1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.594 2009/11/14 20:35:55 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
15 TK/01 Added patch by Johannes Berg that expands the main option
16 "spamd_servers" if it starts with a dollar sign.
18 TK/02 Write list of recipients to X-Envelope-Sender header when building
19 the mbox-format spool file for content scanning (suggested by Jakob
22 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
23 (http://www.dcc-servers.net/) support via dccifd. Activated by
24 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
25 experimental_spec.txt for more documentation.
27 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
28 by Mark Daniel Reidel <mr@df.eu>.
30 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
31 When building exim an external PCRE library is now needed -
32 PCRE is a system library on the majority of modern systems.
33 See entry on PCRE_LIBS in EDITME file.
35 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
36 conversation. Added nologin parameter to request.
37 Patch contributed by Kirill Miazine
39 TF/01 Do not log submission mode rewrites if they do not change the address.
41 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
43 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
44 log files in place. Contributed by Roberto Lima
46 NM/04 Bugzilla 667: close socket used by dovecot authenticator
48 TF/03 Bugzilla 615: When checking the local_parts router precondition
49 after a local_part_suffix or local_part_prefix option, Exim now
50 does not use the address's named list lookup cache, since this
51 contains cached lookups for the whole local part.
53 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
54 Robert Millan. Documentation is in experimental-spec.txt
56 TF/04 Bugzilla 668: Fix parallel build (make -j).
58 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
60 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
61 Patch provided by Jan Srzednicki
63 TF/05 Leading white space used to be stripped from $spam_report which
64 wrecked the formatting. Now it is preserved.
66 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
67 that they are available at delivery time.
69 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
71 TF/08 TLS error reporting now respects the incoming_interface and
72 incoming_port log selectors.
74 TF/09 Produce a more useful error message if an SMTP transport's hosts
75 setting expands to an empty string.
77 NM/06 Bugzilla 744: EXPN did not work under TLS.
78 Patch provided by Phil Pennock
80 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
81 Patch provided by Richard Godbee
83 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
84 acl_smtp_notquit, added index entry.
86 NM/09 Bugzilla 787: Potential buffer overflow in string_format
87 Patch provided by Eugene Bujak
89 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
90 Patch provided by Maxim Dounin
92 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
93 Patch provided by Phil Pennock
95 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
97 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
98 Patch provided by Brad "anomie" Jorsch
100 NM/14 Bugzilla 622: Added timeout setting to transport filter.
101 Patch provided by Dean Brooks
103 TK/05 Add native DKIM support (does not depend on external libraries).
105 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
106 Patch provided by Graeme Fowler
108 NM/16 Bugzilla 851: Documentation example syntax fix.
110 NM/17 Changed NOTICE file to remove references to embedded PCRE.
112 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
114 NM/19 Bugzilla 745: TLS version reporting
115 Patch provided by Phil Pennock
117 NM/20 Bugzilla 167: bool: condition support
118 Patch provided by Phil Pennock
120 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
121 Patch provided by Phil Pennock
123 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
124 Patch provided by Brad "anomie" Jorsch
126 NM/23 Bugzilla 687: Fix misparses in eximstats
127 Patch provided by Heiko Schlittermann
129 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
130 Patch provided by Heiko Schlittermann
132 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
133 plus update to original patch.
135 NM/26 Bugzilla 799: Documentation correction for ratelimit
137 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
138 Patch provided by David Brownlee
140 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
142 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
144 NM/30 Bugzilla 888: TLS documentation bugfixes
146 NM/31 Bugzilla 896: Dovecot buffer overrun fix
148 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
149 Unlike the original bugzilla I have changed all shell scripts in src tree
151 NM/33 Bugzilla 898: Transport filter timeout fix
152 Patch by Todd Rinaldo
154 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
155 Patch by Serge Demonchaux
157 NM/35 Bugzilla 39: Base64 decode bug fixes
158 Patch by Jakob Hirsch
160 NM/36 Bugzilla 909: Correct connect() call in dcc code
162 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
164 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
166 NM/39 Bugzilla 911: Fixed MakeLinks build script
172 TK/01 Add preliminary DKIM support. Currently requires a forked version of
173 ALT-N's libdkim that I have put here:
174 http://duncanthrax.net/exim-experimental/
176 Note to Michael Haardt: I had to rename some vars in sieve.c. They
177 were called 'true' and it seems that C99 defines that as a reserved
178 keyword to be used with 'bool' variable types. That means you could
179 not include C99-style headers which use bools without triggering
180 build errors in sieve.c.
182 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
183 as mailq or other aliases. Changed the --help handling significantly
184 to do whats expected. exim_usage() emits usage/help information.
186 SC/01 Added the -bylocaldomain option to eximstats.
188 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
190 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
192 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
198 PH/01 Another patch from the Sieve maintainer.
200 PH/02 When an IPv6 address is converted to a string for single-key lookup
201 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
202 dots are used instead of colons so that keys in lsearch files need not
203 contain colons. This was done some time before quoting was made available
204 in lsearch files. However, iplsearch files do require colons in IPv6 keys
205 (notated using the quote facility) so as to distinguish them from IPv4
206 keys. This meant that lookups for IP addresses in host lists did not work
207 for iplsearch lookups.
209 This has been fixed by arranging for IPv6 addresses to be expressed with
210 colons if the lookup type is iplsearch. This is not incompatible, because
211 previously such lookups could never work.
213 The situation is now rather anomolous, since one *can* have colons in
214 ordinary lsearch keys. However, making the change in all cases is
215 incompatible and would probably break a number of configurations.
217 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
220 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
221 conversion specification without a maximum field width, thereby enabling
222 a rogue spamd server to cause a buffer overflow. While nobody in their
223 right mind would setup Exim to query an untrusted spamd server, an
224 attacker that gains access to a server running spamd could potentially
225 exploit this vulnerability to run arbitrary code as the Exim user.
227 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
228 $primary_hostname instead of what libspf2 thinks the hosts name is.
230 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
231 a directory entry by the name of the lookup key. Previously, if a
232 symlink pointed to a non-existing file or a file in a directory that
233 Exim lacked permissions to read, a lookup for a key matching that
234 symlink would fail. Now it is enough that a matching directory entry
235 exists, symlink or not. (Bugzilla 503.)
237 PH/03 The body_linecount and body_zerocount variables are now exported in the
240 PH/04 Added the $dnslist_matched variable.
242 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
243 This means they are set thereafter only if the connection becomes
246 PH/06 Added the client_condition to authenticators so that some can be skipped
247 by clients under certain conditions.
249 PH/07 The error message for a badly-placed control=no_multiline_responses left
250 "_responses" off the end of the name.
252 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
254 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
255 (without spaces) instead of just copying the configuration text.
257 PH/10 Added the /noupdate option to the ratelimit ACL condition.
259 PH/11 Added $max_received_linelength.
261 PH/12 Added +ignore_defer and +include_defer to host lists.
263 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
264 way in which PCRE > 7.0 is built.
266 PH/14 Implemented queue_only_load_latch.
268 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
269 MAIL command. The effect was to mangle the value on 64-bit systems.
271 PH/16 Another patch from the Sieve maintainer.
273 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
275 PH/18 If a system quota error occurred while trying to create the file for
276 a maildir delivery, the message "Mailbox is full" was not appended to the
277 bounce if the delivery eventually timed out. Change 4.67/27 below applied
278 only to a quota excession during the actual writing of the file.
280 PH/19 It seems that peer DN values may contain newlines (and other non-printing
281 characters?) which causes problems in log lines. The DN values are now
282 passed through string_printing() before being added to log lines.
284 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
285 and InterBase are left for another time.)
287 PH/21 Added message_body_newlines option.
289 PH/22 Guard against possible overflow in moan_check_errorcopy().
291 PH/23 POSIX allows open() to be a macro; guard against that.
293 PH/24 If the recipient of an error message contained an @ in the local part
294 (suitably quoted, of course), incorrect values were put in $domain and
295 $local_part during the evaluation of errors_copy.
301 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
302 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
305 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
306 issue a MAIL command.
308 PH/02 In an ACL statement such as
310 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
312 if a client was not listed at all, or was listed with a value other than
313 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
314 the condition was not true (as it should be), so access was not denied.
315 The bug was that the ! inversion was incorrectly passed on to the second
316 item. This has been fixed.
318 PH/03 Added additional dnslists conditions == and =& which are different from
319 = and & when the dns lookup returns more than one IP address.
321 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
322 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
324 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
325 FSYNC, which compiles an option called disable_fsync that allows for
326 bypassing fsync(). The documentation is heavily laced with warnings.
328 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
330 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
331 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
332 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
333 including adding "make clean"; (3) Added -fPIC when compiling the test
334 dynamically loaded module, to get rid of a warning.
336 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
337 message fails, move_frozen_messages = true and ignore_bounce_errors_after
338 = 0s. The bug is otherwise harmless.
340 PH/07 There was a bug in the dovecot authenticator such that the value of
341 $auth1 could be overwritten, and so not correctly preserved, after a
342 successful authentication. This usually meant that the value preserved by
343 the server_setid option was incorrect.
345 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
347 PH/09 Installed PCRE release 7.0.
349 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
350 run for batched SMTP input. It is now run at the start of every message
351 in the batch. While fixing this I discovered that the process information
352 (output by running exiwhat) was not always getting set for -bs and -bS
353 input. This is fixed, and it now also says "batched" for BSMTP.
355 PH/11 Added control=no_pipelining.
357 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
358 patch, slightly modified), and move the expansion of helo_data till after
359 the connection is made in the smtp transport (so it can use these
362 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
364 PH/14 Added log_selector = +pid.
366 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
368 PH/16 Add ${if forany and ${if forall.
370 PH/17 Added dsn_from option to vary the From: line in DSNs.
372 PH/18 Flush SMTP output before performing a callout, unless control =
373 no_callout_flush is set.
375 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
376 was true (the default) a successful delivery failed to delete the retry
377 item, thus causing premature timeout of the address. The bug is now
380 PH/20 Added hosts_avoid_pipelining to the smtp transport.
382 PH/21 Long custom messages for fakedefer and fakereject are now split up
383 into multiline reponses in the same way that messages for "deny" and
384 other ACL rejections are.
386 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
387 with slight modification.
389 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
390 draft, changing the syntax and factoring some duplicate code".
392 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
393 for deliveries of the second and subsequent messages over the same SMTP
396 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
397 ${reduce, with only minor "tidies".
399 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
401 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
402 expansion side effects.
404 PH/27 When a message times out after an over-quota error from an Exim-imposed
405 quota, the bounce message says "mailbox is full". This message was not
406 being given when it was a system quota that was exceeded. It now should
409 MH/03 Made $recipients available in local_scan(). local_scan() already has
410 better access to the recipient list through recipients_list[], but
411 $recipients can be useful in postmaster-provided expansion strings.
413 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
414 in the case of a MAIL command with additional options following the
415 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
416 were accidentally chopped off.
418 PH/29 SMTP synchronization checks are implemented when a command is read -
419 there is a check that no more input is waiting when there shouldn't be
420 any. However, for some commands, a delay in an ACL can mean that it is
421 some time before the response is written. In this time, more input might
422 arrive, invalidly. So now there are extra checks after an ACL has run for
423 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
424 pipelining has not been advertised.
426 PH/30 MH's patch to allow iscntrl() characters to be list separators.
428 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
429 returned in the SMTP response when smtp_return_error_details was false.
432 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
433 instead of the C I/O that was originally supplied, because problems were
436 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
437 Exim which did not show up earlier: it was assuming that a call to
438 SSL_CTX_set_info_callback() might give an error value. In fact, there is
439 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
440 was a macro that became an assignment, so it seemed to work. This has
441 changed to a proper function call with a void return, hence the compile
442 error. Exim's code has been fixed.
444 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
447 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
449 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
451 PH/37 If a message is not accepted after it has had an id assigned (e.g.
452 because it turns out to be too big or there is a timeout) there is no
453 "Completed" line in the log. When some messages of this type were
454 selected by exigrep, they were listed as "not completed". Others were
455 picked up by some special patterns. I have improved the selection
456 criteria to be more general.
458 PH/38 The host_find_failed option in the manualroute router can now be set
459 to "ignore", to completely ignore a host whose IP address cannot be
460 found. If all hosts are ignored, the behaviour is controlled by the new
461 host_all_ignored option.
463 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
464 homing or because of multiple MX records with /mx) generated more than
465 one IP address, and the following item turned out to be the local host,
466 all the secondary addresses of the first item were incorrectly removed
467 from the list, along with the local host and any following hosts (which
468 is what is supposed to happen).
470 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
471 whoever called Exim into the -H file. In the case of the daemon it was
472 behaving confusingly. When first started, it used values for whoever
473 started the daemon, but after a SIGHUP it used the Exim user (because it
474 calls itself on a restart). I have changed the code so that it now always
477 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
478 message are rejected with the same error (e.g. no authentication or bad
479 sender address), and a DATA command is nevertheless sent (as can happen
480 with PIPELINING or a stupid MUA), the error message that was given to the
481 RCPT commands is included in the rejection of the DATA command. This is
482 intended to be helpful for MUAs that show only the final error to their
485 PH/42 Another patch from the Sieve maintainer.
487 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
488 Eximstats - Fixed some broken HTML links and added missing column headers
490 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
491 columns for Rejects, Temp Rejects, Ham, and Spam rows.
493 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
495 PH/43 Yet another patch from the Sieve maintainer.
497 PH/44 I found a way to check for a TCP/IP connection going away before sending
498 the response to the final '.' that terminates a message, but only in the
499 case where the client has not sent further data following the '.'
500 (unfortunately, this is allowed). However, in many cases there won't be
501 any further data because there won't be any more messages to send. A call
502 to select() can be used: if it shows that the input is "ready", there is
503 either input waiting, or the socket has been closed. An attempt to read
504 the next input character can distinguish the two cases. Previously, Exim
505 would have sent an OK response which the client would never have see.
506 This could lead to message repetition. This fix should cure that, at
507 least in a lot of common cases.
509 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
510 advertised in response to EHLO.
516 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
517 fixed by 4.65/MH/01 (is this a record?) are fixed:
519 (i) An empty string was always treated as zero by the numeric comparison
520 operators. This behaviour has been restored.
522 (ii) It is documented that the numeric comparison operators always treat
523 their arguments as decimal numbers. This was broken in that numbers
524 starting with 0 were being interpreted as octal.
526 While fixing these problems I realized that there was another issue that
527 hadn't been noticed. Values of message_size_limit (both the global option
528 and the transport option) were treated as octal if they started with 0.
529 The documentation was vague. These values are now always treated as
530 decimal, and I will make that clear in the documentation.
536 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
537 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
540 MH/01 Don't check that the operands of numeric comparison operators are
541 integers when their expansion is in "skipping" mode (fixes bug
542 introduced by 4.64-PH/07).
544 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
545 child addresses, Exim now panics and dies. Previously, because the count
546 is held in a short int, deliveries were likely to be lost. As such a
547 large number of recipients for a single message is ridiculous
548 (performance will be very, very poor), I have chosen to impose a limit
549 rather than extend the field.
555 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
556 leftover -K file (the existence of which was triggered by #402).
557 While we were at it, introduced process PID as part of the -K
558 filename. This should rule out race conditions when creating
561 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
562 processing considerably. Previous code took too long for large mails,
563 triggering a timeout which in turn triggers #401.
565 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
566 in the DK code in transports.c. sendfile() is not really portable,
567 hence the _LINUX specificness.
569 TF/01 In the add_headers option to the mail command in an Exim filter,
570 there was a bug that Exim would claim a syntax error in any
571 header after the first one which had an odd number of characters
574 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
575 callout verification, Exim cached a "reject" for the entire domain. This
576 is correct for most verifications, but it is not correct for a recipient
577 verification with use_sender or use_postmaster set, because in that case
578 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
579 case of MAIL FROM:<> rejection from other early rejections (e.g.
580 rejection of HELO). When verifying a recipient using a non-null MAIL
581 address, the cache is ignored if it shows MAIL FROM:<> rejection.
582 Whatever the result of the callout, the value of the domain cache is
583 left unchanged (for any other kind of callout, getting as far as trying
584 RCPT means that the domain itself is ok).
586 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
589 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
590 manifest itself as EPIPE rather than ECONNECT. When tidying away a
591 session, the daemon ignores ECONNECT errors and logs others; it now
592 ignores EPIPE as well.
594 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
595 (quoted-printable decoding).
597 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
598 later the small subsequent patch to fix an introduced bug.
600 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
602 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
604 PH/08 An error is now given if message_size_limit is specified negative.
606 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
607 to be given (somewhat) arbitrary names.
609 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
612 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
613 miscellaneous code fixes
615 PH/10 Added the log_reject_target ACL modifier to specify where to log
618 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
619 hostname. This is wrong, because it relates to the incoming message (and
620 probably the interface on which it is arriving) and not to the outgoing
621 callout (which could be using a different interface). This has been
622 changed to use the value of the helo_data option from the smtp transport
623 instead - this is what is used when a message is actually being sent. If
624 there is no remote transport (possible with a router that sets up host
625 addresses), $smtp_active_hostname is used.
627 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
628 tweaks were necessary in order to get it to work (see also 21 below):
629 (a) The code assumed that strncpy() returns a negative number on buffer
630 overflow, which isn't the case. Replaced with Exim's string_format()
632 (b) There were several signed/unsigned issues. I just did the minimum
633 hacking in of casts. There is scope for a larger refactoring.
634 (c) The code used strcasecmp() which is not a standard C function.
635 Replaced with Exim's strcmpic() function.
636 (d) The code set only $1; it now sets $auth1 as well.
637 (e) A simple test gave the error "authentication client didn't specify
638 service in request". It would seem that Dovecot has changed its
639 interface. Fortunately there's a specification; I followed it and
640 changed what the client sends and it appears to be working now.
642 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
645 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
646 address A is aliased to B and C, where B exists and C does not. Without
647 -v the output is "A verified" because verification stops after a
648 successful redirection if more than one address is generated. However,
649 with -v the child addresses are also verified. Exim was outputting "A
650 failed to verify" and then showing the successful verification for C,
651 with its parentage. It now outputs "B failed to verify", showing B's
652 parentage before showing the successful verification of C.
654 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
655 look up a TXT record in a specific list after matching in a combined
658 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
659 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
660 they consult the DNS. I had assumed they would set it the way they
661 wanted; and indeed my experiments on Linux seem to show that in some
662 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
663 To be on the safe side, however, I have now made the interface to
664 host_find_byname() similar to host_find_bydns(), with an argument
665 containing the DNS resolver options. The host_find_byname() function now
666 sets these options at its start, just as host_find_bydns() does. The smtp
667 transport options dns_qualify_single and dns_search_parents are passed to
668 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
669 of host_find_byname() use the default settings of RES_DEFNAMES
670 (qualify_single) but not RES_DNSRCH (search_parents).
672 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
673 spool_read_header() do less string testing, by means of a preliminary
674 switch on the second character of optional "-foo" lines. (This is
675 overdue, caused by the large number of possibilities that now exist.
676 Originally there were few.) While I was there, I also converted the
677 str(n)cmp tests so they don't re-test the leading "-" and the first
678 character, in the hope this might squeeze out yet more improvement.
680 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
681 flag allowing group syntax was set by the header_syntax check but not
682 turned off, possible causing trouble later; (2) The flag was not being
683 set at all for the header_verify test, causing "group"-style headers to
684 be rejected. I have now set it in this case, and also caused header_
685 verify to ignore an empty address taken from a group. While doing this, I
686 came across some other cases where the code for allowing group syntax
687 while scanning a header line wasn't quite right (mostly, not resetting
688 the flag correctly in the right place). These bugs could have caused
689 trouble for malformed header lines. I hope it is now all correct.
691 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
692 with the "reply" argument non-NULL. The code, however (which originally
693 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
694 but it didn't always do it. This confused somebody who was copying the
695 code for some other use. I have removed all the tests.
697 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
698 feature that was used to support insecure browsers during the U.S. crypto
699 embargo. It requires special client support, and Exim is probably the
700 only MTA that supported it -- and would never use it because real RSA is
701 always available. This code has been removed, because it had the bad
702 effect of slowing Exim down by computing (never used) parameters for the
703 RSA_EXPORT functionality.
705 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
706 authenticator to fail if there's a tab character in the incoming data
707 (there should never be unless someone is messing about, as it's supposed
708 to be base64-encoded). Also added, on Timo's advice, the "secured" option
709 if the connection is using TLS or if the remote IP is the same as the
710 local IP, and the "valid-client-cert option" if a client certificate has
713 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
714 authenticators. This can be used for authorization after authentication
715 succeeds. (In the case of plaintext, it servers for both authentication
718 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
719 if any retry times were supplied.
721 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
722 connection, where there is no client IP address to check. In this
723 situation, the verify now always succeeds.
725 PH/25 Applied John Jetmore's -Mset patch.
727 PH/26 Added -bem to be like -Mset, but loading a message from a file.
729 PH/27 In a string expansion for a processed (not raw) header when multiple
730 headers of the same name were present, leading whitespace was being
731 removed from all of them, but trailing whitespace was being removed only
732 from the last one. Now trailing whitespace is removed from each header
733 before concatenation. Completely empty headers in a concatenation (as
736 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
737 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
739 PH/29 [Removed. This was a change that I later backed out, and forgot to
740 correct the ChangeLog entry (that I had efficiently created) before
741 committing the later change.]
743 PH/30 Exim was sometimes attempting to deliver messages that had suffered
744 address errors (4xx response to RCPT) over the same connection as other
745 messages routed to the same hosts. Such deliveries are always "forced",
746 so retry times are not inspected. This resulted in far too many retries
747 for the affected addresses. The effect occurred only when there were more
748 hosts than the hosts_max_try setting in the smtp transport when it had
749 the 4xx errors. Those hosts that it had tried were not added to the list
750 of hosts for which the message was waiting, so if all were tried, there
751 was no problem. Two fixes have been applied:
753 (i) If there are any address or message errors in an SMTP delivery, none
754 of the hosts (tried or untried) are now added to the list of hosts
755 for which the message is waiting, so the message should not be a
756 candidate for sending over the same connection that was used for a
757 successful delivery of some other message. This seems entirely
758 reasonable: after all the message is NOT "waiting for some host".
759 This is so "obvious" that I'm not sure why it wasn't done
760 previously. Hope I haven't missed anything, but it can't do any
761 harm, as the worst effect is to miss an optimization.
763 (ii) If, despite (i), such a delivery is accidentally attempted, the
764 routing retry time is respected, so at least it doesn't keep
765 hammering the server.
767 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
768 in ${readsocket because some servers need this prod.
770 PH/32 Added some extra debug output when updating a wait-xxx database.
772 PH/33 The hint "could be header name not terminated by colon", which has been
773 given for certain expansion errors for a long time, was not being given
774 for the ${if def:h_colon_omitted{... case.
776 PH/34 The spec says: "With one important exception, whenever a domain list is
777 being scanned, $domain contains the subject domain." There was at least
778 one case where this was not true.
780 PH/35 The error "getsockname() failed: connection reset by peer" was being
781 written to the panic log as well as the main log, but it isn't really
782 panic-worthy as it just means the connection died rather early on. I have
783 removed the panic log writing for the ECONNRESET error when getsockname()
786 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
787 runs only) independently of the message's sender address. This meant
788 that, if the 4xx error was in fact related to the sender, a different
789 message to the same recipient with a different sender could confuse
790 things. In particualar, this can happen when sending to a greylisting
791 server, but other circumstances could also provoke similar problems.
792 I have changed the default so that the retry time for these errors is now
793 based a combination of the sender and recipient addresses. This change
794 can be overridden by setting address_retry_include_sender=false in the
797 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
798 remote server are returned as part of bounce messages. This was not
799 happening for LMTP over a pipe (the lmtp transport), but now it is the
800 same for both kinds of LMTP.
802 PH/38 Despite being documented as not happening, Exim was rewriting addresses
803 in header lines that were in fact CNAMEs. This is no longer the case.
805 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
806 and queue runs started by the daemon processed all messages. This has
807 been fixed so that -R and -S can now usefully be given with -q<time>.
809 PH/40 Import PCRE release 6.7 (fixes some bugs).
811 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
813 PH/42 Give an error if -q is specified more than once.
815 PH/43 Renamed the variables $interface_address and $interface_port as
816 $received_ip_address and $received_port, to make it clear that these
817 values apply to message reception, and not to the outgoing interface when
818 a message is delivered. (The old names remain recognized, of course.)
820 PH/44 There was no timeout on the connect() call when using a Unix domain
821 socket in the ${readsocket expansion. There now is.
823 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
824 be meaningful with "accept".
826 SC/01 Eximstats V1.43
827 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
829 SC/02 Eximstats V1.44
830 Use a glob alias rather than an array ref in the generated
831 parser. This improves both readability and performance.
833 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
834 Collect SpamAssassin and rejection statistics.
835 Don't display local sender or destination tables unless
836 there is data to show.
837 Added average volumes into the top table text output.
839 SC/04 Eximstats V1.46
840 Collect data on the number of addresses (recipients)
841 as well as the number of messages.
843 SC/05 Eximstats V1.47
844 Added 'Message too big' to the list of mail rejection
845 reasons (thanks to Marco Gaiarin).
847 SC/06 Eximstats V1.48
848 Mainlog lines which have GMT offsets and are too short to
849 have a flag are now skipped.
851 SC/07 Eximstats V1.49 (Alain Williams)
852 Added the -emptyok flag.
854 SC/08 Eximstats V1.50
855 Fixes for obtaining the IP address from reject messages.
857 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
858 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
859 whitesspace changes from 4.64-PH/27
861 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
864 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
865 are found, allow negative numbers in numeric criteria)
867 JJ/06 exipick.20061117.2, added new $message_body_missing variable
869 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
870 to match changes made in 4.64-PH/43
872 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
874 PH/47 Put in an explicit test for a DNS lookup of an address record where the
875 "domain" is actually an IP address, and force a failure. This locks out
876 those revolvers/nameservers that support "A-for-A" lookups, in
877 contravention of the specifications.
879 PH/48 When a host name was looked up from an IP address, and the subsequent
880 forward lookup of the name timed out, the host name was left in
881 $sender_host_name, contrary to the specification.
883 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
884 restricted to single-key lookups, Exim was not diagnosing an error if
885 * or *@ was used with a query-style lookup.
887 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
889 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
890 long ago, but noone interested enough thought of it. Let's just say that
891 the "1.1" means that there are some new functions that weren't there at
892 some point in the past.
894 PH/51 Error processing for expansion failure of helo_data from an smtp
895 transport during callout processing was broken.
897 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
898 tested/used via the -bh/-bhc/-bs options.
900 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
901 bug, fixed in subsequent PCRE releases).
903 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
904 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
906 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
912 SC/01 Use a glob alias rather than an array ref in eximstats generated
913 parser. This improves both readability and performance.
915 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
916 Don't display local sender or destination tables in eximstats unless
917 there is data to show.
918 Added average volumes into the eximstats top table text output.
920 SC/03 Collect data on the number of addresses (recipients) as well
921 as the number of messages in eximstats.
923 TF/01 Correct an error in the documentation for the redirect router. Exim
924 does (usually) call initgroups() when daemonizing.
926 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
927 with consistent privilege compared to when running as a daemon.
929 TF/03 Note in the spec that $authenticated_id is not set for local
930 submissions from trusted users.
932 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
933 Thanks to Dean Brooks <dean@iglou.com> for the patch.
935 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
936 by adding some example configuration directives to the default
937 configuration file. A little bit of work is required to uncomment the
938 directives and define how usernames and passwords are checked, but
939 there is now a framework to start from.
941 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
942 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
943 without this. I don't know how relevant this is to other LDAP libraries.
945 PH/02 Add the verb name to the "unknown ACL verb" error.
947 PH/03 Magnus Holmgren's patch for filter_prepend_home.
949 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
951 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
952 directory not expanded when it should be if an expanded home directory
953 was set for the address (which is overridden by the transport).
955 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
958 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
959 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
960 because it is too late at that time, and has no effect.
962 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
963 security issue with \' (bugzilla #107). I could not use the
964 PQescapeStringConn() function, because it needs a PGconn value as one of
967 PH/08 When testing addresses using -bt, indicate those final addresses that
968 are duplicates that would not cause an additional delivery. At least one
969 person was confused, thinking that -bt output corresponded to deliveries.
970 (Suppressing duplicates isn't a good idea as you lose the information
971 about possibly different redirections that led to the duplicates.)
973 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
974 systems where poll() doesn't work, in particular OS X.
976 PH/10 Added more information to debugging output for retry time not reached.
978 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
979 operations in malware.c.
981 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
984 PH/13 If write_rejectlog was set false when logging was sent to syslog with
985 syslog_duplication set false, log lines that would normally be written
986 both the the main log and to the reject log were not written to syslog at
989 PH/14 In the default configuration, change the use of "message" in ACL warn
990 statements to "add_header".
992 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
993 not followed by a command (e.g. "seen endif").
995 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
996 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
999 PH/17 Added extra conditions to the default value of delay_warning_condition
1003 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1004 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1005 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1008 The Auto-Submitted: and various List- headers are standardised, whereas I
1009 don't think Precedence: ever was.
1011 PH/18 Refactored debugging code in route_finduser() to show more information,
1012 in particular, the error code if getpwnam() issues one.
1014 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1015 This is apparently needed in addition to the PH/07 change above to avoid
1016 any possible encoding problems.
1018 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1019 but not after initializing Perl.
1021 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1022 output them only if debugging. By default they are written stderr,
1023 apparently, which is not desirable.
1025 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1028 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1031 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1033 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1034 authenticated or an ident call has been made. Suppress the default
1035 values for $authenticated_id and $authenticated_sender (but permit -oMai
1036 and -oMas) when testing with -bh.
1038 PH/24 Re-jigged the order of the tests in the default configuration so that the
1039 tests for valid domains and recipients precede the DNS black list and CSA
1040 tests, on the grounds that those ones are more expensive.
1042 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1043 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1044 This bug exists in every version of Exim that I still have, right back to
1047 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1048 However, an attempt to turn on case-sensitivity in a regex key by
1049 including (?-i) didn't work because the subject string was already
1050 lowercased, and the effects were non-intuitive. It turns out that a
1051 one-line patch can be used to allow (?-i) to work as expected.
1057 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1058 other effects) broke the use of negated acl sub-conditions.
1060 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1063 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1064 "Deny" causes Exim to reject the incoming connection with a 554 error.
1065 Unfortunately, if there is a major crisis, such as a disk failure,
1066 tcp-wrappers gives "deny", whereas what one would like would be some
1067 kind of temporary error. A kludge has been added to help with this.
1068 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1069 554 error is used if errno is still zero or contains ENOENT (which occurs
1070 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1073 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1075 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1076 errors. Otherwise a message that provokes a temporary error (when other
1077 messages do not) can cause a whole host to time out.
1079 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1080 addresses were routed directly to files or pipes from a redirect router.
1081 File deliveries just didn't batch; pipe deliveries might have suffered
1084 PH/06 A failure to get a lock for a hints database would erroneously always say
1085 "Failed to get write lock", even when it was really a read lock.
1087 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1088 of 0600. This has been changed to use the value of the lockfile_mode
1089 option (which defaults to 0600).
1091 PH/08 Applied small patch from the Sieve maintainer.
1093 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1094 folder from quota calculations, a direct delivery into this folder messed
1095 up the contents of the maildirsize file. This was because the regex was
1096 used only to exclude .Trash (or whatever) when the size of the mailbox
1097 was calculated. There was no check that a delivery was happening into an
1098 excluded directory. This bug has been fixed by ignoring all quota
1099 processing for deliveries into excluded directories.
1101 PH/10 Added the maildirfolder_create_regex option to appendfile.
1107 PH/01 The code for finding all the local interface addresses on a FreeBSD
1108 system running IPv6 was broken. This may well have applied to all BSD
1109 systems, as well as to others that have similar system calls. The broken
1110 code found IPv4 interfaces correctly, but gave incorrect values for the
1111 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1112 that it would not match correctly against @[] and not recognize the IPv6
1115 PH/02 The ipliteral router was not recognizing addresses of the form user@
1116 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1118 PH/03 Added disable_ipv6.
1120 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1121 decoded form, because it is most often used to construct To: headers
1122 lines in autoreplies, and the decoded form may well be syntactically
1123 invalid. However, $reply_address has leading white space removed, and all
1124 newlines turned into spaces so that the autoreply transport does not
1127 PH/05 If group was specified without a user on a router, and no group or user
1128 was specified on a transport, the group from the router was ignored.
1130 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1131 for visible compile-time settings that can be used to change these
1132 numbers, for those that want even more. Backwards compatibility with old
1133 spool files has been maintained. However, going back to a previous Exim
1134 release will lost any variables that are in spool files.
1136 PH/07 Two small changes when running in the test harness: increase delay when
1137 passing a TCP/IP connection to a new process, in case the original
1138 process has to generate a bounce, and remove special handling of
1139 127.0.0.2 (sic), which is no longer necessary.
1141 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1142 be the same on different OS.
1144 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1147 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1148 whether --show-vars was specified or not
1150 JJ/02 exipick: Added support for new ACL variable spool format introduced
1153 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1154 syntactically invalid From: or Reply-to: line, and a filter used this to
1155 generate an autoreply, and therefore failed to obtain an address for the
1156 autoreply, Exim could try to deliver to a non-existent relative file
1157 name, causing unrelated and misleading errors. What now happens is that
1158 it logs this as a hard delivery error, but does not attempt to create a
1161 PH/11 The exinext utility has a -C option for testing purposes, but although
1162 the given file was scanned by exinext itself; it wasn't being passed on
1163 when Exim was called.
1165 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1166 an end-of-file indication when reading a command response.
1168 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1169 compiled. In many other places in Exim, IPv6 addresses are always
1170 recognized, so I have changed this. It also means that IPv4 domain
1171 literals of the form [IPV4:n.n.n.n] are now always recognized.
1173 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1174 used if the router is not running as root, for example, when verifying at
1175 ACL time, or when using -bh. The debugging output from this situation was
1176 non-existent - all you got was a failure to exec. I have made two
1179 (a) Failures to set uid/gid, the current directory, or a process leader
1180 in a subprocess such as that created by queryprogram now generate
1181 suitable debugging ouput when -d is set.
1183 (b) The queryprogram router detects when it is not running as root,
1184 outputs suitable debugging information if -d is set, and then runs
1185 the subprocess without attempting to change uid/gid.
1187 PH/15 Minor change to Makefile for building test_host (undocumented testing
1190 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1191 additional section of a DNS packet that returns MX or SRV records.
1192 Instead, it always explicitly searches for A/AAAA records. This avoids
1193 major problems that occur when a DNS server includes only records of one
1194 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1195 fixed another bug: if SRV records were looked up and the corresponding
1196 address records were *not* found in the additional section, the port
1197 values from the SRV records were lost.
1199 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1200 using the correct key (the original address) when searching the retry
1201 rules in order to find which one to use for generating the retry hint.
1203 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1204 adding the default one. Similarly, if it contains a Reply-To: header, the
1205 errors_reply_to option, if set, is not used.
1207 PH/19 When calculating a retry time, Exim used to measure the "time since
1208 failure" by looking at the "first failed" field in the retry record. Now
1209 it does not use this if it is later than than the arrival time of the
1210 message. Instead it uses the arrival time. This makes for better
1211 behaviour in cases where some deliveries succeed, thus re-setting the
1212 "first failed" field. An example is a quota failure for a huge message
1213 when small messages continue to be delivered. Without this change, the
1214 "time since failure" will always be short, possible causing more frequent
1215 delivery attempts for the huge message than are intended.
1216 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1218 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1219 $1, $2, $3) because the numerical variables can be reset during some
1220 expansion items (e.g. "match"), thereby losing the authentication data.
1222 PH/21 Make -bV show the size of off_t variables so that the test suite can
1223 decide whether to run tests for quotas > 2G.
1225 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1226 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1227 filecount value is greater than 2G or if a quota value is greater than 2G
1228 on a system where the size of off_t is not greater than 4, a panic error
1231 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1232 never match. The debug and -bh output now contains an explicit error
1233 message indicating a malformed IPv4 address or mask.
1235 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1236 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1237 PH/23 above applies.
1239 PH/25 Do not write to syslog when running in the test harness. The only
1240 occasion when this arises is a failure to open the main or panic logs
1241 (for which there is an explicit test).
1243 PH/26 Added the /no_tell option to "control=freeze".
1245 PH/27 If a host name lookup failed very early in a connection, for example, if
1246 the IP address matched host_lookup and the reverse lookup yielded a name
1247 that did not have a forward lookup, an error message of the form "no IP
1248 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1249 could be logged. Now it outputs the IP address instead of "NULL".
1251 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1252 allows the sender and the authenticated sender to be set when
1253 submitting a message from within Exim. Since child_open_exim() is
1254 documented for local_scan(), the new function should be too.
1256 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1257 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1258 results in an empty string is now treated as unset.
1260 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1262 PH/31 Added sender_verify_fail logging option.
1264 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1265 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1266 tidied the source and removed it altogether.
1268 PH/33 When a queue run was abandoned because the load average was too high, a
1269 log line was always written; now it is written only if the queue_run log
1270 selector is set. In addition, the log line for abandonment now contains
1271 information about the queue run such as the pid. This is always present
1272 in "start" and "stop" lines but was omitted from the "abandon" line.
1274 PH/34 Omit spaces between a header name and the colon in the error message that
1275 is given when verify = headers_syntax fails (if there are lots of them,
1276 the message gets confusing).
1278 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1279 names, as there are now some PTR records that contain slashes. This check
1280 is only to protect against broken name servers that fall over on strange
1281 characters, so the fact that it applies to all lookups doesn't matter.
1283 PH/36 Now that the new test suite is complete, we can remove some of the
1284 special code in Exim that was needed for the old test suite. For example,
1285 sorting DNS records because real resolvers return them in an arbitrary
1286 order. The new test suite's fake resolver always returns records in the
1289 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1290 bounces) except when queue_only is set, to avoid logging races between
1291 the different processes.
1293 PH/38 Panic-die if .include specifies a non-absolute path.
1295 PH/39 A tweak to the "H" retry rule from its user.
1297 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1298 a label. They prevented compilation on older perls.
1300 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1301 a warning to be raised on newish perls.
1303 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1304 on queue. Changes to match documented behaviour of showing count of
1305 messages matching specified criteria.
1307 PH/40 Changed the default ident timeout from 30s to 5s.
1309 PH/41 Added support for the use of login_cap features, on those BSD systems
1310 that have them, for controlling the resources used by pipe deliveries.
1312 PH/42 The content-scanning code uses fopen() to create files in which to put
1313 message data. Previously it was not paying any attention to the mode of
1314 the files. Exim runs with umask(0) because the rest of the code creates
1315 files with open(), and sets the required mode explicitly. Thus, these
1316 files were ending up world-writeable. This was not a big issue, because,
1317 being within the spool directory, they were not world-accessible. I have
1318 created a function called modefopen, which takes an additional mode
1319 argument. It sets umask(777), creates the file, chmods it to the required
1320 mode, then resets the umask. All the relevant calls to fopen() in the
1321 content scanning code have been changed to use this function.
1323 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1324 to 24 hours. This avoids potential overflow problems when processing G
1325 and H retry rules. I suspect nobody ever tinkers with this value.
1327 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1329 PH/45 When the plaintext authenticator is running as a client, the server's
1330 challenges are checked to ensure they are valid base64 strings. By
1331 default, the authentication attempt is cancelled if an invalid string is
1332 received. Setting client_ignore_invalid_base64 true ignores these errors.
1333 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1334 they are received. Thus, the responses can be made to depend on the
1335 challenges. If an invalid string is ignored, an empty string is placed in
1338 PH/46 Messages that are created by the autoreply transport now contains a
1339 References: header, in accordance with RFCs 2822 and 3834.
1341 PH/47 Added authenticated_sender_force to the smtp transport.
1343 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1345 PH/49 Installed latest patch from the Sieve maintainer.
1347 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1348 was also set, the appendfile transport was unnecessarily scanning a
1349 directory of message files (e.g. for maildir delivery) to find the count
1350 of files (along with the size), even though it did not need this
1351 information. It now does the scan only if it needs to find either the
1352 size of the count of files.
1354 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1356 PH/52 Two bugs concerned with error handling when the smtp transport is
1359 (i) Exim was not creating retry information for temporary errors given
1360 for individual recipients after the DATA command when the smtp transport
1361 was used in LMTP mode. This meant that they could be retried too
1362 frequently, and not timed out correctly.
1364 (ii) Exim was setting the flag that allows error details to be returned
1365 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1366 recipients that were returned after the DATA command.
1368 PH/53 This is related to PH/52, but is more general: for any failing address,
1369 when detailed error information was permitted to be returned to the
1370 sender, but the error was temporary, then after the final timeout, only
1371 "retry timeout exceeded" was returned. Now it returns the full error as
1372 well as "retry timeout exceeded".
1374 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1375 do this, and (what is worse) MTAs that accept it.
1377 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1378 will now be deprecated.
1380 PH/56 New os.c-cygwin from the Cygwin maintainer.
1382 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1383 formats (previously only available in exim formats via -bpr, -bpru,
1384 and -bpra. Now also available in native and exiqgrep formats)
1386 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1387 with very large, slow to parse queues
1389 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1391 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1393 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1394 responses to authentication challenges, though it was showing the
1395 challenges; (ii) I've removed the CR characters from the debug output for
1398 PH/58 Allow for the insertion of a newline as well as a space when a string
1399 is turned into more than one encoded-word during RFC 2047 encoding. The
1400 Sieve code now uses this.
1402 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1403 data_4xx, lost_connection, tls_required.
1405 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1406 message was being sent as an SMTP response.
1408 PH/61 Add -l and -k options to exicyclog.
1410 PH/62 When verifying, if an address was redirected to one new address, so that
1411 verification continued, and the new address failed or deferred after
1412 having set something in $address_data, the value of $address_data was not
1413 passed back to the ACL. This was different to the case when no
1414 redirection occurred. The value is now passed back in both cases.
1416 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1417 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1418 use login_cap.h, so on its own it isn't the distinguishing feature. The
1419 new name refers directly to the setclassresources() function.
1421 PH/65 Added configuration files for NetBSD3.
1423 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1425 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1426 is preferred over IPv4.
1428 PH/68 The bounce_return_message and bounce_return_body options were not being
1429 honoured for bounces generated during the reception of non-SMTP messages.
1430 In particular, this applied to messages rejected by the ACL. This bug has
1431 been fixed. However, if bounce_return_message is true and bounce_return_
1432 body is false, the headers that are returned for a non-SMTP message
1433 include only those that have been read before the error was detected.
1434 (In the case of an ACL rejection, they have all been read.)
1436 PH/69 The HTML version of the specification is now built in a directory called
1437 spec_html instead of spec.html, because the latter looks like a path with
1438 a MIME-type, and this confuses some software.
1440 PH/70 Catch two compiler warnings in sieve.c.
1442 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1443 function verify_get_ident() calls ip_connect() to connect a socket, but
1444 if the "connect()" function timed out, ip_connect() used to close the
1445 socket. However, verify_get_ident() also closes the socket later, and in
1446 between Exim writes to the log, which may get opened at this point. When
1447 the socket was closed in ip_connect(), the log could get the same file
1448 descriptor number as the socket. This naturally causes chaos. The fix is
1449 not to close the socket in ip_connect(); the socket should be closed by
1450 the function that creates it. There was only one place in the code where
1451 this was missing, in the iplookup router, which I don't think anybody now
1452 uses, but I've fixed it anyway.
1454 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1455 well as to direct DNS lookups. Otherwise the handling of names in host
1456 lists is inconsistent and therefore confusing.
1462 PH/01 Two changes to the default runtime configuration:
1464 (1) Move the checks for relay_from_hosts and authenticated clients from
1465 after to before the (commented out) DNS black list checks.
1467 (2) Add control=submission to the relay_from_hosts and authenticated
1468 clients checks, on the grounds that messages accepted by these
1469 statements are most likely to be submissions.
1471 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1473 (1) Generate an error if the third argument for the ${prvs expansion is
1476 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1479 (3) Reset the variables that are obtained from the first argument of
1480 ${prvscheck and used in the second argument before leaving the code,
1481 because their memory is reclaimed, so using them afterwards may do
1484 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1485 one (it's much easier than Tom thought :-).
1487 (5) Because of (4), we can now allow for the use of $prvscheck_result
1488 inside the third argument.
1490 PH/03 For some reason, the default setting of PATH when running a command from
1491 a pipe transport was just "/usr/bin". I have changed it to
1494 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1495 anything to be listed in the output from -bV.
1497 PH/05 When a filter generated an autoreply, the entire To: header line was
1498 quoted in the delivery log line, like this:
1500 => >A.N.Other <ano@some.domain> <original@ddress> ...
1502 This has been changed so that it extracts the operative address. There
1503 may be more than one such address. If so, they are comma-separated, like
1506 => >ano@some.domain,ona@other.domain <original@ddress> ...
1508 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1509 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1510 not being looked up in the rDNS to get a host name, Exim was showing the
1511 IP address twice in Received: lines, even though the IP addresses were
1512 identical. For example:
1514 Received: from [1.2.3.4] (helo=[1.2.3.4])
1516 However, if the real host name was known, it was omitting the HELO data
1517 if it matched the actual IP address. This has been tidied up so that it
1518 doesn't show the same IP address twice.
1520 PH/07 When both +timestamp and +memory debugging was on, the value given by
1521 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1522 called by the debug printing, thereby overwriting the timestamp buffer.
1523 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1525 PH/08 When the original message was included in an autoreply transport, it
1526 always said "this is a copy of the message, including all the headers",
1527 even if body_only or headers_only was set. It now gives an appropriate
1530 PH/09 Applied a patch from the Sieve maintainer which:
1532 o fixes some comments
1533 o adds the (disabled) notify extension core
1534 o adds some debug output for the result of if/elsif tests
1535 o points to the current vacation draft in the documentation
1536 and documents the missing references header update
1540 o fixes a bug in processing the envelope test (when testing
1541 multiple envelope elements, the last element determinted the
1544 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1545 Electronic Mail") by including:
1547 Auto-submitted: auto-generated
1549 in the messages that it generates (bounce messages and others, such as
1550 warnings). In the case of bounce messages for non-SMTP mesages, there was
1551 also a typo: it was using "Auto_submitted" (underscore instead of
1552 hyphen). Since every message generated by Exim is necessarily in response
1553 to another message, thes have all been changed to:
1555 Auto-Submitted: auto-replied
1557 in accordance with these statements in the RFC:
1559 The auto-replied keyword:
1561 - SHOULD be used on messages sent in direct response to another
1562 message by an automatic process,
1564 - MUST NOT be used on manually-generated messages,
1566 - MAY be used on Delivery Status Notifications (DSNs) and Message
1567 Disposition Notifications (MDNs),
1569 - MUST NOT be used on messages generated by automatic or periodic
1570 processes, except for messages which are automatic responses to
1573 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1574 to the default Received: header definition.
1576 PH/12 Added log selector acl_warn_skipped (default on).
1578 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1579 variables because (a) they are in the wrong storage pool and (b) even if
1580 they were copied, it wouldn't work properly because of the caching.
1582 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1583 checking when decoding. Apparently there are clients that generate
1584 overlong encoded strings. Why am I not surprised?
1586 PH/15 If the first argument of "${if match_address" was not empty, but did not
1587 contain an "@" character, Exim crashed. Now it writes a panic log message
1588 and treats the condition as false.
1590 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1592 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1593 extension "envelope-auth". The code is finished and in agreement with
1594 other implementations, but there is no documentation so far and in fact,
1595 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1596 not changing the active code.
1598 Print executed "if" and "elsif" statements when debugging is used. This
1599 helps a great deal to understand what a filter does.
1601 Document more things not specified clearly in RFC3028. I had all this
1602 sorted out, when out of a sudden new issues came to my mind. Oops."
1604 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1607 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1608 canonical form (as documented). However, after a host name lookup from
1609 the IP address, check_host() was doing a simple string comparison with
1610 addresses acquired from the DNS when checking that the found name did
1611 have the original IP as one of its addresses. Since any found IPv6
1612 addresses are likely to be in abbreviated form, the comparison could
1613 fail. Luckily, there already exists a function for doing the comparison
1614 by converting both addresses to binary, so now that is used instead of
1615 the text comparison.
1617 PH/20 There was another similar case to PH/19, when a complete host name was
1618 given in a host list; looking up its IP address could give an abbreviated
1619 form, whereas the current host's name might or might not be abbreviated.
1620 The same fix has been applied.
1626 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1627 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1630 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1631 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1633 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1635 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1636 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1637 Also, for "Auto-Submitted", treat anything other than "no" as
1638 non-personal, in accordance with RFC 3834. (Previously it treated
1639 anything starting "auto-" as non-personal.)
1641 TF/01 The control=submission/name=... option had a problem with syntax
1642 errors if the name included a slash character. The /name= option
1643 now slurps the rest of the string, so it can include any characters
1644 but it must come last in the list of options (after /sender_retain
1647 PH/05 Some modifications to the interface to the fake nameserver for the new
1655 TK/01 Added the "success_on_redirect" address verification option. See
1656 NewStuff for rationale and an example.
1658 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1660 PH/02 Patch to exigrep to allow it to work on syslog lines.
1662 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1663 fread() to skip over the body file's header line, because in Cygwin the
1664 header line is locked and is inaccessible.
1666 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1667 co-exist for some time) to make it clear that it is the Exim ID that is
1668 referenced, not the Message-ID: header line.
1670 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1671 string_format() function, because snprintf() does not exist on all
1674 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1675 ${stat: expansion item.
1677 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1678 protocol synchronization error", to keep the pedants happy.
1680 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1681 well as for IRIX systems, when gcc is being used. See the host.c source
1684 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1686 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1689 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1690 transport and to the smtp transport in LMTP mode.
1692 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1694 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1695 run a filter in a subprocess. This could lead to confusion in subsequent
1696 lookups in the parent process. There should also be a search_tidyup() at
1697 the end of the subprocess.
1699 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1700 only if the host matched helo_try_verify_hosts, which caused the
1701 verification to occur when the EHLO/HELO command was issued. The ACL just
1702 tested the remembered result. Now, if a previous verification attempt has
1703 not happened, "verify = helo" does it there and then.
1705 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1707 TK/03 Fix log output including CR from clamd.
1709 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1710 exist provoked a memory error which could cause a segfault.
1712 PH/15 Installed PCRE 6.2
1714 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1716 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1717 of the problem. Specifically, suggested +O2 rather than +O1 for the
1720 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1722 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1723 record, the port was not correctly logged when the outgoing_port log
1724 selector was set (it logged the transort's default port).
1726 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1727 fallback_hosts, and "hosts" in the smtp transport.
1729 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1730 host errors such as "Connection refused".
1732 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1733 authentication with radiusclient 0.4.9:
1735 - Error returned from rc_read_config was caught wrongly
1736 - Username/password not passed on to radius server due to wrong length.
1738 The presumption is that some radiusclient API changes for 4.51/PH/17
1739 were not taken care of correctly. The code is still untested by me (my
1740 Linux distribution still has 0.3.2 of radiusclient), but it was
1741 contributed by a Radius user.
1743 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1744 expanding the "port" option of the smtp transport.
1746 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1747 while reading a MIME header. Thanks to Tom Hughes for a patch.
1749 PH/24 Include config.h inside local_scan.h so that configuration settings are
1752 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1753 that in an ACL for RCPT (for example), you can examine exactly what was
1756 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1757 commands, but it was not correctly comparing the address with the actual
1758 client host address. Thus, it would show the EHLO address in Received:
1759 header lines when this was not necessary.
1761 PH/27 Added the % operator to ${eval:}.
1763 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1764 it should be ignoring failures (because with -C, for example, it has lost
1765 privilege). It wasn't ignoring creation failures other than "already
1768 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1771 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1772 because an input line was too long, either on its own, or by virtue of
1773 too many continuations, the temporary file was not being removed, and the
1774 return code was incorrect.
1776 PH/31 Missing "BOOL" in function definition in filtertest.c.
1778 PH/32 Applied Sieve patches from the maintainer.
1780 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1782 PH/33 Added "verify = not_blind".
1784 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1785 Local/Makefile (with some defaults set). These are used in built scripts
1786 such as exicyclog, but they have never been used in the exim_install
1787 script (though there are many overriding facilities there). I have
1788 arranged that the exim_install script now takes note of these two
1791 PH/35 Installed configuration files for Dragonfly.
1793 PH/36 When a locally submitted message by a trusted user did not contain a
1794 From: header, and the sender address was obtained from -f or from an SMTP
1795 MAIL command, and the trusted user did not use -F to supply a sender
1796 name, $originator_name was incorrectly used when constructing a From:
1797 header. Furthermore, $originator_name was used for submission mode
1798 messages from external hosts without From: headers in a similar way,
1799 which is clearly wrong.
1801 PH/37 Added control=suppress_local_fixups.
1803 PH/38 When log_selector = +received_sender was set, and the addition of the
1804 sender made the log line's construction buffer exactly full, or one byte
1805 less than full, an overflow happened when the terminating "\n" was
1808 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1809 when the result of a list match is failure because a DNS lookup failed.
1811 PH/40 RM_COMMAND is now used in the building process.
1813 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1814 the "build-* directories that it finds.
1816 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1817 address in a domain literal was a prefix of an interface address.
1819 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1820 when verifying a sender address, unless rewrite_headers is false.
1822 PH/44 Wrote a long comment about why errors_to addresses are verified as
1823 recipients, not senders.
1825 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1826 the ratelimit ACL was added.
1828 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1830 PH/46 Added extra information about PostgreSQL errors to the error string.
1832 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1833 avoiding the need to install special zones in a real server. This is
1834 backwards compatible; if it can't find the fake resolver, it drops back.
1835 Thus, both old and new test suites can be run.
1837 TF/02 Added util/ratelimit.pl
1839 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1840 clock is set back in time.
1842 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1843 Candler <B.Candler@pobox.com>.
1845 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1846 OK for addresses that are the result of redirections.
1848 PH/48 A number of further additions for the benefit of the new test suite,
1849 including a fake gethostbyname() that interfaces to the fake DNS resolver
1852 TF/06 The fix for widen_domains has also been applied to qualify_single and
1853 search_parents which are the other dnslookup options that can cause
1856 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1859 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1861 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1862 service" if it encountered an MX record with an empty target hostname.
1863 The message is now "an MX or SRV record indicated no SMTP service".
1865 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1866 if the DNS of the sending site is misconfigured. This is quite a
1867 common situation. This change restores the behaviour of treating a
1868 helo verification defer as a failure.
1870 PH/51 If self=fail was set on a router, the bounce message did not include the
1871 actual error message.
1877 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1879 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1880 command itself ended in error, the underlying message about the transport
1881 filter timeout was being overwritten with the pipe command error. Now the
1882 underlying error message should be appended to the second error message.
1884 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1886 PH/02 Reduce the amount of output that "make" produces by default. Full output
1887 can still be requested.
1889 PH/03 The warning log line about a condition test deferring for a "warn" verb
1890 was being output only once per connection, rather than after each
1891 occurrence (because it was using the same function as for successful
1892 "warn" verbs). This seems wrong, so I have changed it.
1894 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1895 it should not have, which might have caused a crash in the right
1896 circumstances, but probably never did.
1898 PH/04 Installed a modified version of Tony Finch's patch to make submission
1899 mode fix the return path as well as the Sender: header line, and to
1900 add a /name= option so that you can make the user's friendly name appear
1903 TF/03 Added the control = fakedefer ACL modifier.
1905 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1906 Mark Lowes for thorough testing.
1908 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1910 TK/03 Merged latest SRS patch from Miles Wilton.
1912 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1913 with the definition in sysexits.h (which is #included earlier).
1914 Fortunately, Exim does not actually use EX_OK. The code used to try to
1915 preserve the sysexits.h value, by assumimg that macro definitions were
1916 scanned for macro replacements. I have been disabused of this notion,
1917 so now the code just undefines EX_OK before #including unistd.h.
1919 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1920 in the smtp transport. When a block could not be written in a single
1921 write() function, the timeout was being re-applied to each part-write.
1922 This seems wrong - if the receiver was accepting one byte at a time it
1923 would take for ever. The timeout is now adjusted when this happens. It
1924 doesn't have to be particularly precise.
1926 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1927 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1929 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1930 without a domain if the check to <postmaster@domain> fails.
1932 SC/01 Eximstats: added -xls and the ability to specify output files
1933 (patch written by Frank Heydlauf).
1935 SC/02 Eximstats: use FileHandles for outputing results.
1937 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1939 SC/04 Eximstats: fixed display of large numbers with -nvr option
1941 SC/05 Eximstats: fixed merging of reports with empty tables.
1943 SC/06 Eximstats: added the -include_original_destination flag
1945 SC/07 Eximstats: removed tabs and trailing whitespace.
1947 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1949 TK/06 MBOX spool code: Add real "From " MBOX separator line
1950 so the .eml file is really in mbox format (even though
1951 most programs do not really care). Patch from Alex Miller.
1953 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1954 The latter is generated from $received_to and is only set if the
1955 message has one envelope recipient. SA can use these headers,
1956 obviously out-of-the-box. Patch from Alex Miller.
1958 PH/08 The ${def test on a variable was returning false if the variable's
1959 value was "0", contrary to what the specification has always said!
1960 The result should be true unless the variable is empty.
1962 PH/09 The syntax error of a character other than { following "${if
1963 def:variable_name" (after optional whitespace) was not being diagnosed.
1964 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1965 accidental colon was present, for example, could give incorrect results.
1967 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1968 result is used (not including appendfile, where other changes are about
1971 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1972 This involved changing a lot of size variables from int to off_t. It
1973 should work with maildirs and everything.
1975 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1976 spamd dying while we are connected to it.
1978 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1981 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1982 function for BDB 4.3.
1984 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1986 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1987 the sources for PCRE have been re-arranged and more files are now
1990 PH/15 The code I had for printing potentially long long variables in PH/11
1991 above was not the best (it lost precision). The length of off_t variables
1992 is now inspected at build time, and an appropriate printing format (%ld
1993 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1994 to be "long long int" or "long int". This is needed for the internal
1995 formatting function string_vformat().
1997 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1998 the configuration file to be ":syslog", then the script "guesses" where
1999 the logs files are, rather than using the compiled in default. In our
2000 case the guess is not the same as the compiled default, so the script
2001 suddenly stopped working when I started to use syslog. The patch checks
2002 to see if log_file_path is "". If so, it attempts to read it from exim
2003 with no configuration file to get the compiled in version, before it
2004 falls back to the previous guessing code."
2006 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2007 implementing BATV in an Exim configuration. See NewStuff for the gory
2010 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2011 Makefile that are specific to HP-UX.
2013 PH/18 If the "use_postmaster" option was set for a recipient callout together
2014 with the "random" option, the postmaster address was used as the MAIL
2015 FROM address for the random test, but not for the subsequent recipient
2016 test. It is now used for both.
2018 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2019 patch removes a few documentation additions to RFC 3028, because the
2020 latest draft now contains them. It adds the new en;ascii-case comparator
2021 and a new error check for 8bit text in MIME parts. Comparator and
2022 require names are now matched exactly. I enabled the subaddress
2023 extension, but it is not well tested yet (read: it works for me)."
2025 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2026 rework some of the code of TK/09 above to avoid the hardwired use of
2027 "%lld" and "long long". Replaced the call to snprintf() with a call to
2030 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2031 records point to non-existent hosts", "retry timeout exceeded", and
2032 "retry time not reached for any host after a long failure period".
2034 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2035 experimental DomainKeys support:
2037 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2038 (2) On an error such as an illegally used "control", the wrong name for
2039 the control was given.
2041 These problems did NOT occur unless DomainKeys support was compiled.
2043 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2045 PH/24 Added ${if match_ip condition.
2047 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2048 databases so that it will be absolutely obvious if a crash occurs in the
2049 DB library. This is a regular occurrence (often caused by mis-matched
2052 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2053 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2054 picked up on a user's system that detects such things. There doesn't seem
2055 to be a gcc warning option for this - only an attribute that has to be
2056 put on the function's prototype. It seems that in Fedora Core 4 they have
2057 set this on a number of new functions. No doubt there will be more in due
2060 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2061 specify a transport. However, if an address that was verified by such a
2062 router was the subject of a callout, Exim crashed because it tried to
2063 read the rcpt_include_affixes from the non-existent transport. Now it
2064 just assumes that the setting of that option is false. This bug was
2065 introduced by 4.51/PH/31.
2067 PH/28 Changed -d+all to exclude +memory, because that information is very
2068 rarely of interest, but it makes the output a lot bigger. People tend to
2069 do -d+all out of habit.
2071 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2072 code in os-type was giving problems when libc.so lives in lib64, like on
2075 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2076 aren't the modern standard, and it seems that some systems' include files
2077 don't always have them. Exim was already checking for some of the newer
2078 ones like T_AAAA, and defining it itself. I've added checks for all the
2079 record types that Exim uses.
2081 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2082 not automatically generating a new one, as it is supposed to. This
2083 prevented TLS from working. If the file did exist, but contained invalid
2084 data, a new version was generated, as expected. It was only the case of a
2085 non-existent file that was broken.
2087 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2088 with a change in libdomainkeys > 0.64.
2090 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2091 from DNS. If the selector record carries the flag, it now has
2092 precedence over the domain-wide flag.
2094 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2096 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2097 the use of a port name that isn't defined in /etc/services) occurred, the
2098 message was deferred as in a normal delivery, and thus remained on the
2099 spool, instead of being failed because of the mua_wrapper setting. This
2100 is now fixed, and I tidied up some of the mua_wrapper messages at the
2103 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2104 the messages in a hash of arrays rather than using individual hashes.
2105 This is a bit cleaner and results in dramatic memory savings, albeit
2106 at a slight CPU cost.
2108 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2109 as requested by Marc Sherman.
2111 SC/10 Eximstats: added histograms for user specified patterns as requested
2114 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2116 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2117 fopen() in the content-scanning modules that did not already have it.
2123 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2124 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2126 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2128 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2130 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2131 to test Sieve filters that use "vacation".
2133 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2134 that changes the way the GnuTLS parameters are stored in the cache file.
2135 The new format can be generated externally. For backward compatibility,
2136 if the data in the cache doesn't make sense, Exim assumes it has read an
2137 old-format file, and it generates new data and writes a new file. This
2138 means that you can't go back to an older release without removing the
2141 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2142 work if there are any delivery delays because "one_time" forces the
2143 parent to be marked "delivered", so its unseen clone is never tried
2144 again. For this reason, Exim now forbids the simultaneous setting of
2147 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2148 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2149 entry if you want to know the details. The fix, however, neglected to
2150 consider the case when local delivery batching is involved. The test for
2151 "previously delivered" was not happening when checking to see if an
2152 address could be batched with a previous (undelivered) one; under
2153 certain circumstances this could lead to multiple deliveries to the same
2156 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2157 in its include files, and this causes problems building Exim.
2159 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2160 header_syntax) but Exim was just ignoring anything given after a slash.
2161 In particular, this caused confusion with an attempt to use "verify =
2162 reverse_host_lookup/defer_ok". An error is now given when options are
2163 supplied for verify items that do not have them. (Maybe reverse_host_
2164 lookup should have a defer_ok option, but that's a different point.)
2166 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2167 defined by RFC 821) to 2048, because there were problems with some AUTH
2168 commands, and RFC 1869 says the size should be increased for extended
2169 SMTP commands that take arguments.
2171 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2174 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2175 "unknown" error; now it says that the functionality isn't in the binary.
2177 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2178 an address' error message when a string expansion fails (syntax or
2179 whatever). Otherwise the password may appear in the log. Following change
2180 PH/42 below, there is no longer a chance of it appearing in a bounce
2183 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2185 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2186 of its addresses was ever tried. (Bugzilla bug #2.)
2188 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2189 the result incorrectly in the debug output. (It correctly added a newline
2190 to what was transported.)
2192 TF/01 Added $received_time.
2194 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2195 commented out examples of how to interface to a virus scanner and to
2196 SpamAssassin. Also added commented examples of av_scanner and
2197 spamd_address settings.
2199 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2200 and controls are allowed in which ACLs. There were a couple of minor
2201 errors. Some of the entries in the conditions table (which is a table of
2202 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2203 negation of where the condition IS allowed.
2205 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2207 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2208 header file does not have a version number, so I've had to invent a new
2209 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2210 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2211 radiusclient), but it was contributed by a Radius user.
2213 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2214 files or directories, for OpenSSL.
2216 PH/19 When an Exim process that is running as root has to create an Exim log
2217 file, it does so in a subprocess that runs as exim:exim so as to get the
2218 ownership right at creation (otherwise, other Exim processes might see
2219 the file with the wrong ownership). There was no test for failure of this
2220 fork() call, which would lead to the process getting stuck as it waited
2221 for a non-existent subprocess. Forks do occasionally fail when resources
2222 run out. I reviewed all the other calls to fork(); they all seem to check
2225 PH/20 When checking for unexpected SMTP input at connect time (before writing
2226 the banner), Exim was not dealing correctly with a non-positive return
2227 from the read() function. If the client had disconnected by this time,
2228 the result was a log entry for a synchronization error with an empty
2229 string after "input=" when read() returned zero. If read() returned -1
2230 (an event I could not check), uninitialized data bytes were printed.
2231 There were reports of junk text (parts of files, etc) appearing after
2234 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2236 PH/22 Added support for macro redefinition, and (re)definition in between
2237 driver and ACL definitions.
2239 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2240 forgetting to use the resulting value; it was using the unexpanded value.
2242 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2243 hadn't been configured. The fix is from Juergen Kreileder, who
2244 understands it better than I do:
2246 "Here's what I see happening with three configured cyrus_sasl
2247 authenticators configured (plain, login, cram-md5):
2249 On startup auth_cyrus_sasl_init() gets called for each of these.
2250 This means three calls to sasl_listmech() without a specified mech_list.
2251 => SASL tests which mechs of all available mechs actually work
2252 => three warnings about OTP not working
2253 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2255 With the patch, sasl_listmech() also gets called three times. But now
2256 SASL's mech_list option is set to the server_mech specified in the the
2257 authenticator. Or in other words, the answer from sasl_listmech()
2258 gets limited to just the mech you're testing for (which is different
2260 => the return list contains just 'plain' or 'login', 'cram-md5' or
2261 nothing depending on the value of ob->server_mech.
2263 I've just tested the patch: Authentication still works fine,
2264 unavailable mechs specified in the exim configuration are still
2265 caught, and the auth.log warnings about OTP are gone."
2267 PH/25 When debugging is enabled, the contents of the command line are added
2268 to the debugging output, even when log_selector=+arguments is not
2271 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2272 answer is "GNU", and only if the return is "GNU/something" is the answer
2275 PH/27 $acl_verify_message is now set immediately after the failure of a
2276 verification in an ACL, and so is available in subsequent modifiers. In
2277 particular, the message can be preserved by coding like this:
2279 warn !verify = sender
2280 set acl_m0 = $acl_verify_message
2282 Previously, $acl_verify_message was set only while expanding "message"
2283 and "log_message" when a very denied access.
2285 PH/28 Modified OS/os.c-Linux with
2287 -#ifndef OS_LOAD_AVERAGE
2288 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2290 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2291 nomenclature these days.)
2293 PH/29 Installed patch from the Sieve maintainer that adds the options
2294 sieve_useraddress and sieve_subaddress to the redirect router.
2296 PH/30 In these circumstances:
2297 . Two addresses routed to the same list of hosts;
2298 . First host does not offer TLS;
2299 . First host accepts first address;
2300 . First host gives temporary error to second address;
2301 . Second host offers TLS and a TLS session is established;
2302 . Second host accepts second address.
2303 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2304 and peerdn, if requested) that were in fact used only for the second
2307 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2308 attention to any local part prefix or suffix that was matched by the
2309 router that accepted the address. It now behaves in the same way as it
2310 does for delivery: the affixes are removed from the local part unless
2311 rcpt_include_affixes is set on the transport.
2313 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2314 timeout during the DATA phase of an incoming message.
2316 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2317 applied a patch sanctioned by the Sieve maintainer.
2319 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2320 the uid or gid is negative. A case of a negative gid caused this to be
2321 noticed. The fix allows for either to be negative.
2323 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2324 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2327 PH/36 The change PH/12 above was broken. Fixed it.
2329 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2330 the grounds that routing the same address twice would always produce the
2331 same answer. This might have been true once, but it is certainly no
2332 longer true now. Routing a child address may depend on the previous
2333 routing that produced that child. Some complicated redirection strategies
2334 went wrong when messages had multiple recipients, and made Exim's
2335 behaviour dependent on the order in which the addresses were given.
2337 I have moved the duplicate checking until after the routing is complete.
2338 Exim scans the addresses that are assigned to local and remote
2339 transports, and removes any duplicates. This means that more work will be
2340 done, as duplicates will always all be routed, but duplicates are
2341 presumably rare, so I don't expect this is of any significance.
2343 For deliveries to pipes, files, and autoreplies, the duplicate checking
2344 still happens during the routing process, since they are not going to be
2347 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2348 It corrects a timeout issue with spamd. This is Ian's comment: "The
2349 background is that sometimes spamd either never reads data from a
2350 connection it has accepted, or it never writes response data. The exiscan
2351 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2352 blindly assumes that writes won't block so it may never time out."
2354 PH/39 Allow G after quota size as well as K and M.
2356 PH/40 The value set for $authenticated_id in an authenticator may not contain
2357 binary zeroes or newlines because the value is written to log lines and
2358 to spool files. There was no check on this. Now the value is run through
2359 the string_printing() function so that such characters are converted to
2360 printable escape sequences.
2362 PH/41 $message_linecount is a new variable that contains the total number of
2363 lines in the message. Compare $body_linecount, which is the count for the
2366 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2367 bounce and delay warning messages, except in certain special cases, which
2370 (a) An SMTP error message from a remote host;
2371 (b) A message specified in a :fail: redirection;
2372 (c) A message specified in a "fail" command in a system filter;
2373 (d) A message specified in a FAIL return from the queryprogram router;
2374 (e) A message specified by the cannot_route_message router option.
2376 In these cases only, Exim does include the error details in bounce and
2377 warning messages. There are also a few cases where bland messages such
2378 as "unrouteable address" or "local delivery error" are given.
2380 PH/43 $value is now also set for the "else" part of a ${run expansion.
2382 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2383 being worked on, but at least Exim now implements the latest version to
2386 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2387 process to complete was treated as a delivery failure, a timeout while
2388 writing the message to the pipe was logged, but erroneously treated as a
2389 successful delivery. Such timeouts include transport filter timeouts. For
2390 consistency with the overall process timeout, these timeouts are now
2391 treated as errors, giving rise to delivery failures by default. However,
2392 there is now a new Boolean option for the pipe transport called
2393 timeout_defer, which, if set TRUE, converts the failures into defers for
2394 both kinds of timeout. A transport filter timeout is now identified in
2397 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2398 systems where "make" and "gmake" are different, calling "gmake" at top
2399 level broke things. I've arranged for the value of $(MAKE) to be passed
2400 from the Makefile to this script so that it can call the same version of
2404 A note about Exim versions 4.44 and 4.50
2405 ----------------------------------------
2407 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2408 changes of various kinds. As a consequence, a big documentation update was
2409 needed. This delayed the release for rather longer than seemed good, especially
2410 in the light of a couple of (minor) security issues. Therefore, the changes
2411 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2412 release. So 4.44 and 4.50 are in effect two different branches that both start
2415 I have left the 4.50 change log unchanged; it contains all the changes since
2416 4.43. The change log for 4.44 is below; many of its items are identical to
2417 those for 4.50. This seems to be the most sensible way to preserve the
2418 historical information.
2424 1. Minor wording change to the doc/README.SIEVE file.
2426 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2427 computation of the current number of files was incorrect.
2429 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2430 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2433 4. Give more explanation in the error message when the command for a transport
2434 filter fails to execute.
2436 5. There are several places where Exim runs a non-Exim command in a
2437 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2438 was being done only for the command run by the queryprogram router. It is
2439 now done for all such subprocesses. The other cases are: ${run, transport
2440 filters, and the commands run by the lmtp and pipe transports.
2442 6. Added CONFIGURE_GROUP build-time option.
2444 7. Some older OS have a limit of 256 on the maximum number of file
2445 descriptors. Exim was using setrlimit() to set 1000 as a large value
2446 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2447 systems. I've change it so that if it can't get 1000, it tries for 256.
2449 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2450 was an oversight, and furthermore, ever since the addition of extra
2451 controls (e.g. 4.43/32), the checks on when to allow different forms of
2452 "control" were broken. There should now be diagnostics for all cases when a
2453 control that does not make sense is encountered.
2455 9. Added the /retain_sender option to "control=submission".
2457 10. $recipients is now available in the predata ACL (oversight).
2459 11. Tidy the search cache before the fork to do a delivery from a message
2460 received from the command line. Otherwise the child will trigger a lookup
2461 failure and thereby defer the delivery if it tries to use (for example) a
2462 cached ldap connection that the parent has called unbind on.
2464 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2465 of $address_data from the recipient verification was clobbered by the
2466 sender verification.
2468 13. The value of address_data from a sender verification is now available in
2469 $sender_address_data in subsequent conditions in the ACL statement.
2471 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2473 15. Added a new option "connect=<time>" to callout options, to set a different
2476 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2477 was its contents. (It was OK if the option was not defined at all.)
2479 17. A "Completed" log line is now written for messages that are removed from
2480 the spool by the -Mrm option.
2482 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2483 information about exactly what failed.
2485 19. Added -dd to debug only the daemon process.
2487 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2488 handles timeouts, both on the server side and network timeouts. Renamed the
2489 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2491 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2492 It is now set to "smtps".
2494 22. $host_address is now set to the target address during the checking of
2495 ignore_target_hosts.
2497 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2498 being passed; this would have caused $sender_host_name to have been used if
2499 matching the list had actually called for a host name (not very likely,
2500 since this list is usually IP addresses). A host name is now passed as
2503 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2504 code that specifies a non-restarting handler (typically sigaction() in
2505 modern systems) in an attempt to fix a rare and obscure crash bug.
2507 25. Narrowed the window for a race in the daemon that could cause it to ignore
2508 SIGCHLD signals. This is not a major problem, because they are used only to
2509 wake it up if nothing else does.
2511 26. A malformed maildirsize file could cause Exim to calculate negative values
2512 for the mailbox size or file count. Odd effects could occur as a result.
2513 The maildirsize information is now recalculated if the size or filecount
2516 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2517 support for a long time. Removed HAVE_SYS_VFS_H.
2519 28. Installed the latest version of exipick from John Jetmore.
2521 29. In an address list, if the pattern was not a regular expression, an empty
2522 subject address (from a bounce message) matched only if the pattern was an
2523 empty string. Non-empty patterns were not even tested. This was the wrong
2524 because it is perfectly reasonable to use an empty address as part of a
2525 database query. An empty address is now tested by patterns that are
2526 lookups. However, all the other forms of pattern expect the subject to
2527 contain a local part and a domain, and therefore, for them, an empty
2528 address still always fails if the pattern is not itself empty.
2530 30. Exim went into a mad DNS loop when attempting to do a callout where the
2531 host was specified on an smtp transport, and looking it up yielded more
2532 than one IP address.
2534 31. Re-factored the code for checking spool and log partition space into a
2535 function that finds that data and another that does the check. The former
2536 is then used to implement four new variables: $spool_space, $log_space,
2537 $spool_inodes, and $log_inodes.
2539 32. The RFC2047 encoding function was originally intended for short strings
2540 such as real names; it was not keeping to the 75-character limit for
2541 encoded words that the RFC imposes. It now respects the limit, and
2542 generates multiple encoded words if necessary. To be on the safe side, I
2543 have increased the buffer size for the ${rfc2047: expansion operator from
2546 33. It is now permitted to omit both strings after an "if" condition; if the
2547 condition is true, the result is "true". As before, when the second string
2548 is omitted, a false condition yields an empty string. This makes it less
2549 cumbersome to write custom ACL and router conditions.
2551 34. Failure to deliver a bounce message always caused it to be frozen, even if
2552 there was an errors_to setting on the router. The errors_to setting is now
2555 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2556 canonical form (fully expanded) before being placed in
2557 $sender_host_address.
2559 36. The table in the code that translates DNS record types into text (T_A to
2560 "A" for instance) was missing entries for NS and CNAME. It is just possible
2561 that this could have caused confusion if both these types were looked up
2562 for the same domain, because the text type is used as part of Exim's
2563 per-process caching. But the chance of anyone hitting this buglet seems
2566 37. The dnsdb lookup has been extended in a number of ways.
2568 (1) There is a new type, "zns", which walks up the domain tree until it
2569 finds some nameserver records. It should be used with care.
2571 (2) There is a new type, "mxh", which is like "mx" except that it returns
2572 just the host names, not the priorities.
2574 (3) It is now possible to give a list of domains (or IP addresses) to be
2575 looked up. The behaviour when one of the lookups defers can be
2576 controlled by a keyword.
2578 (4) It is now possible to specify the separator character for use when
2579 multiple records are returned.
2581 38. The dnslists ACL condition has been extended: it is now possible to supply
2582 a list of IP addresses and/or domains to be looked up in a particular DNS
2585 39. Added log_selector=+queue_time_overall.
2587 40. When running the queue in the test harness, wait just a tad after forking a
2588 delivery process, to get repeatability of debugging output.
2590 41. Include certificate and key file names in error message when GnuTLS fails
2591 to set them up, because the GnuTLS error message doesn't include the name
2592 of the failing file when there is a problem reading it.
2594 42. Allow both -bf and -bF in the same test run.
2596 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2598 44. The "Exiscan patch" is now merged into the mainline Exim source.
2600 45. Sometimes the final signoff response after QUIT could fail to get
2601 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2602 < 0 before doing a fflush(). This bug looks as though it goes back to the
2603 introduction of TLS in release 3.20, but "sometimes" must have been rare
2604 because the tests only now provoked it.
2606 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2607 (this can affect the format of dates).
2609 47. exim_tidydb, when checking for the continued existence of a message for
2610 which it has found a message-specific retry record, was not finding
2611 messages that were in split spool directories. Consequently, it was
2612 deleting retry records that should have stayed in existence.
2614 48. Steve fixed some bugs in eximstats.
2616 49. The SPA authentication driver was not abandoning authentication and moving
2617 on to the next authenticator when an expansion was forced to fail,
2618 contradicting the general specification for all authenticators. Instead it
2619 was generating a temporary error. It now behaves as specified.
2621 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2622 (the order specifies the preference for clients). The order is now AES256,
2623 AES128, 3DES, ARCFOUR128.
2625 51. Small patch to Sieve code - explicitly set From: when generating an
2628 52. Exim crashed if a remote delivery caused a very long error message to be
2629 recorded - for instance if somebody sent an entire SpamAssassin report back
2630 as a large number of 550 error lines. This bug was coincidentally fixed by
2631 increasing the size of one of Exim's internal buffers (big_buffer) that
2632 happened as part of the Exiscan merge. However, to be on the safe side, I
2633 have made the code more robust (and fixed the comments that describe what
2636 53. Now that there can be additional text after "Completed" in log lines (if
2637 the queue_time_overall log selector is set), a one-byte patch to exigrep
2638 was needed to allow it to recognize "Completed" as not the last thing in
2641 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2642 patch that reportedly fixes this has been added. I am not expert enough to
2643 create a test for it. This is what the patch creator wrote:
2645 "I found a little strange behaviour of ldap code when working with
2646 Windows 2003 AD Domain, where users was placed in more than one
2647 Organization Units. When I tried to give exim partial DN, the exit code
2648 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2649 But simultaneously result of request was absolutely normal ldap result,
2650 so I produce this patch..."
2652 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2653 REFERENCE, so I have modified the code to exclude the patch when that macro
2656 55. Some experimental protocols are using DNS PTR records for new purposes. The
2657 keys for these records are domain names, not reversed IP addresses. The
2658 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2659 leaves it alone. Component reversal etc. now happens only for IP addresses.
2662 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2664 57. Double the size of the debug message buffer (to 2048) so that more of very
2665 long debug lines gets shown.
2667 58. The exicyclog utility now does better if the number of log files to keep
2668 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2670 59. Two changes related to the smtp_active_hostname option:
2672 (1) $smtp_active_hostname is now available as a variable.
2673 (2) The default for smtp_banner uses $smtp_active_hostname instead
2674 of $primary_hostname.
2676 60. The host_aton() function is supposed to be passed a string that is known
2677 to be a valid IP address. However, in the case of IPv6 addresses, it was
2678 not checking this. This is a hostage to fortune. Exim now panics and dies
2679 if the condition is not met. A case was found where this could be provoked
2680 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2681 components; fortuitously, this particular loophole had already been fixed
2682 by change 4.50/55 above.
2684 If there are any other similar loopholes, the new check in host_aton()
2685 itself should stop them being exploited. The report I received stated that
2686 data on the command line could provoke the exploit when Exim was running as
2687 exim, but did not say which command line option was involved. All I could
2688 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2689 running as the user.
2692 61. There was a buffer overflow vulnerability in the SPA authentication code
2693 (which came originally from the Samba project). I have added a test to the
2694 spa_base64_to_bits() function which I hope fixes it.
2697 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2698 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2700 63. The daemon start-up calls getloadavg() while still root for those OS that
2701 need the first call to be done as root, but it missed one case: when
2702 deliver_queue_load_max is set with deliver_drop_privilege. This is
2703 necessary for the benefit of the queue runner, because there is no re-exec
2704 when deliver_drop_privilege is set.
2706 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2707 This has been fixed.
2709 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2710 was in use, was not putting the data itself into the right store pool;
2711 consequently, it could be overwritten for a subsequent message in the same
2712 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2715 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2717 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2718 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2719 the function were treating the return as a boolean value, which happened to
2720 work because 0=false and not-0=true, but is not correct code.
2722 68. The host_aton() function was not handling scoped IPv6 addresses (those
2723 with, for example, "%eth0" on the end) correctly.
2725 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2726 negated items (that is, ~something) in unsigned ints. Some compilers
2727 apparently mutter when there is no cast.
2729 70. If an address verification called from an ACL failed, and did not produce a
2730 user-specific message (i.e. there was only a "system" message), nothing was
2731 put in $acl_verify_message. In this situation, it now puts the system
2734 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2735 session; change 4.31/43 added the unwanted input to the log line - except
2736 that it did not do this in the start of session case. It now does.
2738 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2739 This is wrong and can cause the other end to generate a synchronization
2740 error if it is another Exim or anything else that does the synchronization
2741 check. A QUIT command is no longer sent after a timeout.
2743 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2744 during host lookups.
2746 74. The defer_ok option of callout verification was not working if it was used
2747 when verifying addresses in header lines, that is, for this case:
2749 verify = header_sender/callout=defer_ok
2751 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2752 those file descriptors could be used for SMTP connections. If anything
2753 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2754 could be sent to the SMTP client, causing chaos. The daemon now opens
2755 stdin, stdout, and stderr to /dev/null when it puts itself into the
2758 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2759 log by default. The user can override this with suitable Perl magic.
2761 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2762 the log message when discard triggers, was not working for the DATA ACL or
2763 for the non-SMTP ACL.
2765 78. Error message wording change in sieve.c.
2767 79. If smtp_accept_max_per_host was set, the number of connections could be
2768 restricted to fewer than expected, because the daemon was trying to set up
2769 a new connection before checking whether the processes handling previous
2770 connections had finished. The check for completed processes is now done
2771 earlier. On busy systems, this bug wouldn't be noticed because something
2772 else would have woken the daemon, and it would have reaped the completed
2775 80. If a message was submitted locally by a user whose login name contained one
2776 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2777 It caused a spool format error. I have fixed the spool reading code. A
2778 related problem was that the "from" clause in the Received: line became
2779 illegal because of the space(s). It is now covered by ${quote_local_part.
2781 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2784 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2786 83. Patch from Sieve maintainer to fix unterminated string problem in
2787 "vacation" handling.
2789 84. Some minor changes to the Linux configuration files to help with other
2790 OS variants using glibc.
2792 85. One more patch for Sieve to update vacation handling to latest spec.
2795 ----------------------------------------------------
2796 See the note above about the 4.44 and 4.50 releases.
2797 ----------------------------------------------------
2803 1. Change 4.43/35 introduced a bug that caused file counts to be
2804 incorrectly computed when quota_filecount was set in an appendfile
2807 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2808 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2811 3. Give more explanation in the error message when the command for a transport
2812 filter fails to execute.
2814 4. There are several places where Exim runs a non-Exim command in a
2815 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2816 was being done only for the command run by the queryprogram router. It is
2817 now done for all such subprocesses. The other cases are: ${run, transport
2818 filters, and the commands run by the lmtp and pipe transports.
2820 5. Some older OS have a limit of 256 on the maximum number of file
2821 descriptors. Exim was using setrlimit() to set 1000 as a large value
2822 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2823 systems. I've change it so that if it can't get 1000, it tries for 256.
2825 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2826 was an oversight, and furthermore, ever since the addition of extra
2827 controls (e.g. 4.43/32), the checks on when to allow different forms of
2828 "control" were broken. There should now be diagnostics for all cases when a
2829 control that does not make sense is encountered.
2831 7. $recipients is now available in the predata ACL (oversight).
2833 8. Tidy the search cache before the fork to do a delivery from a message
2834 received from the command line. Otherwise the child will trigger a lookup
2835 failure and thereby defer the delivery if it tries to use (for example) a
2836 cached ldap connection that the parent has called unbind on.
2838 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2839 of $address_data from the recipient verification was clobbered by the
2840 sender verification.
2842 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2843 was its contents. (It was OK if the option was not defined at all.)
2845 11. A "Completed" log line is now written for messages that are removed from
2846 the spool by the -Mrm option.
2848 12. $host_address is now set to the target address during the checking of
2849 ignore_target_hosts.
2851 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2852 being passed; this would have caused $sender_host_name to have been used if
2853 matching the list had actually called for a host name (not very likely,
2854 since this list is usually IP addresses). A host name is now passed as
2857 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2858 code that specifies a non-restarting handler (typically sigaction() in
2859 modern systems) in an attempt to fix a rare and obscure crash bug.
2861 15. Narrowed the window for a race in the daemon that could cause it to ignore
2862 SIGCHLD signals. This is not a major problem, because they are used only to
2863 wake it up if nothing else does.
2865 16. A malformed maildirsize file could cause Exim to calculate negative values
2866 for the mailbox size or file count. Odd effects could occur as a result.
2867 The maildirsize information is now recalculated if the size or filecount
2870 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2871 support for a long time. Removed HAVE_SYS_VFS_H.
2873 18. Updated exipick to current release from John Jetmore.
2875 19. Allow an empty sender to be matched against a lookup in an address list.
2876 Previously the only cases considered were a regular expression, or an
2879 20. Exim went into a mad DNS lookup loop when doing a callout where the
2880 host was specified on the transport, if the DNS lookup yielded more than
2883 21. The RFC2047 encoding function was originally intended for short strings
2884 such as real names; it was not keeping to the 75-character limit for
2885 encoded words that the RFC imposes. It now respects the limit, and
2886 generates multiple encoded words if necessary. To be on the safe side, I
2887 have increased the buffer size for the ${rfc2047: expansion operator from
2890 22. Failure to deliver a bounce message always caused it to be frozen, even if
2891 there was an errors_to setting on the router. The errors_to setting is now
2894 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2895 canonical form (fully expanded) before being placed in
2896 $sender_host_address.
2898 24. Updated eximstats to version 1.33
2900 25. Include certificate and key file names in error message when GnuTLS fails
2901 to set them up, because the GnuTLS error message doesn't include the name
2902 of the failing file when there is a problem reading it.
2904 26. Expand error message when OpenSSL has problems setting up cert/key files.
2907 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2908 (this can affect the format of dates).
2910 28. exim_tidydb, when checking for the continued existence of a message for
2911 which it has found a message-specific retry record, was not finding
2912 messages that were in split spool directories. Consequently, it was
2913 deleting retry records that should have stayed in existence.
2915 29. eximstats updated to version 1.35
2916 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2917 1.35 - bugfix such that pie charts by volume are generated correctly
2919 30. The SPA authentication driver was not abandoning authentication and moving
2920 on to the next authenticator when an expansion was forced to fail,
2921 contradicting the general specification for all authenticators. Instead it
2922 was generating a temporary error. It now behaves as specified.
2924 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2925 (the order specifies the preference for clients). The order is now AES256,
2926 AES128, 3DES, ARCFOUR128.
2928 31. Small patch to Sieve code - explicitly set From: when generating an
2931 32. Exim crashed if a remote delivery caused a very long error message to be
2932 recorded - for instance if somebody sent an entire SpamAssassin report back
2933 as a large number of 550 error lines. This bug was coincidentally fixed by
2934 increasing the size of one of Exim's internal buffers (big_buffer) that
2935 happened as part of the Exiscan merge. However, to be on the safe side, I
2936 have made the code more robust (and fixed the comments that describe what
2939 33. Some experimental protocols are using DNS PTR records for new purposes. The
2940 keys for these records are domain names, not reversed IP addresses. The
2941 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2942 leaves it alone. Component reversal etc. now happens only for IP addresses.
2945 34. The host_aton() function is supposed to be passed a string that is known
2946 to be a valid IP address. However, in the case of IPv6 addresses, it was
2947 not checking this. This is a hostage to fortune. Exim now panics and dies
2948 if the condition is not met. A case was found where this could be provoked
2949 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2950 components; fortuitously, this particular loophole had already been fixed
2951 by change 4.50/55 or 4.44/33 above.
2953 If there are any other similar loopholes, the new check in host_aton()
2954 itself should stop them being exploited. The report I received stated that
2955 data on the command line could provoke the exploit when Exim was running as
2956 exim, but did not say which command line option was involved. All I could
2957 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2958 running as the user.
2961 35. There was a buffer overflow vulnerability in the SPA authentication code
2962 (which came originally from the Samba project). I have added a test to the
2963 spa_base64_to_bits() function which I hope fixes it.
2966 36. The daemon start-up calls getloadavg() while still root for those OS that
2967 need the first call to be done as root, but it missed one case: when
2968 deliver_queue_load_max is set with deliver_drop_privilege. This is
2969 necessary for the benefit of the queue runner, because there is no re-exec
2970 when deliver_drop_privilege is set.
2972 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2973 was in use, was not putting the data itself into the right store pool;
2974 consequently, it could be overwritten for a subsequent message in the same
2975 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2978 38. Sometimes the final signoff response after QUIT could fail to get
2979 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2980 < 0 before doing a fflush(). This bug looks as though it goes back to the
2981 introduction of TLS in release 3.20, but "sometimes" must have been rare
2982 because the tests only now provoked it.
2988 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2989 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2990 Now it returns an integer. A number of calls were still expecting a T/F
2991 return. Fortuitously, in all cases, the tests worked in OK situations,
2992 which is the norm. However, things would have gone wrong on any write
2993 failures on the smtp file descriptor. This function is used when sending
2994 messages over SMTP and also when doing verify callouts.
2996 2. When Exim is called to do synchronous delivery of a locally submitted
2997 message (the -odf or -odi options), it no longer closes stderr before doing
3000 3. Implemented the mua_wrapper option.
3002 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3004 5. Implemented the functions header_remove(), header_testname(),
3005 header_add_at_position(), and receive_remove_recipient(), and exported them
3008 6. If an ACL "warn" statement specified the addition of headers, Exim already
3009 inserted X-ACL-Warn: at the start if there was no header name. However, it
3010 was not making this test for the second and subsequent header lines if
3011 there were newlines in the string. This meant that an invalid header could
3012 be inserted if Exim was badly configured.
3014 7. Allow an ACL "warn" statement to add header lines at the start or after all
3015 the Received: headers, as well as at the end.
3017 8. Added the rcpt_4xx retry error code.
3019 9. Added postmaster_mailfrom=xxx to callout verification option.
3021 10. Added mailfrom=xxxx to the callout verification option, for verify=
3024 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3025 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3027 12. Inserted some casts to stop certain compilers warning when using pointer
3028 differences as field lengths or precisions in printf-type calls (mostly
3029 affecting debugging statements).
3031 13. Added optional readline() support for -be (dynamically loaded).
3033 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3034 same clock tick as a message's arrival, so that its received time was the
3035 same as the "first fail" time on the retry record, and that message
3036 remained on the queue past the ultimate address timeout, every queue runner
3037 would try a delivery (because it was past the ultimate address timeout) but
3038 after another failure, the ultimate address timeout, which should have then
3039 bounced the address, did not kick in. This was a "< instead of <=" error;
3040 in most cases the first failure would have been in the next clock tick
3041 after the received time, and all would be well.
3043 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3044 being recognized when the domain list was tested by the match_domain
3045 condition in an expansion string.
3047 16. Added the ${str2b64: operator.
3049 17. Exim was always calling setrlimit() to set a large limit for the number of
3050 processes, without checking whether the existing limit was already
3051 adequate. (It did check for the limit on file descriptors.) Furthermore,
3052 errors from getrlimit() and setrlimit() were being ignored. Now they are
3053 logged to the main and panic logs, but Exim does carry on, to try to do its
3054 job under whatever limits there are.
3056 18. Imported PCRE 5.0.
3058 19. Trivial typo in log message " temporarily refused connection" (the leading
3061 20. If the log selector return_path_on_delivery was set and an address was
3062 redirected to /dev/null, the delivery process crashed because it assumed
3063 that a return path would always be set for a "successful" delivery. In this
3064 case, the whole delivery is bypassed as an optimization, and therefore no
3067 21. Internal re-arrangement: the function for sending a challenge and reading
3068 a response while authentication was assuming a zero-terminated challenge
3069 string. It's now changed to take a pointer and a length, to allow for
3070 binary data in such strings.
3072 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3074 23. Exim was not respecting finduser_retries when seeking the login of the
3075 uid under which it was called; it was always trying 10 times. (The default
3076 setting of finduser_retries is zero.) Also, it was sleeping after the final
3077 failure, which is pointless.
3079 24. Implemented tls_on_connect_ports.
3081 25. Implemented acl_smtp_predata.
3083 26. If the domain in control=submission is set empty, Exim assumes that the
3084 authenticated id is a complete email address when it generates From: or
3085 Sender: header lines.
3087 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3088 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3089 chown and chgrp in /bin and hostname in /usr/bin.
3091 28. Exim was keeping the "process log" file open after each use, just as it
3092 does for the main log. This opens the possibility of it remaining open for
3093 long periods when the USR1 signal hits a daemon. Occasional processlog
3094 errors were reported, that could have been caused by this. Anyway, it seems
3095 much more sensible not to leave this file open at all, so that is what now
3098 29. The long-running daemon process does not normally write to the log once it
3099 has entered its main loop, and it closes the log before doing so. This is
3100 so that log files can straightforwardly be renamed and moved. However,
3101 there are a couple of unusual error situations where the daemon does write
3102 log entries, and I had neglected to close the log afterwards.
3104 30. The text of an SMTP error response that was received during a remote
3105 delivery was being truncated at 512 bytes. This is too short for some of
3106 the long messages that one sometimes sees. I've increased the limit to
3109 31. It is now possible to make retry rules that apply only when a message has a
3110 specific sender, in particular, an empty sender.
3112 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3113 it possible to be selective about when SMTP synchronization is enforced.
3115 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3117 32. Implemented hosts_connection_nolog.
3119 33. Added an ACL for QUIT.
3121 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3124 35. Added mailbox_size and mailbox_filecount to appendfile.
3126 36. Added control = no_multiline_responses to ACLs.
3128 37. There was a bug in the logic of the code that waits for the clock to tick
3129 in the case where the clock went backwards by a substantial amount such
3130 that the microsecond fraction of "now" was more than the microsecond
3131 fraction of "then" (but the whole seconds number was less).
3133 38. Added support for the libradius Radius client library this is found on
3134 FreeBSD (previously only the radiusclient library was supported).
3140 1. When certain lookups returned multiple values in the form name=value, the
3141 quoting of the values was not always being done properly. Specifically:
3142 (a) If the value started with a double quote, but contained no whitespace,
3144 (b) If the value contained whitespace other than a space character (i.e.
3145 tabs or newlines or carriage returns) it was not quoted.
3146 This fix has been applied to the mysql and pgsql lookups by writing a
3147 separate quoting function and calling it from the lookup code. The fix
3148 should probably also be applied to nisplus, ibase and oracle lookups, but
3149 since I cannot test any of those, I have not disturbed their existing code.
3151 2. A hit in the callout cache for a specific address caused a log line with no
3152 reason for rejecting RCPT. Now it says "Previous (cached) callout
3153 verification failure".
3155 3. There was an off-by-one bug in the queryprogram router. An over-long
3156 return line was truncated at 256 instead of 255 characters, thereby
3157 overflowing its buffer with the terminating zero. As well as fixing this, I
3158 have increased the buffer size to 1024 (and made a note to document this).
3160 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3161 when Exim is waiting for an SMTP response from a remote server, Exim
3162 restarts its select() call on the socket, thereby resetting its timeout.
3163 This is not a problem when such interrupts are rare. Somebody set up a cron
3164 job to run exiwhat every 2 minutes, which is less than the normal select()
3165 timeout (5 or 10 minutes). This meant that the select() timeout never
3166 kicked in because it was always reset. I have fixed this by comparing the
3167 time when an interrupt arrives with the time at the start of the first call
3168 to select(). If more time than the timeout has elapsed, the interrupt is
3169 treated as a timeout.
3171 5. Some internal re-factoring in preparation for the addition of Sieve
3172 extensions (by MH). In particular, the "personal" test is moved to a
3173 separate function, and given an option for scanning Cc: and Bcc: (which is
3174 not set for Exim filters).
3176 6. When Exim created an email address using the login of the caller as the
3177 local part (e.g. when creating a From: or Sender: header line), it was not
3178 quoting the local part when it contained special characters such as @.
3180 7. Installed new OpenBSD configuration files.
3182 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3183 try to make them clearer.
3185 9. Callout options, other than the timeout value, were being ignored when
3186 verifying sender addresses in header lines. For example, when using
3188 verify = header_sender/callout=no_cache
3190 the cache was (incorrectly) being used.
3192 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3193 only the Cygwin environment.
3195 11. When return_path_on_delivery was set as a log selector, if different remote
3196 addresses in the same message used different return paths and parallel
3197 remote delivery occurred, the wrong values would sometimes be logged.
3198 (Whenever a remote delivery process finished, the return path value from
3199 the most recently started remote delivery process was logged.)
3201 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3202 lines when AUTH and/or TLS are in use. This is the "received protocol"
3203 field. Exim used to use "asmtp" for authenticated SMTP, without any
3204 indication (in the protocol name) for TLS use. Now it follows the RFC and
3205 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3206 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3207 names appear in log lines as well as in Received: header lines.
3209 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3210 extensions, and comparison tests, and to fix some bugs.
3212 14. Changes to the "personal" filter test:
3214 (1) The test was buggy in that it was just doing the equivalent of
3215 "contains" tests on header lines. For example, if a user's address was
3216 anne@some.where, the "personal" test would incorrectly be true for
3218 To: susanne@some.where
3220 This test is now done by extracting each address from the header in turn,
3221 and checking the entire address. Other tests that are part of "personal"
3222 are now done using regular expressions (for example, to check local parts
3223 of addresses in From: header lines).
3225 (2) The list of non-personal local parts in From: addresses has been
3226 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3227 taken from the Sieve specification recommendations.
3229 (3) If the message contains any header line starting with "List-" it is
3230 treated as non-personal.
3232 (4) The test for "circular" in the Subject: header line has been removed
3233 because it now seems ill-conceived.
3235 15. Minor typos in src/EDITME comments corrected.
3237 16. Installed latest exipick from John Jetmore.
3239 17. If headers_add on a router specified a text string that was too long for
3240 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3241 of string_sprintf() is now avoided.
3243 18. $message_body_size was not set (it was always zero) when running the DATA
3244 ACL and the local_scan() function.
3246 19. For the "mail" command in an Exim filter, no default was being set for
3247 the once_repeat time, causing a random time value to be used if "once" was
3248 specified. (If the value happened to be <= 0, no repeat happened.) The
3249 default is now 0s, meaning "never repeat". The "vacation" command was OK
3250 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3251 (I found it when inspecting the code).
3253 20. There is now an overall timeout for performing a callout verification. It
3254 defaults to 4 times the callout timeout, which applies to individual SMTP
3255 commands during the callout. The overall timeout applies when there is more
3256 than one host that can be tried. The timeout is checked before trying the
3257 next host. This prevents very long delays if there are a large number of
3258 hosts and all are timing out (e.g. when the network connections are timing
3259 out). The value of the overall timeout can be changed by specifying an
3260 additional sub-option for "callout", called "maxwait". For example:
3262 verify = sender/callout=5s,maxwait=20s
3264 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3265 to the end before writing, but this should make it even safer).
3267 22. Exim was forgetting that it had advertised PIPELINING for the second and
3268 subsequent messages on an SMTP connection. It was also not resetting its
3269 memory on STARTTLS and an internal HELO.
3271 23. When Exim logs an SMTP synchronization error within a session, it now
3272 records whether PIPELINING has been advertised or not.
3274 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3275 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3276 rather than long int.
3278 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3280 26. Added the never_mail option to autoreply.
3286 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3287 crash if the getsockname() call failed; this can happen if a connection is
3288 closed very soon after it is established. The problem was simply in the
3289 order in which certain operations were done, causing Exim to try to write
3290 to the SMTP stream before it had set up the file descriptor. The bug has
3291 been fixed by making things happen in the correct order.
3297 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3298 before the connection was closed, thus losing the rejection response.
3300 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3301 some early Solaris releases, but causes trouble in current releases where
3302 socklen_t is defined.
3304 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3307 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3310 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3311 that releasing the top of it at the end releases what was used for sub-
3312 expansions (unless the block got too big). However, discard this block if
3313 the first thing is a variable or header, so that we can use its block when
3314 it is dynamic (useful for very large $message_headers, for example).
3316 6. Lookups now cache *every* query, not just the most recent. A new, separate
3317 store pool is used for this. It can be recovered when all lookup caches are
3318 flushed. Lookups now release memory at the end of their result strings.
3319 This has involved some general refactoring of the lookup sources.
3321 7. Some code has been added to the store_xxx() functions to reduce the amount
3322 of flapping under certain conditions.
3324 8. log_incoming_interface used to affect only the <= reception log lines. Now
3325 it causes the local interface and port to be added to several more SMTP log
3326 lines, for example "SMTP connection from", and rejection lines.
3328 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3330 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3332 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3333 could be overwritten at the end of the current message (or the start of a
3334 new message if it was set in a HELO ACL). The value is now preserved for
3335 the duration of the SMTP connection.
3337 12. If a transport had a headers_rewrite setting, and a matching header line
3338 contained an unqualified address, that address was qualified, even if it
3339 did not match any rewriting rules. The underlying bug was that the values
3340 of the flags that permit the existence of unqualified sender and recipient
3341 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3342 non-local messages, and by -bnq for local messages) were not being
3343 preserved with the message after it was received.
3345 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3346 "next input=" as part of the text comprising the host identity instead of
3347 the correct text. The code was using the same buffer for two different
3348 strings. However, depending on which order the printing function evaluated
3349 its arguments, the bug did not always show up. Under Linux, for example, my
3350 test suite worked just fine.
3352 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3353 doesn't work with some older versions of Perl. It has been changed to "my",
3354 which in any case is probably the better facility to use.
3356 15. A really picky compiler found some instances of statements for creating
3357 error messages that either had too many or two few arguments for the format
3360 16. The size of the buffer for calls to the DNS resolver has been increased
3361 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3362 for addresses that have a lot of PTR records. This alleviates a problem; it
3363 does not fully solve it.
3365 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3366 buffer now truncates the list and logs the incident, which is the same
3367 action as happens when Exim is looking up a host name and its aliases.
3368 Previously in this situation something unpredictable would happen;
3369 sometimes it was "internal error: store_reset failed".
3371 18. If a server dropped the connection unexpectedly when an Exim client was
3372 using GnuTLS and trying to read a response, the client delivery process
3373 crashed while trying to generate an error log message.
3375 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3376 string, for example:
3378 warn message = H1: something\nH2: something
3380 the text was added as a single header line from Exim's point of view
3381 though it ended up OK in the delivered message. However, searching for the
3382 second and subsequent header lines using $h_h2: did not work. This has been
3383 fixed. Similarly, if a system filter added multiple headers in this way,
3384 the routers could not see them.
3386 20. Expanded the error message when iplsearch is called with an invalid key to
3387 suggest using net-iplsearch in a host list.
3389 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3390 ACLs are no longer actually imposed (and a message to that effect is
3393 22. If a "gecos" field in a passwd entry contained escaped characters, in
3394 particular, if it contained a \" sequence, Exim got it wrong when building
3395 a From: or a Sender: header from that name. A second bug also caused
3396 incorrect handling when an unquoted " was present following a character
3397 that needed quoting.
3399 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3400 was not being matched caselessly.
3402 24. Arranged for all hyphens in the exim.8 source to be escaped with
3405 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3406 properly. Recipient callout cache records were still being keyed to include
3407 the sender, even when use_sender was set false. This led to far more
3408 callouts that were necessary. The sender is no longer included in the key
3409 when use_sender is false.
3411 26. Added "control = submission" modifier to ACLs.
3413 27. Added the ${base62d: operator to decode base 62 numbers.
3415 28. dnsdb lookups can now access SRV records.
3417 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3418 the configuration file.
3420 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3421 (-v) mode. This makes the output for a verbose queue run more intelligible.
3423 31. Added a use_postmaster feature to recipient callouts.
3425 32. Added the $body_zerocount variable, containing the number of binary zero
3426 bytes in the message body.
3428 33. The time of last modification of the "new" subdirectory is now used as the
3429 "mailbox time last read" when there is a quota error for a maildir
3432 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3434 35. Added +ignore_unknown as a special item in host lists.
3436 36. Code for decoding IPv6 addresses in host lists is now included, even if
3437 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3438 address was recognized as an IP address, but was then not correctly decoded
3439 into binary, causing unexpected and incorrect effects when compared with
3446 1. Very minor rewording of debugging text in manualroute to say "list of
3447 hosts" instead of "hostlist".
3449 2. If verify=header_syntax was set, and a header line with an unqualified
3450 address (no domain) and a large number of spaces between the end of the
3451 name and the colon was received, the reception process suffered a buffer
3452 overflow, and (when I tested it) crashed. This was caused by some obsolete
3453 code that should have been removed. The fix is to remove it!
3455 3. When running in the test harness, delay a bit after writing a bounce
3456 message to get a bit more predictability in the log output.
3458 4. Added a call to search_tidyup() just before forking a reception process. In
3459 theory, someone could use a lookup in the expansion of smtp_accept_max_
3460 per_host which, without the tidyup, could leave open a database connection.
3462 5. Added the variables $recipient_data and $sender_data which get set from a
3463 lookup success in an ACL "recipients" or "senders" condition, or a router
3464 "senders" option, similar to $domain_data and $local_part_data.
3466 6. Moved the writing of debug_print from before to after the "senders" test
3469 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3470 problems for message scanning, either using a data ACL, or using
3471 local_scan() because the Received: header was not generated till after they
3472 were called (in order to set the time as the time of reception completion).
3473 I have revised the way this works. The header is now generated after the
3474 body is received, but before the ACL or local_scan() are called. After they
3475 are run, the timestamp in the header is updated.
3481 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3482 before starting a queue runner without re-exec. This happened only when
3483 deliver_drop_privilege was set or when the Exim user was set to root. The
3484 effect of the bug was that timeouts during subsequent deliveries caused
3485 crashes instead of being properly handled. The handler is now left at its
3486 default (and expected) setting.
3488 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3489 message, again when deliver_drop_privilege is set or Exim is run as root.
3490 The bug described in (1) was not present in this case, but the tidying up
3491 of the other signals was missing. I have made the two cases consistent.
3493 3. The ignore_target_hosts setting on a manualroute router was being ignored
3494 for hosts that were looked up using the /MX notation.
3496 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3499 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3500 operated on the sender address. After changing the $sender_address to <>
3501 for the sender address verify, Exim was re-instated it as the original
3502 (before rewriting) address, but remembering that it had rewritten it, so it
3503 wasn't rewriting it again. This bug also had the effect of breaking the
3504 sender address verification caching when the sender address was rewritten.
3506 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3507 This has been changed so that if the ip literal address matches
3508 ignore_target_hosts, the router declines.
3510 7. Added expansion conditions match_domain, match_address, and match_local_
3511 part (NOT match_host).
3513 8. The placeholder for the Received: header didn't have a length field set.
3515 9. Added code to Exim itself and to exim_lock to test for a specific race
3516 condition that could lead to file corruption when using MBX delivery. The
3517 issue is with the lockfile that is created in /tmp. If this file is removed
3518 after a process has opened it but before that process has acquired a lock,
3519 there is the potential for a second process to recreate the file and also
3520 acquire a lock. This could lead to two Exim processes writing to the file
3521 at the same time. The added code performs the same test as UW imapd; it
3522 checks after acquiring the lock that its file descriptor still refers to
3523 the same named file.
3525 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3526 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3529 11. Added the smtp_active_hostname option. If used, this will typically be made
3530 to depend on the incoming interface address. Because $interface_address is
3531 not set up until the daemon has forked a reception process, error responses
3532 that can happen earlier (such as "too many connections") no longer contain
3535 12. If an expansion in a condition on a "warn" statement fails because a lookup
3536 defers, the "warn" statement is abandoned, and the next ACL statement is
3537 processed. Previously this caused the whole ACL to be aborted.
3539 13. Added the iplsearch lookup type.
3541 14. Added ident_timeout as a log selector.
3543 15. Added tls_certificate_verified as a log selector.
3545 16. Added a global option tls_require_ciphers (compare the smtp transport
3546 option of the same name). This controls incoming TLS connections.
3548 17. I finally figured out how to make tls_require_ciphers do a similar thing
3549 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3550 before starting the TLS session.
3552 18. Tabs are now shown as \t in -bP output.
3554 19. If the log selector return_path_on_delivery was set, Exim crashed when
3555 bouncing a message because it had too many Received: header lines.
3557 20. If two routers both had headers_remove settings, and the first one included
3558 a superfluous trailing colon, the final name in the first list and the
3559 first name in the second list were incorrectly joined into one item (with a
3560 colon in the middle).
3566 1. Added -C and -D options to the exinext utility, mainly to make it easier
3567 to include in the automated testing, but these could be helpful when
3568 multiple configurations are in use.
3570 2. The exinext utility was not formatting the output nicely when there was
3571 an alternate port involved in the retry record key, nor when there was a
3572 message id as well (for retries that were specific to a specific message
3573 and a specific host). It was also confused by IPv6 addresses, because of
3574 the additional colons they contain. I have fixed the IPv4 problem, and
3575 patched it up to do a reasonable job for IPv6.
3577 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3578 delivery, the log line now contains "pipelined" if PIPELINING was used.
3580 4. An SMTP transport process used to panic and die if the bind() call to set
3581 an explicit outgoing interface failed. This has been changed; it is now
3582 treated in the same way as a connect() failure.
3584 5. A reference to $sender_host_name in the part of a conditional expansion
3585 that was being skipped was still causing a DNS lookup. This no longer
3588 6. The def: expansion condition was not recognizing references to header lines
3589 that used bh_ and bheader_.
3591 7. Added the _cache feature to named lists.
3593 8. The code for checking quota_filecount in the appendfile transport was
3594 allowing one more file than it should have been.
3596 9. For compatibility with Sendmail, the command line option
3604 and sets the incoming protocol and host name (for trusted callers). The
3605 host name and its colon can be omitted when only the protocol is to be set.
3606 Note the Exim already has two private options, -pd and -ps, that refer to
3607 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3608 "s", but I don't think that's a major issue.
3610 10. A number of refactoring changes to the code, none of which should affect
3613 (a) The number of logging options was getting close to filling up the
3614 32-bit word that was used as a bit map. I have split them into two classes:
3615 those that are passed in the argument to log_write(), and those that are
3616 only ever tested independently outside of that function. These are now in
3617 separate 32-bit words, so there is plenty of room for expansion again.
3618 There is no change in the user interface or the logging behaviour.
3620 (b) When building, for example, log lines, the code previously used a
3621 macro that called string_cat() twice, in order to add two strings. This is
3622 not really sufficiently general. Furthermore, there was one instance where
3623 it was actually wrong because one of the argument was used twice, and in
3624 one call a function was used. (As it happened, calling the function twice
3625 did not affect the overall behaviour.) The macro has been replaced by a
3626 function that can join an arbitrary number of extra strings onto a growing
3629 (c) The code for expansion conditions now uses a table and a binary chop
3630 instead of a serial search (which was left over from when there were very
3631 few conditions). Also, it now recognizes conditions like "pam" even when
3632 the relevant support is not compiled in: a suitably worded error message is
3633 given if an attempt is made to use such a condition.
3635 11. Added ${time_interval:xxxxx}.
3637 12. A bug was causing one of the ddress fields not to be passed back correctly
3638 from remote delivery subprocesses. The field in question was not being
3639 subsequently used, so this caused to problems in practice.
3641 13. Added new log selectors queue_time and deliver_time.
3643 14. Might have fixed a bug in maildirsizefile handling that threw up
3644 "unexpected character" debug warnings, and recalculated the data
3645 unnecessarily. In any case, I expanded the warning message to give more
3648 15. Added the message "Restricted characters in address" to the statements in
3649 the default ACL that block characters like @ and % in local parts.
3651 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3652 Three changes have been made:
3654 (a) There was a serious bug; a negative response to MAIL caused the whole
3655 recipient domain to be cached as invalid, thereby blocking all messages
3656 to all local parts at the same domain, from all senders. This bug has
3657 been fixed. The domain is no longer cached after a negative response to
3658 MAIL if the sender used is not empty.
3660 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3663 (c) A new callout option, "use_sender" has been added for people who want
3664 the modified behaviour.
3670 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3673 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3674 indeed breaks things for older releases.
3676 3. Added additional logging to the case where there is a problem reading data
3677 from a filter that is running in a subprocess using a pipe, in order to
3678 try to track down a specific problem.
3680 4. Testing facility fudge: when running in the test harness and attempting
3681 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3682 getting "No route to host". Convert this to a timeout.
3684 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3687 6. Some OS don't have socklen_t but use size_t instead. This affects the
3688 fifth argument of getsockopt() amongst other things. This is now
3689 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3690 can be set for individual OS. I have set it for SunOS5, OSF1, and
3691 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3692 some earlier ones do not.
3694 7. Change 4.30/15 was not doing the test caselessly.
3696 8. The standard form for an IPv6 address literal was being rejected by address
3697 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3698 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3699 this, as well as the form without the "IPv6" on the front (but only when
3700 address literals are enabled, of course).
3702 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3704 10. Exim crashed if a message with an empty sender address specified by -f
3705 encountered a router with an errors_to setting. This could be provoked only
3706 by a command such as
3710 where an empty string was supplied; "<>" did not hit this bug.
3712 11. Installed PCRE release 4.5.
3714 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3715 remained set. It is now erased.
3717 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3718 times from message ids (which are base 36 rather than the normal 62).
3720 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3721 were being counted as actual protocol errors, and logged if the log
3722 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3723 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3724 and DATA following a set of rejected RCPTs do not count as protocol errors.
3725 In other words, Exim assumes they were pipelined, though this may not
3726 actually be the case. Of course, in all cases the client gets an
3727 appropriate error code.
3729 15. If a lookup fails in an ACL condition, a message about the failure may
3730 be available; it is used if testing the ACL cannot continue, because most
3731 such messages specify what the cause of the deferral is. However, some
3732 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3733 that caused an old message to be retained and used if a later statement
3734 caused a defer, replacing the real cause of the deferral.
3736 16. If an IP address had so many PTR records that the DNS lookup buffer
3737 was not large enough to hold them, Exim could crash while trying to process
3738 the truncated data. It now detects and logs this case.
3740 17. Further to 4.21/58, another change has been made: if (and only if) the
3741 first line of a message (the first header line) ends with CRLF, a bare LF
3742 in a subsequent header line has a space inserted after it, so as not to
3743 terminate the header.
3745 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3746 unnecessarily, used atoi() instead of strtol(), and didn't check the
3747 termination when getting file sizes from file names by regex.
3749 19. Completely re-implemented the support for maildirsize files, in the light
3750 of a number of problems with the previous contributed implementation
3751 (4.30/29). In particular:
3753 . If the quota is zero, the maildirsize file is maintained, but no quota is
3756 . If the maildir directory does not exist, it is created before any attempt
3757 to write a maildirsize file.
3759 . The quota value in the file is just a cache; if the quota is changed in
3760 the transport, the new value overrides.
3762 . A regular expression is available for excluding directories from the
3765 20. The autoreply transport checks the characters in options that define the
3766 message's headers; it allows continued headers, but it was checking with
3767 isspace() after an embedded newline instead of explicitly looking for a
3770 21. If all the "regular" hosts to which an address was routed had passed their
3771 expiry times, and had not reached their retry times, the address was
3772 bounced, even if fallback hosts were defined. Now Exim should go on to try
3775 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3776 equivalent code in the SMTP transport. Some hosts send humungous responses
3777 to HELO/EHLO, more than 1024 it seems.
3779 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3780 gives compiler warnings in some environments. I've now done it "properly",
3783 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3784 (because of problems with the built-in one) was declared to return uschar *
3785 instead of char *, causing compiler failure.
3787 25. Fixed a file descriptor leak when processing alias/forward files.
3789 26. Fixed a minor format string issue in dbfn.c.
3791 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3793 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3794 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3797 29. When a sender address is verified, it is cached, to save repeating the test
3798 when there is more than one recipient in a message. However, when the
3799 verification involves a callout, it is possible for different callout
3800 options to be set for different recipients. It is too complicated to keep
3801 track of this in the cache, so now Exim always runs a verification when a
3802 callout is required, relying on the callout cache for the optimization.
3803 The overhead is duplication of the address routing, but this should not be
3806 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3807 to be verified with callout=postmaster, and the main callout worked but the
3808 postmaster check failed, the verification correctly failed. However, if a
3809 subsequent RCPT command asked for sender verification *without* the
3810 postmaster check, incorrect caching caused this verification also to fail,
3813 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3814 it was not caching the DNS options (qualify_single, search_parents) that
3815 were used when the lookup failed. A subsequent lookup with different
3816 options therefore always gave the same answer, though there were cases
3817 where it should not have. (Example: a "domains = !$mx_any" option on a
3818 dnslookup router: the "domains" option is always processed without any
3819 widening, but the router might have qualify_single set.) Now Exim uses the
3820 cached value only when the same options are set.
3822 32. Added John Jetmore's "exipick" utility to the distribution.
3824 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3825 than a timeout (e.g. a certificate is required, and is not provided), an
3826 Exim server now closes the connection immediately. Previously it waited for
3827 the client to close - but if the client is SSL, it seems that they each
3828 wait for each other, leading to a delay before one of them times out.
3830 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3831 maintained 0.8.x compatibility because I don't think many are using it, and
3832 it is clearly obsolete.
3834 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3837 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3838 client certificate was expired. A simple patch fixes this, though I don't
3839 understand the full logic of why the verify callback is called multiple
3842 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3843 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3844 which causes problems with some clients (such as the Certicom SSL Plus
3845 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3846 disables the coutermeasure allowing Eudora to connect."
3848 38. Exim was not checking that a write() to a log file succeeded. This could
3849 lead to Bad Things if a log got too big, in particular if it hit a file
3850 size limit. Exim now panics and dies if it cannot write to a log file, just
3851 as it does if it cannot open a log file.
3853 39. Modified OS/Makefile-Linux so that it now contains
3855 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3857 The two -D definitions ensure that Exim is compiled with large file
3858 support, which makes it possible to handle log files that are bigger than
3861 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3862 instance) a domain was checked against a named list that involved a lookup,
3863 causing $domain_data to be set, then another domain was checked against the
3864 same list, then the first domain was re-checked, the value of $domain_data
3865 after the final check could be wrong. In particular, if the second check
3866 failed, it could be set empty. This bug probably also applied to
3869 41. The strip_trailing_dot option was not being applied to the address given
3870 with the -f command-line option.
3872 42. The code for reading a message's header from the spool was incrementing
3873 $received_count, but never initializing it. This meant that the value was
3874 incorrect (doubled) while delivering a message in the same process in which
3875 it was received. In the most common configuration of Exim, this never
3876 happens - a fresh exec is done - but it can happen when
3877 deliver_drop_privilege is set.
3879 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3880 it now includes up to 150 characters of the unexpected data in the log
3883 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3884 and building data strings. The size of both of these buffers was 10 000
3885 bytes - far larger than anybody would *ever* want, thought I. Needless to
3886 say, somebody hit the limit. I have increased the maximum line length to
3887 20 000 and the maximum data length of concatenated lines to 100 000. I have
3888 also fixed two bugs, because there was no checking on these buffers. Tsk,
3889 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3890 buffer is too small.
3892 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3893 lsearch lookups. Now it does.
3895 46. When parsing a route_list item in a manualroute router, a fixed-length
3896 buffer was used for the list of hosts. I made this 1024 bytes long,
3897 thinking that nobody would ever have a list of hosts that long. Wrong.
3898 Somebody had a whole pile of complicated expansion conditions, and the
3899 string was silently truncated, leading to an expansion error. It turns out
3900 that it is easier to change to an unlimited length (owing to other changes
3901 that have happened since this code was originally written) than to build
3902 structure for giving a limitation error. The length of the item that
3903 expands into the list of hosts is now unlimited.
3905 47. The lsearch lookup could not handle data where the length of text line was
3906 more than 4095 characters. Such lines were truncated, leading to shortened
3907 data being returned. It should now handle lines of any length.
3909 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3910 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3913 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3914 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3915 no longer happens in comment lines. A list of replacements is now placed
3916 at the head of all of the source files, except those whose only change is
3917 to replace PERL_COMMAND in the very first #! line.
3919 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3920 messages on the queue, with a bottom-up merge sort, using code contributed
3921 by Michael Haardt. This should make operations like -bp somewhat faster on
3922 large queues. It won't affect queue runners, except when queue_run_in_order
3925 51. Installed eximstats 1.31 in the distribution.
3927 52. Added support for SRV lookups to the dnslookup router.
3929 53. If an ACL referred to $message_body or $message_body_end, the value was not
3930 reset for any messages that followed in the same SMTP session.
3932 54. The store-handling optimization for building very long strings was not
3933 differentiating between the different store pools. I don't think this
3934 actually made any difference in practice, but I've tidied it.
3936 55. While running the routers to verify a sender address, $sender_address
3937 was still set to the sender address. This is wrong, because when routing to
3938 send a bounce to the sender, it would be empty. Therefore, I have changed
3939 it so that, while verifying a sender address, $sender_address is set to <>.
3940 (There is no change to what happens when verifying a recipient address.)
3942 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3943 target A or AAAA records (if not already returned) without resetting the
3944 qualify_single or search_parents options of the DNS resolver. These are
3945 inappropriate in this case because the targets of MX and SRV records must
3946 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3947 target that, when qualified, matched something in the local domain. These
3948 two options are now turned off when doing these lookups.
3950 57. It seems that at least some releases of Reiserfs (which does not have the
3951 concept of a fixed number of inodes) returns zero and not -1 for the
3952 number of available inodes. This interacted badly with check_spool_inodes,
3953 which assumed that -1 was the "no such thing" setting. What I have done is
3954 to check that the total number of inodes is greater than zero before doing
3955 the test of how many are available.
3957 58. When a "warn" ACL statement has a log_message modifier, the message is
3958 remembered, and not repeated. This is to avoid a lot of repetition when a
3959 message has many recipients that cause the same warning to be written.
3960 Howewer, Exim was preserving the list of already written lines for an
3961 entire SMTP session, which doesn't seem right. The memory is now reset if a
3962 new message is started.
3964 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3965 debugging output unless log_rewrite was also set.
3967 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3968 of (int)(handle) when we know that handle contains (void *)(-1).
3970 61. The Exim daemon panic-logs an error return when it closes the incoming
3971 connection. However "connection reset by peer" seems to be common, and
3972 isn't really an error worthy of noting specially, so that particular error
3975 62. When Exim is trying to find all the local interfaces, it used to panic and
3976 die if the ioctl to get the interface flags failed. However, it seems that
3977 on at least one OS (Solaris 9) it is possible to have an interface that is
3978 included in the list of interfaces, but for which you get a failure error
3979 for this call. This happens when the interface is not "plumbed" into a
3980 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3981 failure of the "get flags" call assumes that the interface is down.
3983 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3984 makes life easier for people who are doing arithmetic on fields extracted
3985 from dates, where you often get leading zeros that should not be
3986 interpreted as octal.
3988 64. Added qualify_domain to the redirect router, to override the global
3991 65. If a pathologically long header line contained very many addresses (the
3992 report of this problem mentioned 10 000) and each of them was rewritten,
3993 Exim could use up a very large amount of memory. (It kept on making new
3994 copies of the header line as it rewrote, and never released the old ones.)
3995 At the expense of a bit more processing, the header rewriting function has
3996 been changed so that it no longer eats memory in this way.
3998 66. The generation of the Received: header has been moved from the time that a
3999 message starts to be received, to the time that it finishes. The timestamp
4000 in the Received: header should now be very close to that of the <= log
4001 line. There are two side-effects of this change:
4003 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4004 logged header lines no longer include the local Received: line, because
4005 it has not yet been created. The same applies to a copy of the message
4006 that is returned to a non-SMTP sender when a message is rejected.
4008 (b) When a filter file is tested using -bf, no additional Received: header
4009 is added to the test message. After some thought, I decided that this
4012 This change does not affect the value of $received_for. It is still set
4013 after address rewriting, but before local_scan() is called.
4015 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4017 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4018 gave an unhelpful panic error message, and a defer error. I have managed to
4019 change this behaviour so that it now rejects any supplied certificate,
4020 which seems right, as the list of acceptable certificates is empty.
4022 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4023 gave an unhelpful defer error. I have not managed to make this reject any
4024 supplied certificates, but the error message it gives is "no certificate
4025 supplied", which is not helpful.
4027 70. exigrep's output now also includes lines that are not associated with any
4028 message, but which match the given pattern. Implemented by a patch from
4029 Martin Sluka, which also tidied up the Perl a bit.
4031 71. Recipient callout verification, like sender verification, was using <> in
4032 the MAIL FROM command. This isn't really the right thing, since the actual
4033 sender may affect whether the remote host accepts the recipient or not. I
4034 have changed it to use the actual sender in the callout; this means that
4035 the cache record is now keyed on a recipient/sender pair, not just the
4036 recipient address. There doesn't seem to be a real danger of callout loops,
4037 since a callout by the remote host to check the sender would use <>.
4038 [SEE ABOVE: changed after hitting problems.]
4040 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4041 temporary errors. However, in the case of such a code being given after
4042 the end of a data transmission (i.e. after ".") Exim was failing to write
4043 a retry record for the message. (Yes, there was some broken host that was
4044 actually sending 8xx at this point.)
4046 73. An unknown lookup type in a host list could cause Exim to panic-die when
4047 the list was checked. (An example that provoked this was putting <; in the
4048 middle of a list instead of at the start.) If this happened during a DATA
4049 ACL check, a -D file could be left lying around. This kind of configuration
4050 error no longer causes Exim to die; instead it causes a defer errror. The
4051 incident is still logged to the main and panic logs.
4053 74. Buglet left over from Exim 3 conversion. The message "too many messages
4054 in one connection" was written to the rejectlog but not the mainlog, except
4055 when address rewriting (yes!) was being logged.
4057 75. Added write_rejectlog option.
4059 76. When a system filter was run not as root (that is, when system_filter_user
4060 was set), the values of the $n variables were not being returned to the
4061 main process; thus, they were not subsequently available in the $sn
4064 77. Added +return_path_on_delivery log selector.
4066 78. A connection timeout was being treated differently from recipients deferred
4067 when testing hosts_max_try with a message that was older than the host's
4068 retry timeout. (The host should not be counted, thus allowing all hosts to
4069 be tried at least once before bouncing.) This may have been the cause of an
4070 occasionally reported bug whereby a message would remain on the queue
4071 longer than the retry timeout, but would be bounced if a delivery was
4072 forced. I say "may" because I never totally pinned down the problem;
4073 setting up timeout/retry tests is difficult. See also the next item.
4075 79. The ultimate address timeout was not being applied to errors that involved
4076 a combination of host plus message (for example, a timeout on a MAIL
4077 command). When an address resolved to a number of possible hosts, and they
4078 were not all tried for each delivery (e.g. because of hosts_max_try), a
4079 message could remain on the queue longer than the retry timeout.
4081 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4084 81. Fixed an obscure SMTP outgoing bug which required at least the following
4085 conditions: (a) there was another message waiting for the same server;
4086 (b) the server returned 5xx to all RCPT commands in the first message so
4087 that the message was not completed; (c) the server dropped the connection
4088 or gave a negative response to the RSET that Exim sends to abort the
4089 transaction. The observed case was a dropped connection after DATA that had
4090 been sent in pipelining mode. That is, the server had advertised PIPELINING
4091 but was not implementing it correctly. The effect of the bug was incorrect
4092 behaviour, such as trying another host, and this could lead to a crash.
4098 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4099 and daemon.c were passed as pointers to ints; they should have been
4100 pointers to socklen_t variables (which are typically unsigned ints).
4102 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4105 3. Fixed a really odd bug that affected only the testing scheme; patching a
4106 certain fixed string in the binary changed the value of another string that
4107 happened to be identical to the end of the original first string.
4109 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4110 name", it returns that address as the IP address. On some operating
4111 systems (e.g. Solaris), it also passes back the IP address string as the
4112 "host name". However, on others (e.g. Linux), it passes back an empty
4113 string. Exim wasn't checking for this, and was changing the host name to an
4114 empty string, assuming it had been canonicized.
4116 5. Although rare, it is permitted to have more than one PTR record for a given
4117 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4118 all the names associated with an address, because they do in Solaris.
4119 However, it seems that they do not in Linux for data that comes from the
4120 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4121 I found this out when I moved to a new Linux workstation and tried to run
4122 the Exim test suite.
4124 To get round this problem I have changed the code so that it now does its
4125 own call to the DNS to look up PTR records when searching for a host name.
4126 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4127 addresses that are only in /etc/hosts are still found.
4129 This behaviour is, however, controlled by an option called host_lookup_
4130 order, which defaults to "bydns:byaddr". If people want to use the other
4131 order, or indeed, just use one or the other means of lookup, they can
4132 specify it in this variable.
4134 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4135 some operating systems, this comes back from gethostbyaddr() as an empty
4136 string, and this is what Exim used to test for. However, it seems that in
4137 other systems, "." is yielded. Exim now tests for this case too.
4139 7. The values of check_spool_space and check_log_space are now held internally
4140 as a number of kilobytes instead of an absolute number of bytes. If a
4141 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4142 kilobyte. This means that much larger values can be stored.
4144 8. Exim monitor: an attempt to get the action menu when not actually pointing
4145 at a message produces an empty menu entitled "No message selected". This
4146 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4147 no entries in it ("Shell widget menu has zero width and/or height"). So I
4148 have added a single, blank menu entry in this case.
4150 9. Added ${quote_local_part.
4152 10. MIME decoding is now applied to the contents of Subject: header lines when
4155 11. Now that a reference to $sender_host_address automatically causes a reverse
4156 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4157 host lookup before query-style lookups in lists that might use this
4158 variable. This has therefore been abolished, and the "net-" prefix is no
4159 longer necessary for query-style lookups.
4161 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4162 have been a typo for LFLAGS, so it has been changed.
4164 13. The install script calls Exim with "-C /dev/null" in order to find the
4165 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4166 to be output. Howeve, since Exim outputs its version number before the
4167 error, it didn't break the script. It just looked ugly. I fixed this by
4168 always allowing "-C /dev/null" if the caller is root.
4170 14. Ignore overlarge ACL variable number when reading spool file - insurance
4171 against a later release with more variables having written the file.
4173 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4174 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4175 this, as well as the form without the "IPv6" on the front.
4177 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4178 OS/Makefile-Darwin file.
4180 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4181 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4183 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4184 variable. However, some OS use ints for inodes. I've added cast to long int
4185 to get rid of the compiler warning.
4187 19. I had forgotten to lock out "/../" in configuration file names when
4188 ALT_CONFIG_PREFIX was set.
4190 20. Routers used for verification do not need to specify transports. However,
4191 if such a router generated a host list, and callout was configured, Exim
4192 crashed, because it could not find a port number from the (non-existent)
4193 transport. It now assumes port 25 in this circumstance.
4195 21. Added the -t option to exigrep.
4197 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4198 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4199 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4200 them is set, LOOKUP_LSEARCH is forced.
4202 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4204 24. Added sender and host information to the "rejected by local_scan()" log
4205 line; previously there was no indication of these.
4207 25. Added .include_if_exists.
4209 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4210 Linux. It turns out that not all file systems support this. Apparently some
4211 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4212 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4213 not supported on the file descriptor, is now ignored when Exim is trying to
4214 sync a directory. This applies only to Linux.
4216 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4218 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4219 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4222 29. Added support for maildirsize files from supplied patch (modified a bit).
4224 30. The use of :fail: followed by an empty string could lead Exim to respond to
4225 sender verification failures with (e.g.):
4227 550 Verification failed for <xxx>
4228 550 Sender verify failed
4230 where the first response line was missing the '-' that indicates it is not
4231 the final line of the response.
4233 31. The loop for finding the name of the user that called Exim had a hardwired
4234 limit of 10; it now uses the value of finduser_retries, which is used for
4235 all other user lookups.
4237 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4240 33. Exim was neglecting to zero errno before one call of strtol() when
4241 expanding a string and expecting an integer value. On some systems this
4242 resulted in spurious "integer overflow" errors. Also, it was casting the
4243 result into an int without checking.
4245 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4246 did not work. The code looks as if it has *never* worked, though it appears
4247 to have been documented since at least releast 1.62. I have made it work.
4249 35. The "timeout_DNS" error in retry rules, also documented since at least
4250 1.62, also never worked. As it isn't clear exactly what this means, and
4251 clearly it isn't a major issue, I have abolished the feature by treating it
4252 as "timeout", and writing a warning to the main and panic logs.
4254 36. The display of retry rules for -brt wasn't always showing the error code
4257 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4258 timeout_connect_A, timeout_connect_MX.
4260 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4261 to the empty sender.
4263 39. The daemon was not analysing the content of -oX till after it had closed
4264 stderr and disconnected from the controlling terminal. This meant that any
4265 syntax errors were only noted on the panic log, and the return code from
4266 the command was 0. By re-arranging the code a little, I've made the
4267 decoding happen first, so such errors now appear on stderr, and the return
4268 code is 1. However, the actual setting up of the sockets still happens in
4269 the disconnected process, so errors there are still only recorded on the
4272 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4273 connections (as happens on some IP stacks) was logged at start up time as
4274 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4275 it from "IPv6 and IPv4", which means that two separate sockets are being
4278 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4279 says whether AF_INET or AF_INET6 was passed as an argument.
4281 42. Exiwhat output was messed up when time zones were included in log
4284 43. Exiwhat now gives more information about the daemon's listening ports,
4285 and whether -tls-on-connect was used.
4287 44. The "port" option of the smtp transport is now expanded.
4289 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4290 silently ignored. Now an error message is written to the main and panic
4293 46. There's a new ACL modifier called "logwrite" which writes to a log file
4294 as soon as it is encountered.
4296 47. Added $local_user_uid and $local_user_gid at routing time.
4298 48. Exim crashed when trying to verify a sender address that was being
4301 49. Exim was recognizing only a space character after ".include". It now also
4302 recognizes a tab character.
4304 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4305 extracting the relevant information from the specification. The man page no
4306 longer contains scrambled data for the -d option, and I've added a section
4307 at the front about calling Exim under different names.
4309 51. Added "extra_headers" argument to the "mail" command in filter files.
4311 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4314 53. Installed eximstats 1.29.
4316 54. Added transport_filter_timeout as a generic transport option.
4318 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4319 Cc: header lines. This was required by RFC 822, but it not required by RFC
4322 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4323 incoming messages that did not have them. Now it does so only if the
4324 message originates locally, that is, if there is no associated remote host
4325 address. When Resent- header lines are present, this applies to the Resent-
4326 lines rather than the non-Resent- lines.
4328 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4329 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4331 58. Messages for configuration errors now include the name of the main
4332 configuration file - useful now that there may be more than one file in a
4333 list (.included file names were always shown).
4335 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4336 for those rare installations that do not start the daemon as root or run it
4337 setuid root. I've cut out the call to initgroups() if the daemon is not
4340 60. The Exim user and group can now be bound into the binary as text strings
4341 that are looked up at the start of Exim's processing.
4343 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4345 62. Added $mailstore_basename variable.
4347 63. Installed patch to sieve.c from Michael Haardt.
4349 64. When Exim failed to open the panic log after failing to open the main log,
4350 the original message it was trying to log was written to stderr and debug
4351 output, but if they were not available (the usual case in production), it
4352 was lost. Now it is written to syslog before the two lines that record the
4353 failures to open the logs.
4355 65. Users' Exim filters run in subprocesses under the user's uid. It is
4356 possible for a "deliver" command or an alias in a "personal" command to
4357 provoke an address rewrite. If logging of address rewriting is configured,
4358 this fails because the process is not running as root or exim. There may be
4359 a better way of dealing with this, but for the moment (because 4.30 needs
4360 to be released), I have disabled address rewrite logging when running a
4361 filter in a non-root, non-exim process.
4367 1. The buildconfig auxiliary program wasn't quoting the value set for
4368 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4369 not defined. This bug was masked in 4.22 by the effect that was fixed in
4372 2. Some messages that were rejected after a message id was allocated were
4373 shown as "incomplete" by exigrep. It no longer does this for messages that
4374 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4376 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4377 have allow_domain_literals set, the ID did not get logged in the <= line.
4378 Domain literals are now always recognized in Message-ID: header lines.
4380 4. The first argument for a ${extract expansion item is the key name or field
4381 number. Leading and trailing spaces in this item were not being ignored,
4382 causing some misleading effects.
4384 5. When deliver_drop_privilege was set, single queue runner processes started
4385 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4386 same command in the process it spins off) were not dropping privilege.
4388 6. When the daemon running as "exim" started a queue runner, it always
4389 re-executed Exim in the spun-off process. This is a waste of effort when
4390 deliver_drop_privilege is set. The new process now just calls the
4391 queue-runner function directly.
4397 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4400 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4401 ignored. Though the use of -oP was forcing the writing of a pid file, it
4402 was always written to the default place.
4404 3. If the message "no IP address found for host xxxx" is generated during
4405 incoming verification, it is now followed by identification of the incoming
4406 connection (so you can more easily find what provoked it).
4408 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4410 5. Added some features to "harden" Exim a bit more against certain attacks:
4412 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4413 be put in Local/Makefile. This is like the never_users runtime option,
4414 but it cannot be overridden. The default setting is "root".
4416 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4417 prefix string with which any file named in a -C command line option
4420 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4421 is retained for -C and -D only if the caller of Exim is root. Without
4422 it, the exim user may also use -C and -D and retain privilege.
4424 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4425 command line option is disabled.
4427 6. Macro names set by the -D option must start with an upper case letter, just
4428 like macro names defined in the configuration file.
4430 7. Added "dereference=" facility to LDAP.
4432 8. Two instances of the typo "uknown" in the source files are fixed.
4434 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4435 the Configure-Makefile script screwed up while processing it.
4437 10. Incorporated PCRE 4.4.
4439 11. The SMTP synchronization check was not operating right at the start of an
4440 SMTP session. For example, it could not catch a HELO sent before the client
4441 waited for the greeting. There is now a check for outstanding input at the
4442 point when the greeting is written. Because of the duplex, asynchronous
4443 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4444 way, but not yet received, when the check is performed.
4446 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4447 on TCP/IP sockets, because this apparently causes some broken clients to
4450 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4451 unchanged) from the Cygwin maintainer.
4453 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4454 was supported instead of testing for mbx. Effectively a typo.
4456 15. The spa authenticator server code was not checking that the input it
4457 received was valid base64.
4459 16. The debug output line for the "set" modifier in ACLs was not showing the
4460 name of the variable that was being set.
4462 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4464 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4465 lookup on its own. Something else was needed to trigger the lookup. For
4466 example, a match in host_lookup or the need for a host name in a host list.
4467 Now, if $sender_host_name is referenced and the host name has not yet been
4468 looked up, a lookup is performed. If the lookup fails, the variable remains
4469 empty, and $host_lookup_failed is set to "1".
4471 19. Added "eqi" as a case-independent comparison operator.
4473 20. The saslauthd authentication condition could segfault if neither service
4474 nor realm was specified.
4476 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4477 error message that was logged was misleading, and incoming SMTP
4478 connections were dropped. The message is now more accurate, and temporary
4479 errors are given to SMTP connections.
4481 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4482 (except RSET) with a 421 error, until QUIT is received. However, it was
4483 failing to send a response to QUIT.
4485 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4486 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4487 ACL is now run after the helo_try_verify_hosts code.
4489 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4490 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4491 case-independent, so other case variants are also recognized). Apparently
4492 some systems use these upper case variants.
4494 25. If more than two messages were waiting for the same host, and a transport
4495 filter was specified for the transport, Exim sent two messages over the
4496 same TCP/IP connection, and then failed with "socket operation on non-
4497 socket" when it tried to send the third.
4499 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4501 27. The extern definition of crypt16() in expand.c was not being excluded when
4502 the OS had its own crypt16() function.
4504 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4505 as a preferred synonym for return_size_limit, both as an option and as an
4508 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4510 30. Changed the default configuration ACL to relax the local part checking rule
4511 for addresses that are not in any local domains. For these addresses,
4512 slashes and pipe symbols are allowed within local parts, but the sequence
4513 /../ is explicitly forbidden.
4515 31. SPA server authentication was not clearing the challenge buffer before
4518 32. log_message in a "warn" ACL statement was writing to the reject log as
4519 well as to the main log, which contradicts the documentation and doesn't
4520 seem right (because no rejection is happening). So I have stopped it.
4522 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4523 However, I am unable to do any testing of this.
4525 34. Fixed an infelicity in the appendfile transport. When checking directories
4526 for a mailbox, to see if any needed to be created, it was accidentally
4527 using path names with one or more superfluous leading slashes; tracing
4528 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4530 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4531 contents are added to the log line that is written for every discarded
4532 recipient. (Previously a log_message setting was ignored.)
4534 36. The ${quote: operator now quotes the string if it is empty.
4536 37. The install script runs exim in order to find its version number. If for
4537 some reason other than non-existence or emptiness, which it checks, it
4538 could not run './exim', it was installing it with an empty version number,
4539 i.e. as "exim-". This error state is now caught, and the installation is
4542 38. An argument was missing from the function that creates an error message
4543 when Exim fails to connect to the socket for saslauthd authentication.
4544 This could cause Exim to crash, or give a corrupted message.
4546 39. Added isip, isip4, and isip6 to ${if conditions.
4548 40. The ACL variables $acl_xx are now saved with the message, and can be
4549 accessed later in routers, transports, and filters.
4551 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4552 strings in the file are not string-expanded.
4554 42. If a MAIL command specified a SIZE value that was too large to fit into an
4555 int variable, the check against message_size_limit failed. Such values are
4556 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4557 day this will have to be increased, but I don't think I want to be around
4558 when emails are that large.
4565 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4566 iconv() is not standard in FreeBSD.
4568 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4569 IPv6 enabled. The observed symptom was a segmentation fault on return from
4570 the function os_common_find_running_interfaces() in src/os.c.
4572 3. In the check_special_case() function in daemon.c I had used "errno" as an
4573 argument name, which causes warnings on some systems. This was basically a
4574 typo, since it was named "eno" in the comments!
4576 4. The code that waits for the clock to tick (at a resolution of some fraction
4577 of a second) so as to ensure message-id uniqueness was always waiting for
4578 at least one whole tick, when it could have waited for less. [This is
4579 almost certainly not relevant at current processor speeds, where it is
4580 unlikely to ever wait at all. But we try to future-proof.]
4582 5. The function that sleeps for a time interval that includes fractions of a
4583 second contained a race. It did not block SIGALRM between setting the
4584 timer, and suspending (a couple of lines later). If the interval was short
4585 and the sigsuspend() was delayed until after it had expired, the suspension
4586 never ended. On busy systems this could lead to processes getting stuck for
4589 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4590 process, before it forks any delivery processes. The open lookup caching
4591 mechanism meant that the open file or database connection was passed into
4592 the delivery process. The problem was that delivery processes always tidy
4593 up cached lookup data. This could cause a problem for the next delivery
4594 process started by the queue runner, because the external queue runner
4595 process does not know about the closure. So the next delivery process
4596 still has data in the lookup cache. In the case of a file lookup, there was
4597 no problem because closing a file descriptor in a subprocess doesn't affect
4598 the parent. However, if the lookup was caching a connection to a database,
4599 the connection was closed, and the second delivery process was likely to
4600 see errors such as "PGSQL: query failed: server closed the connection
4601 unexpectedly". The problem has been fixed by closing all cached lookups
4602 in a queue runner before running a delivery process.
4604 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4605 seem to have the "const" qualifier which it has on other OS. I've
4608 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4609 *of the same type* (client or server) with the same public name that an
4610 error should be diagnosed.
4612 9. When Exim looked up a host name for an IP address, but failed to find the
4613 original IP address when looking up the host name (a safety check), it
4614 output the message "<ip address> does not match any IP for NULL", which was
4615 confusing, to say the least. The bug was that the host name should have
4616 appeared instead of "NULL".
4618 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4619 user that is built into the binary, and the -C or -D options is used, root
4620 privilege is dropped before the configuration file is read. In addition,
4621 logging is switched to stderr instead of the normal log files. If the
4622 configuration then re-defines the Exim user, the unprivileged environment
4623 is probably not what is expected, so Exim logs a panic warning message (but
4626 However, if deliver_drop_privilege is set, the unprivileged state may well
4627 be exactly what is intended, so the warning has been cut out in that case,
4628 and Exim is allowed to try to write to its normal log files.
4634 1. smtp_return_error_details was not giving details for temporary sender
4635 or receiver verification errors.
4637 2. Diagnose a configuration error if two authenticators have the same public
4640 3. Exim used not to create the message log file for a message until the first
4641 delivery attempt. This could be confusing when incoming messages were held
4642 for policy or load reasons. The message log file is now created at the time
4643 the message is received, and an initial "Received" line is written to it.
4645 4. The automatically generated man page for command line options had a minor
4646 bug that caused no ill effects; however, a more serious problem was that
4647 the procedure for building the man page automatically didn't always
4648 operate. Consequently, release 4.20 contains an out-of-date version. This
4649 shouldn't happen again.
4651 5. When building Exim with embedded Perl support, the script that builds the
4652 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4653 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4655 6. The freeze_tell option was not being used for messages that were frozen on
4656 arrival, either by an ACL or by local_scan().
4658 7. Added the smtp_incomplete_transaction log selector.
4660 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4661 was accepting AUTH without a new EHLO.
4663 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4664 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4667 10. Logging of TCP/IP connections (when configured) now happens in the main
4668 daemon process instead of the child process, so that the TCP/IP connection
4669 count is more accurate (but it can never be perfect).
4671 11. The use of "drop" in a nested ACL was not being handled correctly in the
4672 outer ACL. Now, if condition failure induced by the nested "drop" causes
4673 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4674 or "require"), the connection is dropped.
4676 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4677 that yield "discard" can now be used with an "accept" or a "discard" verb,
4678 but an error is generated for any others (because I can't see a useful way
4679 to define what should happen).
4681 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4682 are now processed in the same way as lines in the Exim configuration file.
4683 In particular, continuation lines are supported.
4685 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4687 15. Added -ti meaning -t -i.
4689 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4690 domains, and warn by logging if others are found.
4692 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4693 returned from GIFCONF when Exim is trying to find the list of interfaces on
4694 a host. The code in os.c has been modified to copy each ifreq to an aligned
4695 structure in all cases.
4697 Also, in some cases, the returned ifreq's were being copied to a 'struct
4698 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4699 means the last couple of bytes of an IPv6 address could be chopped if the
4700 ifreq contained only a normal sockaddr (14 bytes storage).
4702 18. Named domain lists were not supported in the hosts_treat_as_local option.
4703 An entry such as +xxxx was not recognized, and was treated as a literal
4706 19. Ensure that header lines added by a DATA ACL are included in the reject log
4707 if the ACL subsequently rejects the message.
4709 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4710 MD5 (which is deprecated).
4712 21. When testing a filter file using -bf, Exim was writing a message when it
4713 took the sender from a "From " line in the message, but it was not doing so
4714 when it took $return_path from a Return-Path: header line. It now does.
4716 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4717 with a valid header line field name (a series of printing characters
4718 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4720 23. Changed "disc" in the source to "disk" to conform to the documentation and
4721 the book and for uniformity.
4723 24. Ignore Sendmail's -Ooption=value command line item.
4725 25. When execve() failed while trying to run a command in a pipe transport,
4726 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4727 could be confused with a return value of 69 from the command itself. This
4728 has been changed to 127, the value the shell returns if it is asked to run
4729 a non-existent command. The wording for the related log line suggests a
4730 non-existent command as the problem.
4732 26. If received_header_text expands to an empty string, do not add a Received:
4733 header line to the message. (Well, it adds a token one on the spool, but
4734 marks it "old" so that it doesn't get used or transmitted.)
4736 27. Installed eximstats 1.28 (addition of -nt option).
4738 28. There was no check for failure on the call to getsockname() in the daemon
4739 code. This can fail if there is a shortage of resources on the system, with
4740 ENOMEM, for example. A temporary error is now given on failure.
4742 29. Contrary to the C standard, it seems that in some environments, the
4743 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4744 program. Exim now does this explicitly; it affects the formatting of
4745 timestamps using strftime().
4747 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4748 complaints. I've now initialized all the variables, to avoid this.
4750 32. Header lines added by a system filter were not being "seen" during
4751 transport-time rewrites.
4753 33. The info_callback() function passed to OpenSSL is set up with type void
4754 (*)(SSL *, int, int), as described somewhere. However, when calling the
4755 function (actually a macro) that sets it up, the type void(*)() is
4756 expected. I've put in a cast to prevent warnings from picky compilers.
4758 34. If a DNS black list lookup found a CNAME record, but there were no A
4759 records associated with the domain it pointed at, Exim crashed.
4761 35. If a DNS black list lookup returned more than one A record, Exim ignored
4762 all but the first. It now scans all returned addresses if a particular IP
4763 value is being sought. In this situation, the contents of the
4764 $dnslist_value variable are a list of all the addresses, separated by a
4767 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4768 to accept a host name and all its aliases if the forward lookup for any of
4769 them yielded the IP address of the incoming connection. Now it accepts only
4770 those names whose forward lookup yields the correct IP address. Any other
4771 names are discarded. This closes a loophole whereby a rogue DNS
4772 administrator could create reverse DNS records to break through a
4773 wildcarded host restriction in an ACL.
4775 37. If a user filter or a system filter that ran in a subprocess used any of
4776 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4777 the wrong values were passed to the pipe command ($thisaddress had the
4778 value of $0, $0 had the value of $1, etc). This bug was introduced by
4779 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4781 38. Improved the line breaking for long SMTP error messages from ACLs.
4782 Previously, if there was no break point between 40 and 75 characters, Exim
4783 left the rest of the message alone. Two changes have been made: (a) I've
4784 reduced the minimum length to 35 characters; (b) if it can't find a break
4785 point between 35 and 75 characters, it looks ahead and uses the first one
4786 that it finds. This may give the occasional overlong line, but at least the
4787 remaining text gets split now.
4789 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4790 file descriptors might be low, and that setting 1000 would always raise it.
4791 It turns out that in some environments, the limit is already over 1000 and
4792 that lowering it causes trouble. So now Exim takes care not to decrease it.
4794 40. When delivering a message, the value of $return_path is set to $sender_
4795 address at the start of routing (routers may change the value). By an
4796 oversight, this default was not being set up when an address was tested by
4797 -bt or -bv, which affected the outcome if any router or filter referred to
4800 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4801 log only when "message" or "log_message" are set. However, if one of the
4802 conditions was an address verification, or a call to a nested ACL, the
4803 messages generated by the underlying test were being passed through. This
4804 no longer happens. The underlying message is available in $acl_verify_
4805 message for both "message" and "log_message" expansions, so it can be
4806 passed through if needed.
4808 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4809 new expansion $bh_ to give the encoded byte string without charset
4810 translation. Translation happens only if iconv() is available; HAVE_ICONV
4811 indicates this at build time. HEADERS_CHARSET gives the charset to
4812 translate to; headers_charset can change it in the configuration, and
4813 "headers charset" can change it in an individual filter file.
4815 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4816 that creates RFC 2047 encoded "words" labels them as that charset instead
4817 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4818 expansion operator; (ii) when Exim creates a From: line for a local
4819 message; (iii) when a header line is rewritten to include a "phrase" part.
4821 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4822 buggy, causing it to skip the first lines of messages whose message ID
4823 ended in 'D'. This would not have bitten before Exim release 4.14, because
4824 message IDs were unlikely to end in 'D' before then. The effect was to have
4825 incorrect size information for certain domains.
4827 45. #include "config.h" was missing at the start of the crypt16.c module. This
4828 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4831 46. If there was a timeout during a "random" callout check, Exim treated it as
4832 a failure of the random address, and carried on sending RSET and the real
4833 address. If the delay was just some slowness somewhere, the response to the
4834 original RCPT would be taken as a response to RSET and so on, causing
4835 mayhem of various kinds.
4837 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4838 when I implemented it. It didn't allow for the fact that some option values
4839 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4840 the right test for positive values.
4842 48. Domain names in DNS records are case-independent. Exim always looks them up
4843 in lower case. Some resolvers return domain names in exactly the case they
4844 appear in the zone file, that is, they may contain uppercase letters. Not
4845 all resolvers do this - some return always lower case. Exim was treating a
4846 change of case by a resolver as a change of domain, similar to a widening
4847 of a domain abbreviation. This triggered its re-routing code and so it was
4848 trying to route what was effectively the same domain again. This normally
4849 caused routing to fail (because the router wouldn't handle the domain
4850 twice). Now Exim checks for this case specially, and just changes the
4851 casing of the domain that it ultimately uses when it transmits the message
4854 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4857 50. If a filter generated a file delivery with a non-absolute name (possible if
4858 no home directory exists for the router), the forbid_file option was not
4861 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4862 the existing equality matching.
4864 52. Exim was using ints instead of ino_t variables in some places where it was
4865 dealing with inode numbers.
4867 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4868 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4869 called TMPDIR, and if it finds it is different, it changes its value.
4871 54. The smtp_printf() function is now made available to local_scan() so
4872 additional output lines can be written before returning. There is also an
4873 smtp_fflush() function to enable the detection of a dropped connection.
4874 The variables smtp_input and smtp_batched_input are exported to
4877 55. Changed the default runtime configuration: the message "Unknown user"
4878 has been removed from the ACL, and instead placed on the localuser router,
4879 using the cannot_route_message feature. This means that any verification
4880 failures that generate their own messages won't get overridden. Similarly,
4881 the "Unrouteable address" message that was in the ACL for unverifiable
4882 relay addresses has also been removed.
4884 56. Added hosts_avoid_esmtp to the smtp transport.
4886 57. The exicyclog script was not checking for the esoteric option
4887 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4888 will work only if exicyclog is run under the appropriate euid.
4890 58. Following a discussion on the list, the rules by which Exim recognises line
4891 endings on incoming messages have been changed. The -dropcr and drop_cr
4892 options are now no-ops, retained only for backwards compatibility. The
4893 following line terminators are recognized: LF CRLF CR. However, special
4894 processing applies to CR:
4896 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4897 nor a local message in the state where . is a terminator.
4899 (ii) If a bare CR is encountered in a header line, an extra space is added
4900 after the line terminator so as not to end the header. The reasoning
4901 behind this is that bare CRs in header lines are most likely either
4902 to be mistakes, or people trying to play silly games.
4904 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4905 was being incorrectly given as 18 bytes larger than it should have been.
4906 This is a VOB (very old bug).
4908 60. This may never have affected anything current, but just in case it has:
4909 When the local host is found other than at the start of a list of hosts,
4910 the local host, those with the same MX, and any that follow, are discarded.
4911 When the list in question was part of a longer list of hosts, the following
4912 hosts (not currently being processed) were also being discarded. This no
4913 longer happens. I'm not sure if this situation could ever has previously
4916 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4919 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4920 header. This is something that is recommended in a new Internet Draft, and
4921 is something that is documented as being done by Sendmail. There are two
4922 possible values. For messages generated by the autoreply transport, Exim
4925 Auto-Submitted: auto-replied
4927 whereas for all other generated messages (e.g. bounces) it adds
4929 Auto-Submitted: auto-generated
4931 63. The "personal" condition in filters now includes a test for the
4932 Auto-Submitted: header. If it contains the string "auto-" the message it
4933 not considered personal.
4935 64. Added rcpt_include_affixes as a generic transport option.
4937 65. Added queue_only_override (default true).
4939 66. Added the syslog_duplication option.
4941 67. If what should have been the first header line of a message consisted of
4942 a space followed by a colon, Exim was mis-interpreting it as a header line.
4943 It isn't of course - it is syntactically invalid and should therefore be
4944 treated as the start of the message body. The misbehaviour could have
4945 caused a number of strange effects, including loss of data in subsequent
4946 header lines, and spool format errors.
4948 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4949 client host had authenticated. This control can now be exercised by an ACL
4950 for more flexibility.
4952 69. By default, callouts do not happen when testing with -bh. There is now a
4953 variant, -bhc, which does actually run the callout code, including
4954 consulting and updating the callout cache.
4956 70. Added support for saslauthd authentication, courtesy of Alexander
4959 71. If statvfs() failed on the spool or log directories while checking their
4960 size for availability, Exim confusingly gave the error "space shortage".
4961 Furthermore, in debugging mode it crashed with a floating point exception.
4962 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4963 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4964 serious problem, Exim now writes to the main and panic logs when this
4965 happens, with details of the failure. It then refuses to accept the
4966 incoming message, giving the message "spool directory problem" or "log
4967 directory problem" with a 421 code for SMTP messages.
4969 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4970 0, 1, and 2 exist, because some OS complain for execs without them (see
4971 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4972 descriptors. However, the code omitted to check that the open succeeded,
4973 causing mysterious errors if for some reason the permissions on /dev/null
4974 got screwed. Now Exim writes a message to the main and panic logs, and
4975 bombs out if it can't open /dev/null.
4977 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4978 interact so that it is all more flexible. It is supposed to remain
4979 backwards compatible. Also added extra_local_interfaces.
4981 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4982 to bomb out with an assertion failure - to the client this appears as a
4983 connection drop. This problem occurs in the part of the code that was taken
4984 from the Samba project. Fortunately, the assertion is in a very simple
4985 function, so I have fixed this by reproducing the function inline in the
4986 one place where it is called, and arranging for authentication to fail
4987 instead of killing the process with assert().
4989 75. The SPA client code was not working when the server requested OEM rather
4990 than Unicode encoding.
4992 76. Added code to make require_files with a specific uid setting more usable in
4993 the case where statting the file as root fails - usually a non-root-mounted
4994 NFS file system. When this happens and the failure is EACCES, Exim now
4995 forks a subprocess and does the per-uid checking as the relevant uid.
4997 77. Added process_log_path.
4999 78. If log_file_path was not explicitly set, a setting of check_log_space or
5000 check_log_inodes was ignored.
5002 79. If a space check for the spool or log partitions fails, the incident is now
5003 logged. Of course, in the latter case the data may get lost...
5005 80. Added the %p formatting code to string_format() so that it can be used to
5006 print addresses in debug_print(). Adjusted all the address printing in the
5007 debugging in store.c to use %p rather than %d.
5009 81. There was a concern that a line of code in smtp_in.c could overflow a
5010 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5011 initially expressed, the concern was not well-founded, because trailing
5012 spaces are removed early. However, if the trailing spaces were followed by
5013 a NULL, they did not get removed, so the overflow was possible. Two fixes
5016 (a) I re-wrote the offending code in a cleaner fashion.
5017 (b) If an incoming SMTP command contains a NULL character, it is rejected
5020 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5021 runs initgroups(), so that if the Exim user is in any additional groups,
5022 they will be used during message reception.
5028 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob