.
4768 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4769 domains, and warn by logging if others are found.
4771 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4772 returned from GIFCONF when Exim is trying to find the list of interfaces on
4773 a host. The code in os.c has been modified to copy each ifreq to an aligned
4774 structure in all cases.
4776 Also, in some cases, the returned ifreq's were being copied to a 'struct
4777 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4778 means the last couple of bytes of an IPv6 address could be chopped if the
4779 ifreq contained only a normal sockaddr (14 bytes storage).
4781 18. Named domain lists were not supported in the hosts_treat_as_local option.
4782 An entry such as +xxxx was not recognized, and was treated as a literal
4785 19. Ensure that header lines added by a DATA ACL are included in the reject log
4786 if the ACL subsequently rejects the message.
4788 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4789 MD5 (which is deprecated).
4791 21. When testing a filter file using -bf, Exim was writing a message when it
4792 took the sender from a "From " line in the message, but it was not doing so
4793 when it took $return_path from a Return-Path: header line. It now does.
4795 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4796 with a valid header line field name (a series of printing characters
4797 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4799 23. Changed "disc" in the source to "disk" to conform to the documentation and
4800 the book and for uniformity.
4802 24. Ignore Sendmail's -Ooption=value command line item.
4804 25. When execve() failed while trying to run a command in a pipe transport,
4805 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4806 could be confused with a return value of 69 from the command itself. This
4807 has been changed to 127, the value the shell returns if it is asked to run
4808 a non-existent command. The wording for the related log line suggests a
4809 non-existent command as the problem.
4811 26. If received_header_text expands to an empty string, do not add a Received:
4812 header line to the message. (Well, it adds a token one on the spool, but
4813 marks it "old" so that it doesn't get used or transmitted.)
4815 27. Installed eximstats 1.28 (addition of -nt option).
4817 28. There was no check for failure on the call to getsockname() in the daemon
4818 code. This can fail if there is a shortage of resources on the system, with
4819 ENOMEM, for example. A temporary error is now given on failure.
4821 29. Contrary to the C standard, it seems that in some environments, the
4822 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4823 program. Exim now does this explicitly; it affects the formatting of
4824 timestamps using strftime().
4826 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4827 complaints. I've now initialized all the variables, to avoid this.
4829 32. Header lines added by a system filter were not being "seen" during
4830 transport-time rewrites.
4832 33. The info_callback() function passed to OpenSSL is set up with type void
4833 (*)(SSL *, int, int), as described somewhere. However, when calling the
4834 function (actually a macro) that sets it up, the type void(*)() is
4835 expected. I've put in a cast to prevent warnings from picky compilers.
4837 34. If a DNS black list lookup found a CNAME record, but there were no A
4838 records associated with the domain it pointed at, Exim crashed.
4840 35. If a DNS black list lookup returned more than one A record, Exim ignored
4841 all but the first. It now scans all returned addresses if a particular IP
4842 value is being sought. In this situation, the contents of the
4843 $dnslist_value variable are a list of all the addresses, separated by a
4846 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4847 to accept a host name and all its aliases if the forward lookup for any of
4848 them yielded the IP address of the incoming connection. Now it accepts only
4849 those names whose forward lookup yields the correct IP address. Any other
4850 names are discarded. This closes a loophole whereby a rogue DNS
4851 administrator could create reverse DNS records to break through a
4852 wildcarded host restriction in an ACL.
4854 37. If a user filter or a system filter that ran in a subprocess used any of
4855 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4856 the wrong values were passed to the pipe command ($thisaddress had the
4857 value of $0, $0 had the value of $1, etc). This bug was introduced by
4858 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4860 38. Improved the line breaking for long SMTP error messages from ACLs.
4861 Previously, if there was no break point between 40 and 75 characters, Exim
4862 left the rest of the message alone. Two changes have been made: (a) I've
4863 reduced the minimum length to 35 characters; (b) if it can't find a break
4864 point between 35 and 75 characters, it looks ahead and uses the first one
4865 that it finds. This may give the occasional overlong line, but at least the
4866 remaining text gets split now.
4868 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4869 file descriptors might be low, and that setting 1000 would always raise it.
4870 It turns out that in some environments, the limit is already over 1000 and
4871 that lowering it causes trouble. So now Exim takes care not to decrease it.
4873 40. When delivering a message, the value of $return_path is set to $sender_
4874 address at the start of routing (routers may change the value). By an
4875 oversight, this default was not being set up when an address was tested by
4876 -bt or -bv, which affected the outcome if any router or filter referred to
4879 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4880 log only when "message" or "log_message" are set. However, if one of the
4881 conditions was an address verification, or a call to a nested ACL, the
4882 messages generated by the underlying test were being passed through. This
4883 no longer happens. The underlying message is available in $acl_verify_
4884 message for both "message" and "log_message" expansions, so it can be
4885 passed through if needed.
4887 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4888 new expansion $bh_ to give the encoded byte string without charset
4889 translation. Translation happens only if iconv() is available; HAVE_ICONV
4890 indicates this at build time. HEADERS_CHARSET gives the charset to
4891 translate to; headers_charset can change it in the configuration, and
4892 "headers charset" can change it in an individual filter file.
4894 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4895 that creates RFC 2047 encoded "words" labels them as that charset instead
4896 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4897 expansion operator; (ii) when Exim creates a From: line for a local
4898 message; (iii) when a header line is rewritten to include a "phrase" part.
4900 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4901 buggy, causing it to skip the first lines of messages whose message ID
4902 ended in 'D'. This would not have bitten before Exim release 4.14, because
4903 message IDs were unlikely to end in 'D' before then. The effect was to have
4904 incorrect size information for certain domains.
4906 45. #include "config.h" was missing at the start of the crypt16.c module. This
4907 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4910 46. If there was a timeout during a "random" callout check, Exim treated it as
4911 a failure of the random address, and carried on sending RSET and the real
4912 address. If the delay was just some slowness somewhere, the response to the
4913 original RCPT would be taken as a response to RSET and so on, causing
4914 mayhem of various kinds.
4916 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4917 when I implemented it. It didn't allow for the fact that some option values
4918 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4919 the right test for positive values.
4921 48. Domain names in DNS records are case-independent. Exim always looks them up
4922 in lower case. Some resolvers return domain names in exactly the case they
4923 appear in the zone file, that is, they may contain uppercase letters. Not
4924 all resolvers do this - some return always lower case. Exim was treating a
4925 change of case by a resolver as a change of domain, similar to a widening
4926 of a domain abbreviation. This triggered its re-routing code and so it was
4927 trying to route what was effectively the same domain again. This normally
4928 caused routing to fail (because the router wouldn't handle the domain
4929 twice). Now Exim checks for this case specially, and just changes the
4930 casing of the domain that it ultimately uses when it transmits the message
4933 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4936 50. If a filter generated a file delivery with a non-absolute name (possible if
4937 no home directory exists for the router), the forbid_file option was not
4940 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4941 the existing equality matching.
4943 52. Exim was using ints instead of ino_t variables in some places where it was
4944 dealing with inode numbers.
4946 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4947 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4948 called TMPDIR, and if it finds it is different, it changes its value.
4950 54. The smtp_printf() function is now made available to local_scan() so
4951 additional output lines can be written before returning. There is also an
4952 smtp_fflush() function to enable the detection of a dropped connection.
4953 The variables smtp_input and smtp_batched_input are exported to
4956 55. Changed the default runtime configuration: the message "Unknown user"
4957 has been removed from the ACL, and instead placed on the localuser router,
4958 using the cannot_route_message feature. This means that any verification
4959 failures that generate their own messages won't get overridden. Similarly,
4960 the "Unrouteable address" message that was in the ACL for unverifiable
4961 relay addresses has also been removed.
4963 56. Added hosts_avoid_esmtp to the smtp transport.
4965 57. The exicyclog script was not checking for the esoteric option
4966 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4967 will work only if exicyclog is run under the appropriate euid.
4969 58. Following a discussion on the list, the rules by which Exim recognises line
4970 endings on incoming messages have been changed. The -dropcr and drop_cr
4971 options are now no-ops, retained only for backwards compatibility. The
4972 following line terminators are recognized: LF CRLF CR. However, special
4973 processing applies to CR:
4975 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4976 nor a local message in the state where . is a terminator.
4978 (ii) If a bare CR is encountered in a header line, an extra space is added
4979 after the line terminator so as not to end the header. The reasoning
4980 behind this is that bare CRs in header lines are most likely either
4981 to be mistakes, or people trying to play silly games.
4983 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4984 was being incorrectly given as 18 bytes larger than it should have been.
4985 This is a VOB (very old bug).
4987 60. This may never have affected anything current, but just in case it has:
4988 When the local host is found other than at the start of a list of hosts,
4989 the local host, those with the same MX, and any that follow, are discarded.
4990 When the list in question was part of a longer list of hosts, the following
4991 hosts (not currently being processed) were also being discarded. This no
4992 longer happens. I'm not sure if this situation could ever has previously
4995 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4998 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4999 header. This is something that is recommended in a new Internet Draft, and
5000 is something that is documented as being done by Sendmail. There are two
5001 possible values. For messages generated by the autoreply transport, Exim
5004 Auto-Submitted: auto-replied
5006 whereas for all other generated messages (e.g. bounces) it adds
5008 Auto-Submitted: auto-generated
5010 63. The "personal" condition in filters now includes a test for the
5011 Auto-Submitted: header. If it contains the string "auto-" the message it
5012 not considered personal.
5014 64. Added rcpt_include_affixes as a generic transport option.
5016 65. Added queue_only_override (default true).
5018 66. Added the syslog_duplication option.
5020 67. If what should have been the first header line of a message consisted of
5021 a space followed by a colon, Exim was mis-interpreting it as a header line.
5022 It isn't of course - it is syntactically invalid and should therefore be
5023 treated as the start of the message body. The misbehaviour could have
5024 caused a number of strange effects, including loss of data in subsequent
5025 header lines, and spool format errors.
5027 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5028 client host had authenticated. This control can now be exercised by an ACL
5029 for more flexibility.
5031 69. By default, callouts do not happen when testing with -bh. There is now a
5032 variant, -bhc, which does actually run the callout code, including
5033 consulting and updating the callout cache.
5035 70. Added support for saslauthd authentication, courtesy of Alexander
5038 71. If statvfs() failed on the spool or log directories while checking their
5039 size for availability, Exim confusingly gave the error "space shortage".
5040 Furthermore, in debugging mode it crashed with a floating point exception.
5041 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5042 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5043 serious problem, Exim now writes to the main and panic logs when this
5044 happens, with details of the failure. It then refuses to accept the
5045 incoming message, giving the message "spool directory problem" or "log
5046 directory problem" with a 421 code for SMTP messages.
5048 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5049 0, 1, and 2 exist, because some OS complain for execs without them (see
5050 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5051 descriptors. However, the code omitted to check that the open succeeded,
5052 causing mysterious errors if for some reason the permissions on /dev/null
5053 got screwed. Now Exim writes a message to the main and panic logs, and
5054 bombs out if it can't open /dev/null.
5056 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5057 interact so that it is all more flexible. It is supposed to remain
5058 backwards compatible. Also added extra_local_interfaces.
5060 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5061 to bomb out with an assertion failure - to the client this appears as a
5062 connection drop. This problem occurs in the part of the code that was taken
5063 from the Samba project. Fortunately, the assertion is in a very simple
5064 function, so I have fixed this by reproducing the function inline in the
5065 one place where it is called, and arranging for authentication to fail
5066 instead of killing the process with assert().
5068 75. The SPA client code was not working when the server requested OEM rather
5069 than Unicode encoding.
5071 76. Added code to make require_files with a specific uid setting more usable in
5072 the case where statting the file as root fails - usually a non-root-mounted
5073 NFS file system. When this happens and the failure is EACCES, Exim now
5074 forks a subprocess and does the per-uid checking as the relevant uid.
5076 77. Added process_log_path.
5078 78. If log_file_path was not explicitly set, a setting of check_log_space or
5079 check_log_inodes was ignored.
5081 79. If a space check for the spool or log partitions fails, the incident is now
5082 logged. Of course, in the latter case the data may get lost...
5084 80. Added the %p formatting code to string_format() so that it can be used to
5085 print addresses in debug_print(). Adjusted all the address printing in the
5086 debugging in store.c to use %p rather than %d.
5088 81. There was a concern that a line of code in smtp_in.c could overflow a
5089 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5090 initially expressed, the concern was not well-founded, because trailing
5091 spaces are removed early. However, if the trailing spaces were followed by
5092 a NULL, they did not get removed, so the overflow was possible. Two fixes
5095 (a) I re-wrote the offending code in a cleaner fashion.
5096 (b) If an incoming SMTP command contains a NULL character, it is rejected
5099 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5100 runs initgroups(), so that if the Exim user is in any additional groups,
5101 they will be used during message reception.
5107 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob