1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.595 2009/11/16 19:10:45 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
11 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
17 TK/01 Added patch by Johannes Berg that expands the main option
18 "spamd_servers" if it starts with a dollar sign.
20 TK/02 Write list of recipients to X-Envelope-Sender header when building
21 the mbox-format spool file for content scanning (suggested by Jakob
24 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
25 (http://www.dcc-servers.net/) support via dccifd. Activated by
26 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
27 experimental_spec.txt for more documentation.
29 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
30 by Mark Daniel Reidel <mr@df.eu>.
32 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
33 When building exim an external PCRE library is now needed -
34 PCRE is a system library on the majority of modern systems.
35 See entry on PCRE_LIBS in EDITME file.
37 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
38 conversation. Added nologin parameter to request.
39 Patch contributed by Kirill Miazine
41 TF/01 Do not log submission mode rewrites if they do not change the address.
43 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
45 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
46 log files in place. Contributed by Roberto Lima
48 NM/04 Bugzilla 667: close socket used by dovecot authenticator
50 TF/03 Bugzilla 615: When checking the local_parts router precondition
51 after a local_part_suffix or local_part_prefix option, Exim now
52 does not use the address's named list lookup cache, since this
53 contains cached lookups for the whole local part.
55 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
56 Robert Millan. Documentation is in experimental-spec.txt
58 TF/04 Bugzilla 668: Fix parallel build (make -j).
60 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
62 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
63 Patch provided by Jan Srzednicki
65 TF/05 Leading white space used to be stripped from $spam_report which
66 wrecked the formatting. Now it is preserved.
68 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
69 that they are available at delivery time.
71 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
73 TF/08 TLS error reporting now respects the incoming_interface and
74 incoming_port log selectors.
76 TF/09 Produce a more useful error message if an SMTP transport's hosts
77 setting expands to an empty string.
79 NM/06 Bugzilla 744: EXPN did not work under TLS.
80 Patch provided by Phil Pennock
82 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
83 Patch provided by Richard Godbee
85 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
86 acl_smtp_notquit, added index entry.
88 NM/09 Bugzilla 787: Potential buffer overflow in string_format
89 Patch provided by Eugene Bujak
91 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
92 Patch provided by Maxim Dounin
94 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
95 Patch provided by Phil Pennock
97 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
99 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
100 Patch provided by Brad "anomie" Jorsch
102 NM/14 Bugzilla 622: Added timeout setting to transport filter.
103 Patch provided by Dean Brooks
105 TK/05 Add native DKIM support (does not depend on external libraries).
107 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
108 Patch provided by Graeme Fowler
110 NM/16 Bugzilla 851: Documentation example syntax fix.
112 NM/17 Changed NOTICE file to remove references to embedded PCRE.
114 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
116 NM/19 Bugzilla 745: TLS version reporting
117 Patch provided by Phil Pennock
119 NM/20 Bugzilla 167: bool: condition support
120 Patch provided by Phil Pennock
122 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
123 Patch provided by Phil Pennock
125 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
126 Patch provided by Brad "anomie" Jorsch
128 NM/23 Bugzilla 687: Fix misparses in eximstats
129 Patch provided by Heiko Schlittermann
131 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
132 Patch provided by Heiko Schlittermann
134 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
135 plus update to original patch.
137 NM/26 Bugzilla 799: Documentation correction for ratelimit
139 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
140 Patch provided by David Brownlee
142 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
144 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
146 NM/30 Bugzilla 888: TLS documentation bugfixes
148 NM/31 Bugzilla 896: Dovecot buffer overrun fix
150 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
151 Unlike the original bugzilla I have changed all shell scripts in src tree
153 NM/33 Bugzilla 898: Transport filter timeout fix
154 Patch by Todd Rinaldo
156 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
157 Patch by Serge Demonchaux
159 NM/35 Bugzilla 39: Base64 decode bug fixes
160 Patch by Jakob Hirsch
162 NM/36 Bugzilla 909: Correct connect() call in dcc code
164 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
166 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
168 NM/39 Bugzilla 911: Fixed MakeLinks build script
174 TK/01 Add preliminary DKIM support. Currently requires a forked version of
175 ALT-N's libdkim that I have put here:
176 http://duncanthrax.net/exim-experimental/
178 Note to Michael Haardt: I had to rename some vars in sieve.c. They
179 were called 'true' and it seems that C99 defines that as a reserved
180 keyword to be used with 'bool' variable types. That means you could
181 not include C99-style headers which use bools without triggering
182 build errors in sieve.c.
184 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
185 as mailq or other aliases. Changed the --help handling significantly
186 to do whats expected. exim_usage() emits usage/help information.
188 SC/01 Added the -bylocaldomain option to eximstats.
190 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
192 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
194 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
200 PH/01 Another patch from the Sieve maintainer.
202 PH/02 When an IPv6 address is converted to a string for single-key lookup
203 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
204 dots are used instead of colons so that keys in lsearch files need not
205 contain colons. This was done some time before quoting was made available
206 in lsearch files. However, iplsearch files do require colons in IPv6 keys
207 (notated using the quote facility) so as to distinguish them from IPv4
208 keys. This meant that lookups for IP addresses in host lists did not work
209 for iplsearch lookups.
211 This has been fixed by arranging for IPv6 addresses to be expressed with
212 colons if the lookup type is iplsearch. This is not incompatible, because
213 previously such lookups could never work.
215 The situation is now rather anomolous, since one *can* have colons in
216 ordinary lsearch keys. However, making the change in all cases is
217 incompatible and would probably break a number of configurations.
219 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
222 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
223 conversion specification without a maximum field width, thereby enabling
224 a rogue spamd server to cause a buffer overflow. While nobody in their
225 right mind would setup Exim to query an untrusted spamd server, an
226 attacker that gains access to a server running spamd could potentially
227 exploit this vulnerability to run arbitrary code as the Exim user.
229 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
230 $primary_hostname instead of what libspf2 thinks the hosts name is.
232 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
233 a directory entry by the name of the lookup key. Previously, if a
234 symlink pointed to a non-existing file or a file in a directory that
235 Exim lacked permissions to read, a lookup for a key matching that
236 symlink would fail. Now it is enough that a matching directory entry
237 exists, symlink or not. (Bugzilla 503.)
239 PH/03 The body_linecount and body_zerocount variables are now exported in the
242 PH/04 Added the $dnslist_matched variable.
244 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
245 This means they are set thereafter only if the connection becomes
248 PH/06 Added the client_condition to authenticators so that some can be skipped
249 by clients under certain conditions.
251 PH/07 The error message for a badly-placed control=no_multiline_responses left
252 "_responses" off the end of the name.
254 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
256 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
257 (without spaces) instead of just copying the configuration text.
259 PH/10 Added the /noupdate option to the ratelimit ACL condition.
261 PH/11 Added $max_received_linelength.
263 PH/12 Added +ignore_defer and +include_defer to host lists.
265 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
266 way in which PCRE > 7.0 is built.
268 PH/14 Implemented queue_only_load_latch.
270 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
271 MAIL command. The effect was to mangle the value on 64-bit systems.
273 PH/16 Another patch from the Sieve maintainer.
275 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
277 PH/18 If a system quota error occurred while trying to create the file for
278 a maildir delivery, the message "Mailbox is full" was not appended to the
279 bounce if the delivery eventually timed out. Change 4.67/27 below applied
280 only to a quota excession during the actual writing of the file.
282 PH/19 It seems that peer DN values may contain newlines (and other non-printing
283 characters?) which causes problems in log lines. The DN values are now
284 passed through string_printing() before being added to log lines.
286 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
287 and InterBase are left for another time.)
289 PH/21 Added message_body_newlines option.
291 PH/22 Guard against possible overflow in moan_check_errorcopy().
293 PH/23 POSIX allows open() to be a macro; guard against that.
295 PH/24 If the recipient of an error message contained an @ in the local part
296 (suitably quoted, of course), incorrect values were put in $domain and
297 $local_part during the evaluation of errors_copy.
303 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
304 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
307 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
308 issue a MAIL command.
310 PH/02 In an ACL statement such as
312 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
314 if a client was not listed at all, or was listed with a value other than
315 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
316 the condition was not true (as it should be), so access was not denied.
317 The bug was that the ! inversion was incorrectly passed on to the second
318 item. This has been fixed.
320 PH/03 Added additional dnslists conditions == and =& which are different from
321 = and & when the dns lookup returns more than one IP address.
323 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
324 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
326 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
327 FSYNC, which compiles an option called disable_fsync that allows for
328 bypassing fsync(). The documentation is heavily laced with warnings.
330 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
332 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
333 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
334 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
335 including adding "make clean"; (3) Added -fPIC when compiling the test
336 dynamically loaded module, to get rid of a warning.
338 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
339 message fails, move_frozen_messages = true and ignore_bounce_errors_after
340 = 0s. The bug is otherwise harmless.
342 PH/07 There was a bug in the dovecot authenticator such that the value of
343 $auth1 could be overwritten, and so not correctly preserved, after a
344 successful authentication. This usually meant that the value preserved by
345 the server_setid option was incorrect.
347 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
349 PH/09 Installed PCRE release 7.0.
351 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
352 run for batched SMTP input. It is now run at the start of every message
353 in the batch. While fixing this I discovered that the process information
354 (output by running exiwhat) was not always getting set for -bs and -bS
355 input. This is fixed, and it now also says "batched" for BSMTP.
357 PH/11 Added control=no_pipelining.
359 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
360 patch, slightly modified), and move the expansion of helo_data till after
361 the connection is made in the smtp transport (so it can use these
364 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
366 PH/14 Added log_selector = +pid.
368 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
370 PH/16 Add ${if forany and ${if forall.
372 PH/17 Added dsn_from option to vary the From: line in DSNs.
374 PH/18 Flush SMTP output before performing a callout, unless control =
375 no_callout_flush is set.
377 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
378 was true (the default) a successful delivery failed to delete the retry
379 item, thus causing premature timeout of the address. The bug is now
382 PH/20 Added hosts_avoid_pipelining to the smtp transport.
384 PH/21 Long custom messages for fakedefer and fakereject are now split up
385 into multiline reponses in the same way that messages for "deny" and
386 other ACL rejections are.
388 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
389 with slight modification.
391 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
392 draft, changing the syntax and factoring some duplicate code".
394 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
395 for deliveries of the second and subsequent messages over the same SMTP
398 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
399 ${reduce, with only minor "tidies".
401 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
403 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
404 expansion side effects.
406 PH/27 When a message times out after an over-quota error from an Exim-imposed
407 quota, the bounce message says "mailbox is full". This message was not
408 being given when it was a system quota that was exceeded. It now should
411 MH/03 Made $recipients available in local_scan(). local_scan() already has
412 better access to the recipient list through recipients_list[], but
413 $recipients can be useful in postmaster-provided expansion strings.
415 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
416 in the case of a MAIL command with additional options following the
417 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
418 were accidentally chopped off.
420 PH/29 SMTP synchronization checks are implemented when a command is read -
421 there is a check that no more input is waiting when there shouldn't be
422 any. However, for some commands, a delay in an ACL can mean that it is
423 some time before the response is written. In this time, more input might
424 arrive, invalidly. So now there are extra checks after an ACL has run for
425 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
426 pipelining has not been advertised.
428 PH/30 MH's patch to allow iscntrl() characters to be list separators.
430 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
431 returned in the SMTP response when smtp_return_error_details was false.
434 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
435 instead of the C I/O that was originally supplied, because problems were
438 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
439 Exim which did not show up earlier: it was assuming that a call to
440 SSL_CTX_set_info_callback() might give an error value. In fact, there is
441 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
442 was a macro that became an assignment, so it seemed to work. This has
443 changed to a proper function call with a void return, hence the compile
444 error. Exim's code has been fixed.
446 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
449 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
451 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
453 PH/37 If a message is not accepted after it has had an id assigned (e.g.
454 because it turns out to be too big or there is a timeout) there is no
455 "Completed" line in the log. When some messages of this type were
456 selected by exigrep, they were listed as "not completed". Others were
457 picked up by some special patterns. I have improved the selection
458 criteria to be more general.
460 PH/38 The host_find_failed option in the manualroute router can now be set
461 to "ignore", to completely ignore a host whose IP address cannot be
462 found. If all hosts are ignored, the behaviour is controlled by the new
463 host_all_ignored option.
465 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
466 homing or because of multiple MX records with /mx) generated more than
467 one IP address, and the following item turned out to be the local host,
468 all the secondary addresses of the first item were incorrectly removed
469 from the list, along with the local host and any following hosts (which
470 is what is supposed to happen).
472 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
473 whoever called Exim into the -H file. In the case of the daemon it was
474 behaving confusingly. When first started, it used values for whoever
475 started the daemon, but after a SIGHUP it used the Exim user (because it
476 calls itself on a restart). I have changed the code so that it now always
479 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
480 message are rejected with the same error (e.g. no authentication or bad
481 sender address), and a DATA command is nevertheless sent (as can happen
482 with PIPELINING or a stupid MUA), the error message that was given to the
483 RCPT commands is included in the rejection of the DATA command. This is
484 intended to be helpful for MUAs that show only the final error to their
487 PH/42 Another patch from the Sieve maintainer.
489 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
490 Eximstats - Fixed some broken HTML links and added missing column headers
492 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
493 columns for Rejects, Temp Rejects, Ham, and Spam rows.
495 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
497 PH/43 Yet another patch from the Sieve maintainer.
499 PH/44 I found a way to check for a TCP/IP connection going away before sending
500 the response to the final '.' that terminates a message, but only in the
501 case where the client has not sent further data following the '.'
502 (unfortunately, this is allowed). However, in many cases there won't be
503 any further data because there won't be any more messages to send. A call
504 to select() can be used: if it shows that the input is "ready", there is
505 either input waiting, or the socket has been closed. An attempt to read
506 the next input character can distinguish the two cases. Previously, Exim
507 would have sent an OK response which the client would never have see.
508 This could lead to message repetition. This fix should cure that, at
509 least in a lot of common cases.
511 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
512 advertised in response to EHLO.
518 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
519 fixed by 4.65/MH/01 (is this a record?) are fixed:
521 (i) An empty string was always treated as zero by the numeric comparison
522 operators. This behaviour has been restored.
524 (ii) It is documented that the numeric comparison operators always treat
525 their arguments as decimal numbers. This was broken in that numbers
526 starting with 0 were being interpreted as octal.
528 While fixing these problems I realized that there was another issue that
529 hadn't been noticed. Values of message_size_limit (both the global option
530 and the transport option) were treated as octal if they started with 0.
531 The documentation was vague. These values are now always treated as
532 decimal, and I will make that clear in the documentation.
538 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
539 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
542 MH/01 Don't check that the operands of numeric comparison operators are
543 integers when their expansion is in "skipping" mode (fixes bug
544 introduced by 4.64-PH/07).
546 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
547 child addresses, Exim now panics and dies. Previously, because the count
548 is held in a short int, deliveries were likely to be lost. As such a
549 large number of recipients for a single message is ridiculous
550 (performance will be very, very poor), I have chosen to impose a limit
551 rather than extend the field.
557 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
558 leftover -K file (the existence of which was triggered by #402).
559 While we were at it, introduced process PID as part of the -K
560 filename. This should rule out race conditions when creating
563 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
564 processing considerably. Previous code took too long for large mails,
565 triggering a timeout which in turn triggers #401.
567 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
568 in the DK code in transports.c. sendfile() is not really portable,
569 hence the _LINUX specificness.
571 TF/01 In the add_headers option to the mail command in an Exim filter,
572 there was a bug that Exim would claim a syntax error in any
573 header after the first one which had an odd number of characters
576 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
577 callout verification, Exim cached a "reject" for the entire domain. This
578 is correct for most verifications, but it is not correct for a recipient
579 verification with use_sender or use_postmaster set, because in that case
580 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
581 case of MAIL FROM:<> rejection from other early rejections (e.g.
582 rejection of HELO). When verifying a recipient using a non-null MAIL
583 address, the cache is ignored if it shows MAIL FROM:<> rejection.
584 Whatever the result of the callout, the value of the domain cache is
585 left unchanged (for any other kind of callout, getting as far as trying
586 RCPT means that the domain itself is ok).
588 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
591 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
592 manifest itself as EPIPE rather than ECONNECT. When tidying away a
593 session, the daemon ignores ECONNECT errors and logs others; it now
594 ignores EPIPE as well.
596 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
597 (quoted-printable decoding).
599 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
600 later the small subsequent patch to fix an introduced bug.
602 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
604 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
606 PH/08 An error is now given if message_size_limit is specified negative.
608 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
609 to be given (somewhat) arbitrary names.
611 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
614 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
615 miscellaneous code fixes
617 PH/10 Added the log_reject_target ACL modifier to specify where to log
620 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
621 hostname. This is wrong, because it relates to the incoming message (and
622 probably the interface on which it is arriving) and not to the outgoing
623 callout (which could be using a different interface). This has been
624 changed to use the value of the helo_data option from the smtp transport
625 instead - this is what is used when a message is actually being sent. If
626 there is no remote transport (possible with a router that sets up host
627 addresses), $smtp_active_hostname is used.
629 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
630 tweaks were necessary in order to get it to work (see also 21 below):
631 (a) The code assumed that strncpy() returns a negative number on buffer
632 overflow, which isn't the case. Replaced with Exim's string_format()
634 (b) There were several signed/unsigned issues. I just did the minimum
635 hacking in of casts. There is scope for a larger refactoring.
636 (c) The code used strcasecmp() which is not a standard C function.
637 Replaced with Exim's strcmpic() function.
638 (d) The code set only $1; it now sets $auth1 as well.
639 (e) A simple test gave the error "authentication client didn't specify
640 service in request". It would seem that Dovecot has changed its
641 interface. Fortunately there's a specification; I followed it and
642 changed what the client sends and it appears to be working now.
644 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
647 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
648 address A is aliased to B and C, where B exists and C does not. Without
649 -v the output is "A verified" because verification stops after a
650 successful redirection if more than one address is generated. However,
651 with -v the child addresses are also verified. Exim was outputting "A
652 failed to verify" and then showing the successful verification for C,
653 with its parentage. It now outputs "B failed to verify", showing B's
654 parentage before showing the successful verification of C.
656 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
657 look up a TXT record in a specific list after matching in a combined
660 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
661 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
662 they consult the DNS. I had assumed they would set it the way they
663 wanted; and indeed my experiments on Linux seem to show that in some
664 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
665 To be on the safe side, however, I have now made the interface to
666 host_find_byname() similar to host_find_bydns(), with an argument
667 containing the DNS resolver options. The host_find_byname() function now
668 sets these options at its start, just as host_find_bydns() does. The smtp
669 transport options dns_qualify_single and dns_search_parents are passed to
670 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
671 of host_find_byname() use the default settings of RES_DEFNAMES
672 (qualify_single) but not RES_DNSRCH (search_parents).
674 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
675 spool_read_header() do less string testing, by means of a preliminary
676 switch on the second character of optional "-foo" lines. (This is
677 overdue, caused by the large number of possibilities that now exist.
678 Originally there were few.) While I was there, I also converted the
679 str(n)cmp tests so they don't re-test the leading "-" and the first
680 character, in the hope this might squeeze out yet more improvement.
682 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
683 flag allowing group syntax was set by the header_syntax check but not
684 turned off, possible causing trouble later; (2) The flag was not being
685 set at all for the header_verify test, causing "group"-style headers to
686 be rejected. I have now set it in this case, and also caused header_
687 verify to ignore an empty address taken from a group. While doing this, I
688 came across some other cases where the code for allowing group syntax
689 while scanning a header line wasn't quite right (mostly, not resetting
690 the flag correctly in the right place). These bugs could have caused
691 trouble for malformed header lines. I hope it is now all correct.
693 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
694 with the "reply" argument non-NULL. The code, however (which originally
695 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
696 but it didn't always do it. This confused somebody who was copying the
697 code for some other use. I have removed all the tests.
699 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
700 feature that was used to support insecure browsers during the U.S. crypto
701 embargo. It requires special client support, and Exim is probably the
702 only MTA that supported it -- and would never use it because real RSA is
703 always available. This code has been removed, because it had the bad
704 effect of slowing Exim down by computing (never used) parameters for the
705 RSA_EXPORT functionality.
707 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
708 authenticator to fail if there's a tab character in the incoming data
709 (there should never be unless someone is messing about, as it's supposed
710 to be base64-encoded). Also added, on Timo's advice, the "secured" option
711 if the connection is using TLS or if the remote IP is the same as the
712 local IP, and the "valid-client-cert option" if a client certificate has
715 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
716 authenticators. This can be used for authorization after authentication
717 succeeds. (In the case of plaintext, it servers for both authentication
720 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
721 if any retry times were supplied.
723 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
724 connection, where there is no client IP address to check. In this
725 situation, the verify now always succeeds.
727 PH/25 Applied John Jetmore's -Mset patch.
729 PH/26 Added -bem to be like -Mset, but loading a message from a file.
731 PH/27 In a string expansion for a processed (not raw) header when multiple
732 headers of the same name were present, leading whitespace was being
733 removed from all of them, but trailing whitespace was being removed only
734 from the last one. Now trailing whitespace is removed from each header
735 before concatenation. Completely empty headers in a concatenation (as
738 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
739 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
741 PH/29 [Removed. This was a change that I later backed out, and forgot to
742 correct the ChangeLog entry (that I had efficiently created) before
743 committing the later change.]
745 PH/30 Exim was sometimes attempting to deliver messages that had suffered
746 address errors (4xx response to RCPT) over the same connection as other
747 messages routed to the same hosts. Such deliveries are always "forced",
748 so retry times are not inspected. This resulted in far too many retries
749 for the affected addresses. The effect occurred only when there were more
750 hosts than the hosts_max_try setting in the smtp transport when it had
751 the 4xx errors. Those hosts that it had tried were not added to the list
752 of hosts for which the message was waiting, so if all were tried, there
753 was no problem. Two fixes have been applied:
755 (i) If there are any address or message errors in an SMTP delivery, none
756 of the hosts (tried or untried) are now added to the list of hosts
757 for which the message is waiting, so the message should not be a
758 candidate for sending over the same connection that was used for a
759 successful delivery of some other message. This seems entirely
760 reasonable: after all the message is NOT "waiting for some host".
761 This is so "obvious" that I'm not sure why it wasn't done
762 previously. Hope I haven't missed anything, but it can't do any
763 harm, as the worst effect is to miss an optimization.
765 (ii) If, despite (i), such a delivery is accidentally attempted, the
766 routing retry time is respected, so at least it doesn't keep
767 hammering the server.
769 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
770 in ${readsocket because some servers need this prod.
772 PH/32 Added some extra debug output when updating a wait-xxx database.
774 PH/33 The hint "could be header name not terminated by colon", which has been
775 given for certain expansion errors for a long time, was not being given
776 for the ${if def:h_colon_omitted{... case.
778 PH/34 The spec says: "With one important exception, whenever a domain list is
779 being scanned, $domain contains the subject domain." There was at least
780 one case where this was not true.
782 PH/35 The error "getsockname() failed: connection reset by peer" was being
783 written to the panic log as well as the main log, but it isn't really
784 panic-worthy as it just means the connection died rather early on. I have
785 removed the panic log writing for the ECONNRESET error when getsockname()
788 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
789 runs only) independently of the message's sender address. This meant
790 that, if the 4xx error was in fact related to the sender, a different
791 message to the same recipient with a different sender could confuse
792 things. In particualar, this can happen when sending to a greylisting
793 server, but other circumstances could also provoke similar problems.
794 I have changed the default so that the retry time for these errors is now
795 based a combination of the sender and recipient addresses. This change
796 can be overridden by setting address_retry_include_sender=false in the
799 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
800 remote server are returned as part of bounce messages. This was not
801 happening for LMTP over a pipe (the lmtp transport), but now it is the
802 same for both kinds of LMTP.
804 PH/38 Despite being documented as not happening, Exim was rewriting addresses
805 in header lines that were in fact CNAMEs. This is no longer the case.
807 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
808 and queue runs started by the daemon processed all messages. This has
809 been fixed so that -R and -S can now usefully be given with -q<time>.
811 PH/40 Import PCRE release 6.7 (fixes some bugs).
813 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
815 PH/42 Give an error if -q is specified more than once.
817 PH/43 Renamed the variables $interface_address and $interface_port as
818 $received_ip_address and $received_port, to make it clear that these
819 values apply to message reception, and not to the outgoing interface when
820 a message is delivered. (The old names remain recognized, of course.)
822 PH/44 There was no timeout on the connect() call when using a Unix domain
823 socket in the ${readsocket expansion. There now is.
825 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
826 be meaningful with "accept".
828 SC/01 Eximstats V1.43
829 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
831 SC/02 Eximstats V1.44
832 Use a glob alias rather than an array ref in the generated
833 parser. This improves both readability and performance.
835 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
836 Collect SpamAssassin and rejection statistics.
837 Don't display local sender or destination tables unless
838 there is data to show.
839 Added average volumes into the top table text output.
841 SC/04 Eximstats V1.46
842 Collect data on the number of addresses (recipients)
843 as well as the number of messages.
845 SC/05 Eximstats V1.47
846 Added 'Message too big' to the list of mail rejection
847 reasons (thanks to Marco Gaiarin).
849 SC/06 Eximstats V1.48
850 Mainlog lines which have GMT offsets and are too short to
851 have a flag are now skipped.
853 SC/07 Eximstats V1.49 (Alain Williams)
854 Added the -emptyok flag.
856 SC/08 Eximstats V1.50
857 Fixes for obtaining the IP address from reject messages.
859 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
860 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
861 whitesspace changes from 4.64-PH/27
863 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
866 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
867 are found, allow negative numbers in numeric criteria)
869 JJ/06 exipick.20061117.2, added new $message_body_missing variable
871 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
872 to match changes made in 4.64-PH/43
874 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
876 PH/47 Put in an explicit test for a DNS lookup of an address record where the
877 "domain" is actually an IP address, and force a failure. This locks out
878 those revolvers/nameservers that support "A-for-A" lookups, in
879 contravention of the specifications.
881 PH/48 When a host name was looked up from an IP address, and the subsequent
882 forward lookup of the name timed out, the host name was left in
883 $sender_host_name, contrary to the specification.
885 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
886 restricted to single-key lookups, Exim was not diagnosing an error if
887 * or *@ was used with a query-style lookup.
889 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
891 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
892 long ago, but noone interested enough thought of it. Let's just say that
893 the "1.1" means that there are some new functions that weren't there at
894 some point in the past.
896 PH/51 Error processing for expansion failure of helo_data from an smtp
897 transport during callout processing was broken.
899 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
900 tested/used via the -bh/-bhc/-bs options.
902 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
903 bug, fixed in subsequent PCRE releases).
905 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
906 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
908 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
914 SC/01 Use a glob alias rather than an array ref in eximstats generated
915 parser. This improves both readability and performance.
917 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
918 Don't display local sender or destination tables in eximstats unless
919 there is data to show.
920 Added average volumes into the eximstats top table text output.
922 SC/03 Collect data on the number of addresses (recipients) as well
923 as the number of messages in eximstats.
925 TF/01 Correct an error in the documentation for the redirect router. Exim
926 does (usually) call initgroups() when daemonizing.
928 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
929 with consistent privilege compared to when running as a daemon.
931 TF/03 Note in the spec that $authenticated_id is not set for local
932 submissions from trusted users.
934 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
935 Thanks to Dean Brooks <dean@iglou.com> for the patch.
937 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
938 by adding some example configuration directives to the default
939 configuration file. A little bit of work is required to uncomment the
940 directives and define how usernames and passwords are checked, but
941 there is now a framework to start from.
943 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
944 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
945 without this. I don't know how relevant this is to other LDAP libraries.
947 PH/02 Add the verb name to the "unknown ACL verb" error.
949 PH/03 Magnus Holmgren's patch for filter_prepend_home.
951 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
953 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
954 directory not expanded when it should be if an expanded home directory
955 was set for the address (which is overridden by the transport).
957 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
960 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
961 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
962 because it is too late at that time, and has no effect.
964 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
965 security issue with \' (bugzilla #107). I could not use the
966 PQescapeStringConn() function, because it needs a PGconn value as one of
969 PH/08 When testing addresses using -bt, indicate those final addresses that
970 are duplicates that would not cause an additional delivery. At least one
971 person was confused, thinking that -bt output corresponded to deliveries.
972 (Suppressing duplicates isn't a good idea as you lose the information
973 about possibly different redirections that led to the duplicates.)
975 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
976 systems where poll() doesn't work, in particular OS X.
978 PH/10 Added more information to debugging output for retry time not reached.
980 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
981 operations in malware.c.
983 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
986 PH/13 If write_rejectlog was set false when logging was sent to syslog with
987 syslog_duplication set false, log lines that would normally be written
988 both the the main log and to the reject log were not written to syslog at
991 PH/14 In the default configuration, change the use of "message" in ACL warn
992 statements to "add_header".
994 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
995 not followed by a command (e.g. "seen endif").
997 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
998 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1001 PH/17 Added extra conditions to the default value of delay_warning_condition
1005 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1006 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1007 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1010 The Auto-Submitted: and various List- headers are standardised, whereas I
1011 don't think Precedence: ever was.
1013 PH/18 Refactored debugging code in route_finduser() to show more information,
1014 in particular, the error code if getpwnam() issues one.
1016 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1017 This is apparently needed in addition to the PH/07 change above to avoid
1018 any possible encoding problems.
1020 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1021 but not after initializing Perl.
1023 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1024 output them only if debugging. By default they are written stderr,
1025 apparently, which is not desirable.
1027 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1030 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1033 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1035 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1036 authenticated or an ident call has been made. Suppress the default
1037 values for $authenticated_id and $authenticated_sender (but permit -oMai
1038 and -oMas) when testing with -bh.
1040 PH/24 Re-jigged the order of the tests in the default configuration so that the
1041 tests for valid domains and recipients precede the DNS black list and CSA
1042 tests, on the grounds that those ones are more expensive.
1044 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1045 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1046 This bug exists in every version of Exim that I still have, right back to
1049 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1050 However, an attempt to turn on case-sensitivity in a regex key by
1051 including (?-i) didn't work because the subject string was already
1052 lowercased, and the effects were non-intuitive. It turns out that a
1053 one-line patch can be used to allow (?-i) to work as expected.
1059 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1060 other effects) broke the use of negated acl sub-conditions.
1062 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1065 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1066 "Deny" causes Exim to reject the incoming connection with a 554 error.
1067 Unfortunately, if there is a major crisis, such as a disk failure,
1068 tcp-wrappers gives "deny", whereas what one would like would be some
1069 kind of temporary error. A kludge has been added to help with this.
1070 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1071 554 error is used if errno is still zero or contains ENOENT (which occurs
1072 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1075 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1077 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1078 errors. Otherwise a message that provokes a temporary error (when other
1079 messages do not) can cause a whole host to time out.
1081 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1082 addresses were routed directly to files or pipes from a redirect router.
1083 File deliveries just didn't batch; pipe deliveries might have suffered
1086 PH/06 A failure to get a lock for a hints database would erroneously always say
1087 "Failed to get write lock", even when it was really a read lock.
1089 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1090 of 0600. This has been changed to use the value of the lockfile_mode
1091 option (which defaults to 0600).
1093 PH/08 Applied small patch from the Sieve maintainer.
1095 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1096 folder from quota calculations, a direct delivery into this folder messed
1097 up the contents of the maildirsize file. This was because the regex was
1098 used only to exclude .Trash (or whatever) when the size of the mailbox
1099 was calculated. There was no check that a delivery was happening into an
1100 excluded directory. This bug has been fixed by ignoring all quota
1101 processing for deliveries into excluded directories.
1103 PH/10 Added the maildirfolder_create_regex option to appendfile.
1109 PH/01 The code for finding all the local interface addresses on a FreeBSD
1110 system running IPv6 was broken. This may well have applied to all BSD
1111 systems, as well as to others that have similar system calls. The broken
1112 code found IPv4 interfaces correctly, but gave incorrect values for the
1113 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1114 that it would not match correctly against @[] and not recognize the IPv6
1117 PH/02 The ipliteral router was not recognizing addresses of the form user@
1118 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1120 PH/03 Added disable_ipv6.
1122 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1123 decoded form, because it is most often used to construct To: headers
1124 lines in autoreplies, and the decoded form may well be syntactically
1125 invalid. However, $reply_address has leading white space removed, and all
1126 newlines turned into spaces so that the autoreply transport does not
1129 PH/05 If group was specified without a user on a router, and no group or user
1130 was specified on a transport, the group from the router was ignored.
1132 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1133 for visible compile-time settings that can be used to change these
1134 numbers, for those that want even more. Backwards compatibility with old
1135 spool files has been maintained. However, going back to a previous Exim
1136 release will lost any variables that are in spool files.
1138 PH/07 Two small changes when running in the test harness: increase delay when
1139 passing a TCP/IP connection to a new process, in case the original
1140 process has to generate a bounce, and remove special handling of
1141 127.0.0.2 (sic), which is no longer necessary.
1143 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1144 be the same on different OS.
1146 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1149 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1150 whether --show-vars was specified or not
1152 JJ/02 exipick: Added support for new ACL variable spool format introduced
1155 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1156 syntactically invalid From: or Reply-to: line, and a filter used this to
1157 generate an autoreply, and therefore failed to obtain an address for the
1158 autoreply, Exim could try to deliver to a non-existent relative file
1159 name, causing unrelated and misleading errors. What now happens is that
1160 it logs this as a hard delivery error, but does not attempt to create a
1163 PH/11 The exinext utility has a -C option for testing purposes, but although
1164 the given file was scanned by exinext itself; it wasn't being passed on
1165 when Exim was called.
1167 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1168 an end-of-file indication when reading a command response.
1170 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1171 compiled. In many other places in Exim, IPv6 addresses are always
1172 recognized, so I have changed this. It also means that IPv4 domain
1173 literals of the form [IPV4:n.n.n.n] are now always recognized.
1175 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1176 used if the router is not running as root, for example, when verifying at
1177 ACL time, or when using -bh. The debugging output from this situation was
1178 non-existent - all you got was a failure to exec. I have made two
1181 (a) Failures to set uid/gid, the current directory, or a process leader
1182 in a subprocess such as that created by queryprogram now generate
1183 suitable debugging ouput when -d is set.
1185 (b) The queryprogram router detects when it is not running as root,
1186 outputs suitable debugging information if -d is set, and then runs
1187 the subprocess without attempting to change uid/gid.
1189 PH/15 Minor change to Makefile for building test_host (undocumented testing
1192 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1193 additional section of a DNS packet that returns MX or SRV records.
1194 Instead, it always explicitly searches for A/AAAA records. This avoids
1195 major problems that occur when a DNS server includes only records of one
1196 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1197 fixed another bug: if SRV records were looked up and the corresponding
1198 address records were *not* found in the additional section, the port
1199 values from the SRV records were lost.
1201 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1202 using the correct key (the original address) when searching the retry
1203 rules in order to find which one to use for generating the retry hint.
1205 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1206 adding the default one. Similarly, if it contains a Reply-To: header, the
1207 errors_reply_to option, if set, is not used.
1209 PH/19 When calculating a retry time, Exim used to measure the "time since
1210 failure" by looking at the "first failed" field in the retry record. Now
1211 it does not use this if it is later than than the arrival time of the
1212 message. Instead it uses the arrival time. This makes for better
1213 behaviour in cases where some deliveries succeed, thus re-setting the
1214 "first failed" field. An example is a quota failure for a huge message
1215 when small messages continue to be delivered. Without this change, the
1216 "time since failure" will always be short, possible causing more frequent
1217 delivery attempts for the huge message than are intended.
1218 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1220 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1221 $1, $2, $3) because the numerical variables can be reset during some
1222 expansion items (e.g. "match"), thereby losing the authentication data.
1224 PH/21 Make -bV show the size of off_t variables so that the test suite can
1225 decide whether to run tests for quotas > 2G.
1227 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1228 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1229 filecount value is greater than 2G or if a quota value is greater than 2G
1230 on a system where the size of off_t is not greater than 4, a panic error
1233 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1234 never match. The debug and -bh output now contains an explicit error
1235 message indicating a malformed IPv4 address or mask.
1237 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1238 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1239 PH/23 above applies.
1241 PH/25 Do not write to syslog when running in the test harness. The only
1242 occasion when this arises is a failure to open the main or panic logs
1243 (for which there is an explicit test).
1245 PH/26 Added the /no_tell option to "control=freeze".
1247 PH/27 If a host name lookup failed very early in a connection, for example, if
1248 the IP address matched host_lookup and the reverse lookup yielded a name
1249 that did not have a forward lookup, an error message of the form "no IP
1250 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1251 could be logged. Now it outputs the IP address instead of "NULL".
1253 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1254 allows the sender and the authenticated sender to be set when
1255 submitting a message from within Exim. Since child_open_exim() is
1256 documented for local_scan(), the new function should be too.
1258 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1259 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1260 results in an empty string is now treated as unset.
1262 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1264 PH/31 Added sender_verify_fail logging option.
1266 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1267 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1268 tidied the source and removed it altogether.
1270 PH/33 When a queue run was abandoned because the load average was too high, a
1271 log line was always written; now it is written only if the queue_run log
1272 selector is set. In addition, the log line for abandonment now contains
1273 information about the queue run such as the pid. This is always present
1274 in "start" and "stop" lines but was omitted from the "abandon" line.
1276 PH/34 Omit spaces between a header name and the colon in the error message that
1277 is given when verify = headers_syntax fails (if there are lots of them,
1278 the message gets confusing).
1280 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1281 names, as there are now some PTR records that contain slashes. This check
1282 is only to protect against broken name servers that fall over on strange
1283 characters, so the fact that it applies to all lookups doesn't matter.
1285 PH/36 Now that the new test suite is complete, we can remove some of the
1286 special code in Exim that was needed for the old test suite. For example,
1287 sorting DNS records because real resolvers return them in an arbitrary
1288 order. The new test suite's fake resolver always returns records in the
1291 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1292 bounces) except when queue_only is set, to avoid logging races between
1293 the different processes.
1295 PH/38 Panic-die if .include specifies a non-absolute path.
1297 PH/39 A tweak to the "H" retry rule from its user.
1299 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1300 a label. They prevented compilation on older perls.
1302 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1303 a warning to be raised on newish perls.
1305 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1306 on queue. Changes to match documented behaviour of showing count of
1307 messages matching specified criteria.
1309 PH/40 Changed the default ident timeout from 30s to 5s.
1311 PH/41 Added support for the use of login_cap features, on those BSD systems
1312 that have them, for controlling the resources used by pipe deliveries.
1314 PH/42 The content-scanning code uses fopen() to create files in which to put
1315 message data. Previously it was not paying any attention to the mode of
1316 the files. Exim runs with umask(0) because the rest of the code creates
1317 files with open(), and sets the required mode explicitly. Thus, these
1318 files were ending up world-writeable. This was not a big issue, because,
1319 being within the spool directory, they were not world-accessible. I have
1320 created a function called modefopen, which takes an additional mode
1321 argument. It sets umask(777), creates the file, chmods it to the required
1322 mode, then resets the umask. All the relevant calls to fopen() in the
1323 content scanning code have been changed to use this function.
1325 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1326 to 24 hours. This avoids potential overflow problems when processing G
1327 and H retry rules. I suspect nobody ever tinkers with this value.
1329 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1331 PH/45 When the plaintext authenticator is running as a client, the server's
1332 challenges are checked to ensure they are valid base64 strings. By
1333 default, the authentication attempt is cancelled if an invalid string is
1334 received. Setting client_ignore_invalid_base64 true ignores these errors.
1335 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1336 they are received. Thus, the responses can be made to depend on the
1337 challenges. If an invalid string is ignored, an empty string is placed in
1340 PH/46 Messages that are created by the autoreply transport now contains a
1341 References: header, in accordance with RFCs 2822 and 3834.
1343 PH/47 Added authenticated_sender_force to the smtp transport.
1345 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1347 PH/49 Installed latest patch from the Sieve maintainer.
1349 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1350 was also set, the appendfile transport was unnecessarily scanning a
1351 directory of message files (e.g. for maildir delivery) to find the count
1352 of files (along with the size), even though it did not need this
1353 information. It now does the scan only if it needs to find either the
1354 size of the count of files.
1356 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1358 PH/52 Two bugs concerned with error handling when the smtp transport is
1361 (i) Exim was not creating retry information for temporary errors given
1362 for individual recipients after the DATA command when the smtp transport
1363 was used in LMTP mode. This meant that they could be retried too
1364 frequently, and not timed out correctly.
1366 (ii) Exim was setting the flag that allows error details to be returned
1367 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1368 recipients that were returned after the DATA command.
1370 PH/53 This is related to PH/52, but is more general: for any failing address,
1371 when detailed error information was permitted to be returned to the
1372 sender, but the error was temporary, then after the final timeout, only
1373 "retry timeout exceeded" was returned. Now it returns the full error as
1374 well as "retry timeout exceeded".
1376 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1377 do this, and (what is worse) MTAs that accept it.
1379 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1380 will now be deprecated.
1382 PH/56 New os.c-cygwin from the Cygwin maintainer.
1384 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1385 formats (previously only available in exim formats via -bpr, -bpru,
1386 and -bpra. Now also available in native and exiqgrep formats)
1388 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1389 with very large, slow to parse queues
1391 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1393 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1395 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1396 responses to authentication challenges, though it was showing the
1397 challenges; (ii) I've removed the CR characters from the debug output for
1400 PH/58 Allow for the insertion of a newline as well as a space when a string
1401 is turned into more than one encoded-word during RFC 2047 encoding. The
1402 Sieve code now uses this.
1404 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1405 data_4xx, lost_connection, tls_required.
1407 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1408 message was being sent as an SMTP response.
1410 PH/61 Add -l and -k options to exicyclog.
1412 PH/62 When verifying, if an address was redirected to one new address, so that
1413 verification continued, and the new address failed or deferred after
1414 having set something in $address_data, the value of $address_data was not
1415 passed back to the ACL. This was different to the case when no
1416 redirection occurred. The value is now passed back in both cases.
1418 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1419 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1420 use login_cap.h, so on its own it isn't the distinguishing feature. The
1421 new name refers directly to the setclassresources() function.
1423 PH/65 Added configuration files for NetBSD3.
1425 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1427 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1428 is preferred over IPv4.
1430 PH/68 The bounce_return_message and bounce_return_body options were not being
1431 honoured for bounces generated during the reception of non-SMTP messages.
1432 In particular, this applied to messages rejected by the ACL. This bug has
1433 been fixed. However, if bounce_return_message is true and bounce_return_
1434 body is false, the headers that are returned for a non-SMTP message
1435 include only those that have been read before the error was detected.
1436 (In the case of an ACL rejection, they have all been read.)
1438 PH/69 The HTML version of the specification is now built in a directory called
1439 spec_html instead of spec.html, because the latter looks like a path with
1440 a MIME-type, and this confuses some software.
1442 PH/70 Catch two compiler warnings in sieve.c.
1444 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1445 function verify_get_ident() calls ip_connect() to connect a socket, but
1446 if the "connect()" function timed out, ip_connect() used to close the
1447 socket. However, verify_get_ident() also closes the socket later, and in
1448 between Exim writes to the log, which may get opened at this point. When
1449 the socket was closed in ip_connect(), the log could get the same file
1450 descriptor number as the socket. This naturally causes chaos. The fix is
1451 not to close the socket in ip_connect(); the socket should be closed by
1452 the function that creates it. There was only one place in the code where
1453 this was missing, in the iplookup router, which I don't think anybody now
1454 uses, but I've fixed it anyway.
1456 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1457 well as to direct DNS lookups. Otherwise the handling of names in host
1458 lists is inconsistent and therefore confusing.
1464 PH/01 Two changes to the default runtime configuration:
1466 (1) Move the checks for relay_from_hosts and authenticated clients from
1467 after to before the (commented out) DNS black list checks.
1469 (2) Add control=submission to the relay_from_hosts and authenticated
1470 clients checks, on the grounds that messages accepted by these
1471 statements are most likely to be submissions.
1473 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1475 (1) Generate an error if the third argument for the ${prvs expansion is
1478 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1481 (3) Reset the variables that are obtained from the first argument of
1482 ${prvscheck and used in the second argument before leaving the code,
1483 because their memory is reclaimed, so using them afterwards may do
1486 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1487 one (it's much easier than Tom thought :-).
1489 (5) Because of (4), we can now allow for the use of $prvscheck_result
1490 inside the third argument.
1492 PH/03 For some reason, the default setting of PATH when running a command from
1493 a pipe transport was just "/usr/bin". I have changed it to
1496 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1497 anything to be listed in the output from -bV.
1499 PH/05 When a filter generated an autoreply, the entire To: header line was
1500 quoted in the delivery log line, like this:
1502 => >A.N.Other <ano@some.domain> <original@ddress> ...
1504 This has been changed so that it extracts the operative address. There
1505 may be more than one such address. If so, they are comma-separated, like
1508 => >ano@some.domain,ona@other.domain <original@ddress> ...
1510 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1511 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1512 not being looked up in the rDNS to get a host name, Exim was showing the
1513 IP address twice in Received: lines, even though the IP addresses were
1514 identical. For example:
1516 Received: from [1.2.3.4] (helo=[1.2.3.4])
1518 However, if the real host name was known, it was omitting the HELO data
1519 if it matched the actual IP address. This has been tidied up so that it
1520 doesn't show the same IP address twice.
1522 PH/07 When both +timestamp and +memory debugging was on, the value given by
1523 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1524 called by the debug printing, thereby overwriting the timestamp buffer.
1525 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1527 PH/08 When the original message was included in an autoreply transport, it
1528 always said "this is a copy of the message, including all the headers",
1529 even if body_only or headers_only was set. It now gives an appropriate
1532 PH/09 Applied a patch from the Sieve maintainer which:
1534 o fixes some comments
1535 o adds the (disabled) notify extension core
1536 o adds some debug output for the result of if/elsif tests
1537 o points to the current vacation draft in the documentation
1538 and documents the missing references header update
1542 o fixes a bug in processing the envelope test (when testing
1543 multiple envelope elements, the last element determinted the
1546 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1547 Electronic Mail") by including:
1549 Auto-submitted: auto-generated
1551 in the messages that it generates (bounce messages and others, such as
1552 warnings). In the case of bounce messages for non-SMTP mesages, there was
1553 also a typo: it was using "Auto_submitted" (underscore instead of
1554 hyphen). Since every message generated by Exim is necessarily in response
1555 to another message, thes have all been changed to:
1557 Auto-Submitted: auto-replied
1559 in accordance with these statements in the RFC:
1561 The auto-replied keyword:
1563 - SHOULD be used on messages sent in direct response to another
1564 message by an automatic process,
1566 - MUST NOT be used on manually-generated messages,
1568 - MAY be used on Delivery Status Notifications (DSNs) and Message
1569 Disposition Notifications (MDNs),
1571 - MUST NOT be used on messages generated by automatic or periodic
1572 processes, except for messages which are automatic responses to
1575 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1576 to the default Received: header definition.
1578 PH/12 Added log selector acl_warn_skipped (default on).
1580 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1581 variables because (a) they are in the wrong storage pool and (b) even if
1582 they were copied, it wouldn't work properly because of the caching.
1584 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1585 checking when decoding. Apparently there are clients that generate
1586 overlong encoded strings. Why am I not surprised?
1588 PH/15 If the first argument of "${if match_address" was not empty, but did not
1589 contain an "@" character, Exim crashed. Now it writes a panic log message
1590 and treats the condition as false.
1592 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1594 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1595 extension "envelope-auth". The code is finished and in agreement with
1596 other implementations, but there is no documentation so far and in fact,
1597 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1598 not changing the active code.
1600 Print executed "if" and "elsif" statements when debugging is used. This
1601 helps a great deal to understand what a filter does.
1603 Document more things not specified clearly in RFC3028. I had all this
1604 sorted out, when out of a sudden new issues came to my mind. Oops."
1606 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1609 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1610 canonical form (as documented). However, after a host name lookup from
1611 the IP address, check_host() was doing a simple string comparison with
1612 addresses acquired from the DNS when checking that the found name did
1613 have the original IP as one of its addresses. Since any found IPv6
1614 addresses are likely to be in abbreviated form, the comparison could
1615 fail. Luckily, there already exists a function for doing the comparison
1616 by converting both addresses to binary, so now that is used instead of
1617 the text comparison.
1619 PH/20 There was another similar case to PH/19, when a complete host name was
1620 given in a host list; looking up its IP address could give an abbreviated
1621 form, whereas the current host's name might or might not be abbreviated.
1622 The same fix has been applied.
1628 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1629 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1632 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1633 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1635 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1637 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1638 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1639 Also, for "Auto-Submitted", treat anything other than "no" as
1640 non-personal, in accordance with RFC 3834. (Previously it treated
1641 anything starting "auto-" as non-personal.)
1643 TF/01 The control=submission/name=... option had a problem with syntax
1644 errors if the name included a slash character. The /name= option
1645 now slurps the rest of the string, so it can include any characters
1646 but it must come last in the list of options (after /sender_retain
1649 PH/05 Some modifications to the interface to the fake nameserver for the new
1657 TK/01 Added the "success_on_redirect" address verification option. See
1658 NewStuff for rationale and an example.
1660 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1662 PH/02 Patch to exigrep to allow it to work on syslog lines.
1664 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1665 fread() to skip over the body file's header line, because in Cygwin the
1666 header line is locked and is inaccessible.
1668 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1669 co-exist for some time) to make it clear that it is the Exim ID that is
1670 referenced, not the Message-ID: header line.
1672 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1673 string_format() function, because snprintf() does not exist on all
1676 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1677 ${stat: expansion item.
1679 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1680 protocol synchronization error", to keep the pedants happy.
1682 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1683 well as for IRIX systems, when gcc is being used. See the host.c source
1686 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1688 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1691 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1692 transport and to the smtp transport in LMTP mode.
1694 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1696 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1697 run a filter in a subprocess. This could lead to confusion in subsequent
1698 lookups in the parent process. There should also be a search_tidyup() at
1699 the end of the subprocess.
1701 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1702 only if the host matched helo_try_verify_hosts, which caused the
1703 verification to occur when the EHLO/HELO command was issued. The ACL just
1704 tested the remembered result. Now, if a previous verification attempt has
1705 not happened, "verify = helo" does it there and then.
1707 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1709 TK/03 Fix log output including CR from clamd.
1711 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1712 exist provoked a memory error which could cause a segfault.
1714 PH/15 Installed PCRE 6.2
1716 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1718 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1719 of the problem. Specifically, suggested +O2 rather than +O1 for the
1722 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1724 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1725 record, the port was not correctly logged when the outgoing_port log
1726 selector was set (it logged the transort's default port).
1728 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1729 fallback_hosts, and "hosts" in the smtp transport.
1731 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1732 host errors such as "Connection refused".
1734 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1735 authentication with radiusclient 0.4.9:
1737 - Error returned from rc_read_config was caught wrongly
1738 - Username/password not passed on to radius server due to wrong length.
1740 The presumption is that some radiusclient API changes for 4.51/PH/17
1741 were not taken care of correctly. The code is still untested by me (my
1742 Linux distribution still has 0.3.2 of radiusclient), but it was
1743 contributed by a Radius user.
1745 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1746 expanding the "port" option of the smtp transport.
1748 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1749 while reading a MIME header. Thanks to Tom Hughes for a patch.
1751 PH/24 Include config.h inside local_scan.h so that configuration settings are
1754 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1755 that in an ACL for RCPT (for example), you can examine exactly what was
1758 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1759 commands, but it was not correctly comparing the address with the actual
1760 client host address. Thus, it would show the EHLO address in Received:
1761 header lines when this was not necessary.
1763 PH/27 Added the % operator to ${eval:}.
1765 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1766 it should be ignoring failures (because with -C, for example, it has lost
1767 privilege). It wasn't ignoring creation failures other than "already
1770 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1773 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1774 because an input line was too long, either on its own, or by virtue of
1775 too many continuations, the temporary file was not being removed, and the
1776 return code was incorrect.
1778 PH/31 Missing "BOOL" in function definition in filtertest.c.
1780 PH/32 Applied Sieve patches from the maintainer.
1782 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1784 PH/33 Added "verify = not_blind".
1786 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1787 Local/Makefile (with some defaults set). These are used in built scripts
1788 such as exicyclog, but they have never been used in the exim_install
1789 script (though there are many overriding facilities there). I have
1790 arranged that the exim_install script now takes note of these two
1793 PH/35 Installed configuration files for Dragonfly.
1795 PH/36 When a locally submitted message by a trusted user did not contain a
1796 From: header, and the sender address was obtained from -f or from an SMTP
1797 MAIL command, and the trusted user did not use -F to supply a sender
1798 name, $originator_name was incorrectly used when constructing a From:
1799 header. Furthermore, $originator_name was used for submission mode
1800 messages from external hosts without From: headers in a similar way,
1801 which is clearly wrong.
1803 PH/37 Added control=suppress_local_fixups.
1805 PH/38 When log_selector = +received_sender was set, and the addition of the
1806 sender made the log line's construction buffer exactly full, or one byte
1807 less than full, an overflow happened when the terminating "\n" was
1810 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1811 when the result of a list match is failure because a DNS lookup failed.
1813 PH/40 RM_COMMAND is now used in the building process.
1815 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1816 the "build-* directories that it finds.
1818 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1819 address in a domain literal was a prefix of an interface address.
1821 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1822 when verifying a sender address, unless rewrite_headers is false.
1824 PH/44 Wrote a long comment about why errors_to addresses are verified as
1825 recipients, not senders.
1827 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1828 the ratelimit ACL was added.
1830 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1832 PH/46 Added extra information about PostgreSQL errors to the error string.
1834 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1835 avoiding the need to install special zones in a real server. This is
1836 backwards compatible; if it can't find the fake resolver, it drops back.
1837 Thus, both old and new test suites can be run.
1839 TF/02 Added util/ratelimit.pl
1841 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1842 clock is set back in time.
1844 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1845 Candler <B.Candler@pobox.com>.
1847 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1848 OK for addresses that are the result of redirections.
1850 PH/48 A number of further additions for the benefit of the new test suite,
1851 including a fake gethostbyname() that interfaces to the fake DNS resolver
1854 TF/06 The fix for widen_domains has also been applied to qualify_single and
1855 search_parents which are the other dnslookup options that can cause
1858 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1861 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1863 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1864 service" if it encountered an MX record with an empty target hostname.
1865 The message is now "an MX or SRV record indicated no SMTP service".
1867 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1868 if the DNS of the sending site is misconfigured. This is quite a
1869 common situation. This change restores the behaviour of treating a
1870 helo verification defer as a failure.
1872 PH/51 If self=fail was set on a router, the bounce message did not include the
1873 actual error message.
1879 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1881 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1882 command itself ended in error, the underlying message about the transport
1883 filter timeout was being overwritten with the pipe command error. Now the
1884 underlying error message should be appended to the second error message.
1886 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1888 PH/02 Reduce the amount of output that "make" produces by default. Full output
1889 can still be requested.
1891 PH/03 The warning log line about a condition test deferring for a "warn" verb
1892 was being output only once per connection, rather than after each
1893 occurrence (because it was using the same function as for successful
1894 "warn" verbs). This seems wrong, so I have changed it.
1896 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1897 it should not have, which might have caused a crash in the right
1898 circumstances, but probably never did.
1900 PH/04 Installed a modified version of Tony Finch's patch to make submission
1901 mode fix the return path as well as the Sender: header line, and to
1902 add a /name= option so that you can make the user's friendly name appear
1905 TF/03 Added the control = fakedefer ACL modifier.
1907 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1908 Mark Lowes for thorough testing.
1910 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1912 TK/03 Merged latest SRS patch from Miles Wilton.
1914 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1915 with the definition in sysexits.h (which is #included earlier).
1916 Fortunately, Exim does not actually use EX_OK. The code used to try to
1917 preserve the sysexits.h value, by assumimg that macro definitions were
1918 scanned for macro replacements. I have been disabused of this notion,
1919 so now the code just undefines EX_OK before #including unistd.h.
1921 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1922 in the smtp transport. When a block could not be written in a single
1923 write() function, the timeout was being re-applied to each part-write.
1924 This seems wrong - if the receiver was accepting one byte at a time it
1925 would take for ever. The timeout is now adjusted when this happens. It
1926 doesn't have to be particularly precise.
1928 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1929 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1931 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1932 without a domain if the check to <postmaster@domain> fails.
1934 SC/01 Eximstats: added -xls and the ability to specify output files
1935 (patch written by Frank Heydlauf).
1937 SC/02 Eximstats: use FileHandles for outputing results.
1939 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1941 SC/04 Eximstats: fixed display of large numbers with -nvr option
1943 SC/05 Eximstats: fixed merging of reports with empty tables.
1945 SC/06 Eximstats: added the -include_original_destination flag
1947 SC/07 Eximstats: removed tabs and trailing whitespace.
1949 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1951 TK/06 MBOX spool code: Add real "From " MBOX separator line
1952 so the .eml file is really in mbox format (even though
1953 most programs do not really care). Patch from Alex Miller.
1955 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1956 The latter is generated from $received_to and is only set if the
1957 message has one envelope recipient. SA can use these headers,
1958 obviously out-of-the-box. Patch from Alex Miller.
1960 PH/08 The ${def test on a variable was returning false if the variable's
1961 value was "0", contrary to what the specification has always said!
1962 The result should be true unless the variable is empty.
1964 PH/09 The syntax error of a character other than { following "${if
1965 def:variable_name" (after optional whitespace) was not being diagnosed.
1966 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1967 accidental colon was present, for example, could give incorrect results.
1969 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1970 result is used (not including appendfile, where other changes are about
1973 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1974 This involved changing a lot of size variables from int to off_t. It
1975 should work with maildirs and everything.
1977 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1978 spamd dying while we are connected to it.
1980 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1983 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1984 function for BDB 4.3.
1986 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1988 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1989 the sources for PCRE have been re-arranged and more files are now
1992 PH/15 The code I had for printing potentially long long variables in PH/11
1993 above was not the best (it lost precision). The length of off_t variables
1994 is now inspected at build time, and an appropriate printing format (%ld
1995 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1996 to be "long long int" or "long int". This is needed for the internal
1997 formatting function string_vformat().
1999 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2000 the configuration file to be ":syslog", then the script "guesses" where
2001 the logs files are, rather than using the compiled in default. In our
2002 case the guess is not the same as the compiled default, so the script
2003 suddenly stopped working when I started to use syslog. The patch checks
2004 to see if log_file_path is "". If so, it attempts to read it from exim
2005 with no configuration file to get the compiled in version, before it
2006 falls back to the previous guessing code."
2008 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2009 implementing BATV in an Exim configuration. See NewStuff for the gory
2012 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2013 Makefile that are specific to HP-UX.
2015 PH/18 If the "use_postmaster" option was set for a recipient callout together
2016 with the "random" option, the postmaster address was used as the MAIL
2017 FROM address for the random test, but not for the subsequent recipient
2018 test. It is now used for both.
2020 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2021 patch removes a few documentation additions to RFC 3028, because the
2022 latest draft now contains them. It adds the new en;ascii-case comparator
2023 and a new error check for 8bit text in MIME parts. Comparator and
2024 require names are now matched exactly. I enabled the subaddress
2025 extension, but it is not well tested yet (read: it works for me)."
2027 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2028 rework some of the code of TK/09 above to avoid the hardwired use of
2029 "%lld" and "long long". Replaced the call to snprintf() with a call to
2032 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2033 records point to non-existent hosts", "retry timeout exceeded", and
2034 "retry time not reached for any host after a long failure period".
2036 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2037 experimental DomainKeys support:
2039 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2040 (2) On an error such as an illegally used "control", the wrong name for
2041 the control was given.
2043 These problems did NOT occur unless DomainKeys support was compiled.
2045 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2047 PH/24 Added ${if match_ip condition.
2049 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2050 databases so that it will be absolutely obvious if a crash occurs in the
2051 DB library. This is a regular occurrence (often caused by mis-matched
2054 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2055 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2056 picked up on a user's system that detects such things. There doesn't seem
2057 to be a gcc warning option for this - only an attribute that has to be
2058 put on the function's prototype. It seems that in Fedora Core 4 they have
2059 set this on a number of new functions. No doubt there will be more in due
2062 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2063 specify a transport. However, if an address that was verified by such a
2064 router was the subject of a callout, Exim crashed because it tried to
2065 read the rcpt_include_affixes from the non-existent transport. Now it
2066 just assumes that the setting of that option is false. This bug was
2067 introduced by 4.51/PH/31.
2069 PH/28 Changed -d+all to exclude +memory, because that information is very
2070 rarely of interest, but it makes the output a lot bigger. People tend to
2071 do -d+all out of habit.
2073 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2074 code in os-type was giving problems when libc.so lives in lib64, like on
2077 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2078 aren't the modern standard, and it seems that some systems' include files
2079 don't always have them. Exim was already checking for some of the newer
2080 ones like T_AAAA, and defining it itself. I've added checks for all the
2081 record types that Exim uses.
2083 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2084 not automatically generating a new one, as it is supposed to. This
2085 prevented TLS from working. If the file did exist, but contained invalid
2086 data, a new version was generated, as expected. It was only the case of a
2087 non-existent file that was broken.
2089 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2090 with a change in libdomainkeys > 0.64.
2092 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2093 from DNS. If the selector record carries the flag, it now has
2094 precedence over the domain-wide flag.
2096 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2098 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2099 the use of a port name that isn't defined in /etc/services) occurred, the
2100 message was deferred as in a normal delivery, and thus remained on the
2101 spool, instead of being failed because of the mua_wrapper setting. This
2102 is now fixed, and I tidied up some of the mua_wrapper messages at the
2105 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2106 the messages in a hash of arrays rather than using individual hashes.
2107 This is a bit cleaner and results in dramatic memory savings, albeit
2108 at a slight CPU cost.
2110 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2111 as requested by Marc Sherman.
2113 SC/10 Eximstats: added histograms for user specified patterns as requested
2116 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2118 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2119 fopen() in the content-scanning modules that did not already have it.
2125 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2126 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2128 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2130 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2132 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2133 to test Sieve filters that use "vacation".
2135 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2136 that changes the way the GnuTLS parameters are stored in the cache file.
2137 The new format can be generated externally. For backward compatibility,
2138 if the data in the cache doesn't make sense, Exim assumes it has read an
2139 old-format file, and it generates new data and writes a new file. This
2140 means that you can't go back to an older release without removing the
2143 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2144 work if there are any delivery delays because "one_time" forces the
2145 parent to be marked "delivered", so its unseen clone is never tried
2146 again. For this reason, Exim now forbids the simultaneous setting of
2149 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2150 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2151 entry if you want to know the details. The fix, however, neglected to
2152 consider the case when local delivery batching is involved. The test for
2153 "previously delivered" was not happening when checking to see if an
2154 address could be batched with a previous (undelivered) one; under
2155 certain circumstances this could lead to multiple deliveries to the same
2158 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2159 in its include files, and this causes problems building Exim.
2161 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2162 header_syntax) but Exim was just ignoring anything given after a slash.
2163 In particular, this caused confusion with an attempt to use "verify =
2164 reverse_host_lookup/defer_ok". An error is now given when options are
2165 supplied for verify items that do not have them. (Maybe reverse_host_
2166 lookup should have a defer_ok option, but that's a different point.)
2168 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2169 defined by RFC 821) to 2048, because there were problems with some AUTH
2170 commands, and RFC 1869 says the size should be increased for extended
2171 SMTP commands that take arguments.
2173 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2176 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2177 "unknown" error; now it says that the functionality isn't in the binary.
2179 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2180 an address' error message when a string expansion fails (syntax or
2181 whatever). Otherwise the password may appear in the log. Following change
2182 PH/42 below, there is no longer a chance of it appearing in a bounce
2185 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2187 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2188 of its addresses was ever tried. (Bugzilla bug #2.)
2190 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2191 the result incorrectly in the debug output. (It correctly added a newline
2192 to what was transported.)
2194 TF/01 Added $received_time.
2196 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2197 commented out examples of how to interface to a virus scanner and to
2198 SpamAssassin. Also added commented examples of av_scanner and
2199 spamd_address settings.
2201 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2202 and controls are allowed in which ACLs. There were a couple of minor
2203 errors. Some of the entries in the conditions table (which is a table of
2204 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2205 negation of where the condition IS allowed.
2207 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2209 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2210 header file does not have a version number, so I've had to invent a new
2211 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2212 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2213 radiusclient), but it was contributed by a Radius user.
2215 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2216 files or directories, for OpenSSL.
2218 PH/19 When an Exim process that is running as root has to create an Exim log
2219 file, it does so in a subprocess that runs as exim:exim so as to get the
2220 ownership right at creation (otherwise, other Exim processes might see
2221 the file with the wrong ownership). There was no test for failure of this
2222 fork() call, which would lead to the process getting stuck as it waited
2223 for a non-existent subprocess. Forks do occasionally fail when resources
2224 run out. I reviewed all the other calls to fork(); they all seem to check
2227 PH/20 When checking for unexpected SMTP input at connect time (before writing
2228 the banner), Exim was not dealing correctly with a non-positive return
2229 from the read() function. If the client had disconnected by this time,
2230 the result was a log entry for a synchronization error with an empty
2231 string after "input=" when read() returned zero. If read() returned -1
2232 (an event I could not check), uninitialized data bytes were printed.
2233 There were reports of junk text (parts of files, etc) appearing after
2236 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2238 PH/22 Added support for macro redefinition, and (re)definition in between
2239 driver and ACL definitions.
2241 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2242 forgetting to use the resulting value; it was using the unexpanded value.
2244 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2245 hadn't been configured. The fix is from Juergen Kreileder, who
2246 understands it better than I do:
2248 "Here's what I see happening with three configured cyrus_sasl
2249 authenticators configured (plain, login, cram-md5):
2251 On startup auth_cyrus_sasl_init() gets called for each of these.
2252 This means three calls to sasl_listmech() without a specified mech_list.
2253 => SASL tests which mechs of all available mechs actually work
2254 => three warnings about OTP not working
2255 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2257 With the patch, sasl_listmech() also gets called three times. But now
2258 SASL's mech_list option is set to the server_mech specified in the the
2259 authenticator. Or in other words, the answer from sasl_listmech()
2260 gets limited to just the mech you're testing for (which is different
2262 => the return list contains just 'plain' or 'login', 'cram-md5' or
2263 nothing depending on the value of ob->server_mech.
2265 I've just tested the patch: Authentication still works fine,
2266 unavailable mechs specified in the exim configuration are still
2267 caught, and the auth.log warnings about OTP are gone."
2269 PH/25 When debugging is enabled, the contents of the command line are added
2270 to the debugging output, even when log_selector=+arguments is not
2273 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2274 answer is "GNU", and only if the return is "GNU/something" is the answer
2277 PH/27 $acl_verify_message is now set immediately after the failure of a
2278 verification in an ACL, and so is available in subsequent modifiers. In
2279 particular, the message can be preserved by coding like this:
2281 warn !verify = sender
2282 set acl_m0 = $acl_verify_message
2284 Previously, $acl_verify_message was set only while expanding "message"
2285 and "log_message" when a very denied access.
2287 PH/28 Modified OS/os.c-Linux with
2289 -#ifndef OS_LOAD_AVERAGE
2290 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2292 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2293 nomenclature these days.)
2295 PH/29 Installed patch from the Sieve maintainer that adds the options
2296 sieve_useraddress and sieve_subaddress to the redirect router.
2298 PH/30 In these circumstances:
2299 . Two addresses routed to the same list of hosts;
2300 . First host does not offer TLS;
2301 . First host accepts first address;
2302 . First host gives temporary error to second address;
2303 . Second host offers TLS and a TLS session is established;
2304 . Second host accepts second address.
2305 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2306 and peerdn, if requested) that were in fact used only for the second
2309 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2310 attention to any local part prefix or suffix that was matched by the
2311 router that accepted the address. It now behaves in the same way as it
2312 does for delivery: the affixes are removed from the local part unless
2313 rcpt_include_affixes is set on the transport.
2315 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2316 timeout during the DATA phase of an incoming message.
2318 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2319 applied a patch sanctioned by the Sieve maintainer.
2321 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2322 the uid or gid is negative. A case of a negative gid caused this to be
2323 noticed. The fix allows for either to be negative.
2325 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2326 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2329 PH/36 The change PH/12 above was broken. Fixed it.
2331 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2332 the grounds that routing the same address twice would always produce the
2333 same answer. This might have been true once, but it is certainly no
2334 longer true now. Routing a child address may depend on the previous
2335 routing that produced that child. Some complicated redirection strategies
2336 went wrong when messages had multiple recipients, and made Exim's
2337 behaviour dependent on the order in which the addresses were given.
2339 I have moved the duplicate checking until after the routing is complete.
2340 Exim scans the addresses that are assigned to local and remote
2341 transports, and removes any duplicates. This means that more work will be
2342 done, as duplicates will always all be routed, but duplicates are
2343 presumably rare, so I don't expect this is of any significance.
2345 For deliveries to pipes, files, and autoreplies, the duplicate checking
2346 still happens during the routing process, since they are not going to be
2349 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2350 It corrects a timeout issue with spamd. This is Ian's comment: "The
2351 background is that sometimes spamd either never reads data from a
2352 connection it has accepted, or it never writes response data. The exiscan
2353 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2354 blindly assumes that writes won't block so it may never time out."
2356 PH/39 Allow G after quota size as well as K and M.
2358 PH/40 The value set for $authenticated_id in an authenticator may not contain
2359 binary zeroes or newlines because the value is written to log lines and
2360 to spool files. There was no check on this. Now the value is run through
2361 the string_printing() function so that such characters are converted to
2362 printable escape sequences.
2364 PH/41 $message_linecount is a new variable that contains the total number of
2365 lines in the message. Compare $body_linecount, which is the count for the
2368 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2369 bounce and delay warning messages, except in certain special cases, which
2372 (a) An SMTP error message from a remote host;
2373 (b) A message specified in a :fail: redirection;
2374 (c) A message specified in a "fail" command in a system filter;
2375 (d) A message specified in a FAIL return from the queryprogram router;
2376 (e) A message specified by the cannot_route_message router option.
2378 In these cases only, Exim does include the error details in bounce and
2379 warning messages. There are also a few cases where bland messages such
2380 as "unrouteable address" or "local delivery error" are given.
2382 PH/43 $value is now also set for the "else" part of a ${run expansion.
2384 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2385 being worked on, but at least Exim now implements the latest version to
2388 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2389 process to complete was treated as a delivery failure, a timeout while
2390 writing the message to the pipe was logged, but erroneously treated as a
2391 successful delivery. Such timeouts include transport filter timeouts. For
2392 consistency with the overall process timeout, these timeouts are now
2393 treated as errors, giving rise to delivery failures by default. However,
2394 there is now a new Boolean option for the pipe transport called
2395 timeout_defer, which, if set TRUE, converts the failures into defers for
2396 both kinds of timeout. A transport filter timeout is now identified in
2399 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2400 systems where "make" and "gmake" are different, calling "gmake" at top
2401 level broke things. I've arranged for the value of $(MAKE) to be passed
2402 from the Makefile to this script so that it can call the same version of
2406 A note about Exim versions 4.44 and 4.50
2407 ----------------------------------------
2409 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2410 changes of various kinds. As a consequence, a big documentation update was
2411 needed. This delayed the release for rather longer than seemed good, especially
2412 in the light of a couple of (minor) security issues. Therefore, the changes
2413 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2414 release. So 4.44 and 4.50 are in effect two different branches that both start
2417 I have left the 4.50 change log unchanged; it contains all the changes since
2418 4.43. The change log for 4.44 is below; many of its items are identical to
2419 those for 4.50. This seems to be the most sensible way to preserve the
2420 historical information.
2426 1. Minor wording change to the doc/README.SIEVE file.
2428 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2429 computation of the current number of files was incorrect.
2431 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2432 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2435 4. Give more explanation in the error message when the command for a transport
2436 filter fails to execute.
2438 5. There are several places where Exim runs a non-Exim command in a
2439 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2440 was being done only for the command run by the queryprogram router. It is
2441 now done for all such subprocesses. The other cases are: ${run, transport
2442 filters, and the commands run by the lmtp and pipe transports.
2444 6. Added CONFIGURE_GROUP build-time option.
2446 7. Some older OS have a limit of 256 on the maximum number of file
2447 descriptors. Exim was using setrlimit() to set 1000 as a large value
2448 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2449 systems. I've change it so that if it can't get 1000, it tries for 256.
2451 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2452 was an oversight, and furthermore, ever since the addition of extra
2453 controls (e.g. 4.43/32), the checks on when to allow different forms of
2454 "control" were broken. There should now be diagnostics for all cases when a
2455 control that does not make sense is encountered.
2457 9. Added the /retain_sender option to "control=submission".
2459 10. $recipients is now available in the predata ACL (oversight).
2461 11. Tidy the search cache before the fork to do a delivery from a message
2462 received from the command line. Otherwise the child will trigger a lookup
2463 failure and thereby defer the delivery if it tries to use (for example) a
2464 cached ldap connection that the parent has called unbind on.
2466 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2467 of $address_data from the recipient verification was clobbered by the
2468 sender verification.
2470 13. The value of address_data from a sender verification is now available in
2471 $sender_address_data in subsequent conditions in the ACL statement.
2473 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2475 15. Added a new option "connect=<time>" to callout options, to set a different
2478 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2479 was its contents. (It was OK if the option was not defined at all.)
2481 17. A "Completed" log line is now written for messages that are removed from
2482 the spool by the -Mrm option.
2484 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2485 information about exactly what failed.
2487 19. Added -dd to debug only the daemon process.
2489 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2490 handles timeouts, both on the server side and network timeouts. Renamed the
2491 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2493 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2494 It is now set to "smtps".
2496 22. $host_address is now set to the target address during the checking of
2497 ignore_target_hosts.
2499 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2500 being passed; this would have caused $sender_host_name to have been used if
2501 matching the list had actually called for a host name (not very likely,
2502 since this list is usually IP addresses). A host name is now passed as
2505 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2506 code that specifies a non-restarting handler (typically sigaction() in
2507 modern systems) in an attempt to fix a rare and obscure crash bug.
2509 25. Narrowed the window for a race in the daemon that could cause it to ignore
2510 SIGCHLD signals. This is not a major problem, because they are used only to
2511 wake it up if nothing else does.
2513 26. A malformed maildirsize file could cause Exim to calculate negative values
2514 for the mailbox size or file count. Odd effects could occur as a result.
2515 The maildirsize information is now recalculated if the size or filecount
2518 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2519 support for a long time. Removed HAVE_SYS_VFS_H.
2521 28. Installed the latest version of exipick from John Jetmore.
2523 29. In an address list, if the pattern was not a regular expression, an empty
2524 subject address (from a bounce message) matched only if the pattern was an
2525 empty string. Non-empty patterns were not even tested. This was the wrong
2526 because it is perfectly reasonable to use an empty address as part of a
2527 database query. An empty address is now tested by patterns that are
2528 lookups. However, all the other forms of pattern expect the subject to
2529 contain a local part and a domain, and therefore, for them, an empty
2530 address still always fails if the pattern is not itself empty.
2532 30. Exim went into a mad DNS loop when attempting to do a callout where the
2533 host was specified on an smtp transport, and looking it up yielded more
2534 than one IP address.
2536 31. Re-factored the code for checking spool and log partition space into a
2537 function that finds that data and another that does the check. The former
2538 is then used to implement four new variables: $spool_space, $log_space,
2539 $spool_inodes, and $log_inodes.
2541 32. The RFC2047 encoding function was originally intended for short strings
2542 such as real names; it was not keeping to the 75-character limit for
2543 encoded words that the RFC imposes. It now respects the limit, and
2544 generates multiple encoded words if necessary. To be on the safe side, I
2545 have increased the buffer size for the ${rfc2047: expansion operator from
2548 33. It is now permitted to omit both strings after an "if" condition; if the
2549 condition is true, the result is "true". As before, when the second string
2550 is omitted, a false condition yields an empty string. This makes it less
2551 cumbersome to write custom ACL and router conditions.
2553 34. Failure to deliver a bounce message always caused it to be frozen, even if
2554 there was an errors_to setting on the router. The errors_to setting is now
2557 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2558 canonical form (fully expanded) before being placed in
2559 $sender_host_address.
2561 36. The table in the code that translates DNS record types into text (T_A to
2562 "A" for instance) was missing entries for NS and CNAME. It is just possible
2563 that this could have caused confusion if both these types were looked up
2564 for the same domain, because the text type is used as part of Exim's
2565 per-process caching. But the chance of anyone hitting this buglet seems
2568 37. The dnsdb lookup has been extended in a number of ways.
2570 (1) There is a new type, "zns", which walks up the domain tree until it
2571 finds some nameserver records. It should be used with care.
2573 (2) There is a new type, "mxh", which is like "mx" except that it returns
2574 just the host names, not the priorities.
2576 (3) It is now possible to give a list of domains (or IP addresses) to be
2577 looked up. The behaviour when one of the lookups defers can be
2578 controlled by a keyword.
2580 (4) It is now possible to specify the separator character for use when
2581 multiple records are returned.
2583 38. The dnslists ACL condition has been extended: it is now possible to supply
2584 a list of IP addresses and/or domains to be looked up in a particular DNS
2587 39. Added log_selector=+queue_time_overall.
2589 40. When running the queue in the test harness, wait just a tad after forking a
2590 delivery process, to get repeatability of debugging output.
2592 41. Include certificate and key file names in error message when GnuTLS fails
2593 to set them up, because the GnuTLS error message doesn't include the name
2594 of the failing file when there is a problem reading it.
2596 42. Allow both -bf and -bF in the same test run.
2598 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2600 44. The "Exiscan patch" is now merged into the mainline Exim source.
2602 45. Sometimes the final signoff response after QUIT could fail to get
2603 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2604 < 0 before doing a fflush(). This bug looks as though it goes back to the
2605 introduction of TLS in release 3.20, but "sometimes" must have been rare
2606 because the tests only now provoked it.
2608 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2609 (this can affect the format of dates).
2611 47. exim_tidydb, when checking for the continued existence of a message for
2612 which it has found a message-specific retry record, was not finding
2613 messages that were in split spool directories. Consequently, it was
2614 deleting retry records that should have stayed in existence.
2616 48. Steve fixed some bugs in eximstats.
2618 49. The SPA authentication driver was not abandoning authentication and moving
2619 on to the next authenticator when an expansion was forced to fail,
2620 contradicting the general specification for all authenticators. Instead it
2621 was generating a temporary error. It now behaves as specified.
2623 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2624 (the order specifies the preference for clients). The order is now AES256,
2625 AES128, 3DES, ARCFOUR128.
2627 51. Small patch to Sieve code - explicitly set From: when generating an
2630 52. Exim crashed if a remote delivery caused a very long error message to be
2631 recorded - for instance if somebody sent an entire SpamAssassin report back
2632 as a large number of 550 error lines. This bug was coincidentally fixed by
2633 increasing the size of one of Exim's internal buffers (big_buffer) that
2634 happened as part of the Exiscan merge. However, to be on the safe side, I
2635 have made the code more robust (and fixed the comments that describe what
2638 53. Now that there can be additional text after "Completed" in log lines (if
2639 the queue_time_overall log selector is set), a one-byte patch to exigrep
2640 was needed to allow it to recognize "Completed" as not the last thing in
2643 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2644 patch that reportedly fixes this has been added. I am not expert enough to
2645 create a test for it. This is what the patch creator wrote:
2647 "I found a little strange behaviour of ldap code when working with
2648 Windows 2003 AD Domain, where users was placed in more than one
2649 Organization Units. When I tried to give exim partial DN, the exit code
2650 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2651 But simultaneously result of request was absolutely normal ldap result,
2652 so I produce this patch..."
2654 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2655 REFERENCE, so I have modified the code to exclude the patch when that macro
2658 55. Some experimental protocols are using DNS PTR records for new purposes. The
2659 keys for these records are domain names, not reversed IP addresses. The
2660 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2661 leaves it alone. Component reversal etc. now happens only for IP addresses.
2664 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2666 57. Double the size of the debug message buffer (to 2048) so that more of very
2667 long debug lines gets shown.
2669 58. The exicyclog utility now does better if the number of log files to keep
2670 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2672 59. Two changes related to the smtp_active_hostname option:
2674 (1) $smtp_active_hostname is now available as a variable.
2675 (2) The default for smtp_banner uses $smtp_active_hostname instead
2676 of $primary_hostname.
2678 60. The host_aton() function is supposed to be passed a string that is known
2679 to be a valid IP address. However, in the case of IPv6 addresses, it was
2680 not checking this. This is a hostage to fortune. Exim now panics and dies
2681 if the condition is not met. A case was found where this could be provoked
2682 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2683 components; fortuitously, this particular loophole had already been fixed
2684 by change 4.50/55 above.
2686 If there are any other similar loopholes, the new check in host_aton()
2687 itself should stop them being exploited. The report I received stated that
2688 data on the command line could provoke the exploit when Exim was running as
2689 exim, but did not say which command line option was involved. All I could
2690 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2691 running as the user.
2694 61. There was a buffer overflow vulnerability in the SPA authentication code
2695 (which came originally from the Samba project). I have added a test to the
2696 spa_base64_to_bits() function which I hope fixes it.
2699 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2700 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2702 63. The daemon start-up calls getloadavg() while still root for those OS that
2703 need the first call to be done as root, but it missed one case: when
2704 deliver_queue_load_max is set with deliver_drop_privilege. This is
2705 necessary for the benefit of the queue runner, because there is no re-exec
2706 when deliver_drop_privilege is set.
2708 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2709 This has been fixed.
2711 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2712 was in use, was not putting the data itself into the right store pool;
2713 consequently, it could be overwritten for a subsequent message in the same
2714 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2717 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2719 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2720 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2721 the function were treating the return as a boolean value, which happened to
2722 work because 0=false and not-0=true, but is not correct code.
2724 68. The host_aton() function was not handling scoped IPv6 addresses (those
2725 with, for example, "%eth0" on the end) correctly.
2727 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2728 negated items (that is, ~something) in unsigned ints. Some compilers
2729 apparently mutter when there is no cast.
2731 70. If an address verification called from an ACL failed, and did not produce a
2732 user-specific message (i.e. there was only a "system" message), nothing was
2733 put in $acl_verify_message. In this situation, it now puts the system
2736 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2737 session; change 4.31/43 added the unwanted input to the log line - except
2738 that it did not do this in the start of session case. It now does.
2740 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2741 This is wrong and can cause the other end to generate a synchronization
2742 error if it is another Exim or anything else that does the synchronization
2743 check. A QUIT command is no longer sent after a timeout.
2745 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2746 during host lookups.
2748 74. The defer_ok option of callout verification was not working if it was used
2749 when verifying addresses in header lines, that is, for this case:
2751 verify = header_sender/callout=defer_ok
2753 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2754 those file descriptors could be used for SMTP connections. If anything
2755 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2756 could be sent to the SMTP client, causing chaos. The daemon now opens
2757 stdin, stdout, and stderr to /dev/null when it puts itself into the
2760 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2761 log by default. The user can override this with suitable Perl magic.
2763 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2764 the log message when discard triggers, was not working for the DATA ACL or
2765 for the non-SMTP ACL.
2767 78. Error message wording change in sieve.c.
2769 79. If smtp_accept_max_per_host was set, the number of connections could be
2770 restricted to fewer than expected, because the daemon was trying to set up
2771 a new connection before checking whether the processes handling previous
2772 connections had finished. The check for completed processes is now done
2773 earlier. On busy systems, this bug wouldn't be noticed because something
2774 else would have woken the daemon, and it would have reaped the completed
2777 80. If a message was submitted locally by a user whose login name contained one
2778 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2779 It caused a spool format error. I have fixed the spool reading code. A
2780 related problem was that the "from" clause in the Received: line became
2781 illegal because of the space(s). It is now covered by ${quote_local_part.
2783 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2786 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2788 83. Patch from Sieve maintainer to fix unterminated string problem in
2789 "vacation" handling.
2791 84. Some minor changes to the Linux configuration files to help with other
2792 OS variants using glibc.
2794 85. One more patch for Sieve to update vacation handling to latest spec.
2797 ----------------------------------------------------
2798 See the note above about the 4.44 and 4.50 releases.
2799 ----------------------------------------------------
2805 1. Change 4.43/35 introduced a bug that caused file counts to be
2806 incorrectly computed when quota_filecount was set in an appendfile
2809 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2810 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2813 3. Give more explanation in the error message when the command for a transport
2814 filter fails to execute.
2816 4. There are several places where Exim runs a non-Exim command in a
2817 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2818 was being done only for the command run by the queryprogram router. It is
2819 now done for all such subprocesses. The other cases are: ${run, transport
2820 filters, and the commands run by the lmtp and pipe transports.
2822 5. Some older OS have a limit of 256 on the maximum number of file
2823 descriptors. Exim was using setrlimit() to set 1000 as a large value
2824 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2825 systems. I've change it so that if it can't get 1000, it tries for 256.
2827 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2828 was an oversight, and furthermore, ever since the addition of extra
2829 controls (e.g. 4.43/32), the checks on when to allow different forms of
2830 "control" were broken. There should now be diagnostics for all cases when a
2831 control that does not make sense is encountered.
2833 7. $recipients is now available in the predata ACL (oversight).
2835 8. Tidy the search cache before the fork to do a delivery from a message
2836 received from the command line. Otherwise the child will trigger a lookup
2837 failure and thereby defer the delivery if it tries to use (for example) a
2838 cached ldap connection that the parent has called unbind on.
2840 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2841 of $address_data from the recipient verification was clobbered by the
2842 sender verification.
2844 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2845 was its contents. (It was OK if the option was not defined at all.)
2847 11. A "Completed" log line is now written for messages that are removed from
2848 the spool by the -Mrm option.
2850 12. $host_address is now set to the target address during the checking of
2851 ignore_target_hosts.
2853 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2854 being passed; this would have caused $sender_host_name to have been used if
2855 matching the list had actually called for a host name (not very likely,
2856 since this list is usually IP addresses). A host name is now passed as
2859 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2860 code that specifies a non-restarting handler (typically sigaction() in
2861 modern systems) in an attempt to fix a rare and obscure crash bug.
2863 15. Narrowed the window for a race in the daemon that could cause it to ignore
2864 SIGCHLD signals. This is not a major problem, because they are used only to
2865 wake it up if nothing else does.
2867 16. A malformed maildirsize file could cause Exim to calculate negative values
2868 for the mailbox size or file count. Odd effects could occur as a result.
2869 The maildirsize information is now recalculated if the size or filecount
2872 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2873 support for a long time. Removed HAVE_SYS_VFS_H.
2875 18. Updated exipick to current release from John Jetmore.
2877 19. Allow an empty sender to be matched against a lookup in an address list.
2878 Previously the only cases considered were a regular expression, or an
2881 20. Exim went into a mad DNS lookup loop when doing a callout where the
2882 host was specified on the transport, if the DNS lookup yielded more than
2885 21. The RFC2047 encoding function was originally intended for short strings
2886 such as real names; it was not keeping to the 75-character limit for
2887 encoded words that the RFC imposes. It now respects the limit, and
2888 generates multiple encoded words if necessary. To be on the safe side, I
2889 have increased the buffer size for the ${rfc2047: expansion operator from
2892 22. Failure to deliver a bounce message always caused it to be frozen, even if
2893 there was an errors_to setting on the router. The errors_to setting is now
2896 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2897 canonical form (fully expanded) before being placed in
2898 $sender_host_address.
2900 24. Updated eximstats to version 1.33
2902 25. Include certificate and key file names in error message when GnuTLS fails
2903 to set them up, because the GnuTLS error message doesn't include the name
2904 of the failing file when there is a problem reading it.
2906 26. Expand error message when OpenSSL has problems setting up cert/key files.
2909 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2910 (this can affect the format of dates).
2912 28. exim_tidydb, when checking for the continued existence of a message for
2913 which it has found a message-specific retry record, was not finding
2914 messages that were in split spool directories. Consequently, it was
2915 deleting retry records that should have stayed in existence.
2917 29. eximstats updated to version 1.35
2918 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2919 1.35 - bugfix such that pie charts by volume are generated correctly
2921 30. The SPA authentication driver was not abandoning authentication and moving
2922 on to the next authenticator when an expansion was forced to fail,
2923 contradicting the general specification for all authenticators. Instead it
2924 was generating a temporary error. It now behaves as specified.
2926 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2927 (the order specifies the preference for clients). The order is now AES256,
2928 AES128, 3DES, ARCFOUR128.
2930 31. Small patch to Sieve code - explicitly set From: when generating an
2933 32. Exim crashed if a remote delivery caused a very long error message to be
2934 recorded - for instance if somebody sent an entire SpamAssassin report back
2935 as a large number of 550 error lines. This bug was coincidentally fixed by
2936 increasing the size of one of Exim's internal buffers (big_buffer) that
2937 happened as part of the Exiscan merge. However, to be on the safe side, I
2938 have made the code more robust (and fixed the comments that describe what
2941 33. Some experimental protocols are using DNS PTR records for new purposes. The
2942 keys for these records are domain names, not reversed IP addresses. The
2943 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2944 leaves it alone. Component reversal etc. now happens only for IP addresses.
2947 34. The host_aton() function is supposed to be passed a string that is known
2948 to be a valid IP address. However, in the case of IPv6 addresses, it was
2949 not checking this. This is a hostage to fortune. Exim now panics and dies
2950 if the condition is not met. A case was found where this could be provoked
2951 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2952 components; fortuitously, this particular loophole had already been fixed
2953 by change 4.50/55 or 4.44/33 above.
2955 If there are any other similar loopholes, the new check in host_aton()
2956 itself should stop them being exploited. The report I received stated that
2957 data on the command line could provoke the exploit when Exim was running as
2958 exim, but did not say which command line option was involved. All I could
2959 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2960 running as the user.
2963 35. There was a buffer overflow vulnerability in the SPA authentication code
2964 (which came originally from the Samba project). I have added a test to the
2965 spa_base64_to_bits() function which I hope fixes it.
2968 36. The daemon start-up calls getloadavg() while still root for those OS that
2969 need the first call to be done as root, but it missed one case: when
2970 deliver_queue_load_max is set with deliver_drop_privilege. This is
2971 necessary for the benefit of the queue runner, because there is no re-exec
2972 when deliver_drop_privilege is set.
2974 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2975 was in use, was not putting the data itself into the right store pool;
2976 consequently, it could be overwritten for a subsequent message in the same
2977 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2980 38. Sometimes the final signoff response after QUIT could fail to get
2981 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2982 < 0 before doing a fflush(). This bug looks as though it goes back to the
2983 introduction of TLS in release 3.20, but "sometimes" must have been rare
2984 because the tests only now provoked it.
2990 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2991 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2992 Now it returns an integer. A number of calls were still expecting a T/F
2993 return. Fortuitously, in all cases, the tests worked in OK situations,
2994 which is the norm. However, things would have gone wrong on any write
2995 failures on the smtp file descriptor. This function is used when sending
2996 messages over SMTP and also when doing verify callouts.
2998 2. When Exim is called to do synchronous delivery of a locally submitted
2999 message (the -odf or -odi options), it no longer closes stderr before doing
3002 3. Implemented the mua_wrapper option.
3004 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3006 5. Implemented the functions header_remove(), header_testname(),
3007 header_add_at_position(), and receive_remove_recipient(), and exported them
3010 6. If an ACL "warn" statement specified the addition of headers, Exim already
3011 inserted X-ACL-Warn: at the start if there was no header name. However, it
3012 was not making this test for the second and subsequent header lines if
3013 there were newlines in the string. This meant that an invalid header could
3014 be inserted if Exim was badly configured.
3016 7. Allow an ACL "warn" statement to add header lines at the start or after all
3017 the Received: headers, as well as at the end.
3019 8. Added the rcpt_4xx retry error code.
3021 9. Added postmaster_mailfrom=xxx to callout verification option.
3023 10. Added mailfrom=xxxx to the callout verification option, for verify=
3026 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3027 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3029 12. Inserted some casts to stop certain compilers warning when using pointer
3030 differences as field lengths or precisions in printf-type calls (mostly
3031 affecting debugging statements).
3033 13. Added optional readline() support for -be (dynamically loaded).
3035 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3036 same clock tick as a message's arrival, so that its received time was the
3037 same as the "first fail" time on the retry record, and that message
3038 remained on the queue past the ultimate address timeout, every queue runner
3039 would try a delivery (because it was past the ultimate address timeout) but
3040 after another failure, the ultimate address timeout, which should have then
3041 bounced the address, did not kick in. This was a "< instead of <=" error;
3042 in most cases the first failure would have been in the next clock tick
3043 after the received time, and all would be well.
3045 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3046 being recognized when the domain list was tested by the match_domain
3047 condition in an expansion string.
3049 16. Added the ${str2b64: operator.
3051 17. Exim was always calling setrlimit() to set a large limit for the number of
3052 processes, without checking whether the existing limit was already
3053 adequate. (It did check for the limit on file descriptors.) Furthermore,
3054 errors from getrlimit() and setrlimit() were being ignored. Now they are
3055 logged to the main and panic logs, but Exim does carry on, to try to do its
3056 job under whatever limits there are.
3058 18. Imported PCRE 5.0.
3060 19. Trivial typo in log message " temporarily refused connection" (the leading
3063 20. If the log selector return_path_on_delivery was set and an address was
3064 redirected to /dev/null, the delivery process crashed because it assumed
3065 that a return path would always be set for a "successful" delivery. In this
3066 case, the whole delivery is bypassed as an optimization, and therefore no
3069 21. Internal re-arrangement: the function for sending a challenge and reading
3070 a response while authentication was assuming a zero-terminated challenge
3071 string. It's now changed to take a pointer and a length, to allow for
3072 binary data in such strings.
3074 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3076 23. Exim was not respecting finduser_retries when seeking the login of the
3077 uid under which it was called; it was always trying 10 times. (The default
3078 setting of finduser_retries is zero.) Also, it was sleeping after the final
3079 failure, which is pointless.
3081 24. Implemented tls_on_connect_ports.
3083 25. Implemented acl_smtp_predata.
3085 26. If the domain in control=submission is set empty, Exim assumes that the
3086 authenticated id is a complete email address when it generates From: or
3087 Sender: header lines.
3089 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3090 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3091 chown and chgrp in /bin and hostname in /usr/bin.
3093 28. Exim was keeping the "process log" file open after each use, just as it
3094 does for the main log. This opens the possibility of it remaining open for
3095 long periods when the USR1 signal hits a daemon. Occasional processlog
3096 errors were reported, that could have been caused by this. Anyway, it seems
3097 much more sensible not to leave this file open at all, so that is what now
3100 29. The long-running daemon process does not normally write to the log once it
3101 has entered its main loop, and it closes the log before doing so. This is
3102 so that log files can straightforwardly be renamed and moved. However,
3103 there are a couple of unusual error situations where the daemon does write
3104 log entries, and I had neglected to close the log afterwards.
3106 30. The text of an SMTP error response that was received during a remote
3107 delivery was being truncated at 512 bytes. This is too short for some of
3108 the long messages that one sometimes sees. I've increased the limit to
3111 31. It is now possible to make retry rules that apply only when a message has a
3112 specific sender, in particular, an empty sender.
3114 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3115 it possible to be selective about when SMTP synchronization is enforced.
3117 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3119 32. Implemented hosts_connection_nolog.
3121 33. Added an ACL for QUIT.
3123 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3126 35. Added mailbox_size and mailbox_filecount to appendfile.
3128 36. Added control = no_multiline_responses to ACLs.
3130 37. There was a bug in the logic of the code that waits for the clock to tick
3131 in the case where the clock went backwards by a substantial amount such
3132 that the microsecond fraction of "now" was more than the microsecond
3133 fraction of "then" (but the whole seconds number was less).
3135 38. Added support for the libradius Radius client library this is found on
3136 FreeBSD (previously only the radiusclient library was supported).
3142 1. When certain lookups returned multiple values in the form name=value, the
3143 quoting of the values was not always being done properly. Specifically:
3144 (a) If the value started with a double quote, but contained no whitespace,
3146 (b) If the value contained whitespace other than a space character (i.e.
3147 tabs or newlines or carriage returns) it was not quoted.
3148 This fix has been applied to the mysql and pgsql lookups by writing a
3149 separate quoting function and calling it from the lookup code. The fix
3150 should probably also be applied to nisplus, ibase and oracle lookups, but
3151 since I cannot test any of those, I have not disturbed their existing code.
3153 2. A hit in the callout cache for a specific address caused a log line with no
3154 reason for rejecting RCPT. Now it says "Previous (cached) callout
3155 verification failure".
3157 3. There was an off-by-one bug in the queryprogram router. An over-long
3158 return line was truncated at 256 instead of 255 characters, thereby
3159 overflowing its buffer with the terminating zero. As well as fixing this, I
3160 have increased the buffer size to 1024 (and made a note to document this).
3162 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3163 when Exim is waiting for an SMTP response from a remote server, Exim
3164 restarts its select() call on the socket, thereby resetting its timeout.
3165 This is not a problem when such interrupts are rare. Somebody set up a cron
3166 job to run exiwhat every 2 minutes, which is less than the normal select()
3167 timeout (5 or 10 minutes). This meant that the select() timeout never
3168 kicked in because it was always reset. I have fixed this by comparing the
3169 time when an interrupt arrives with the time at the start of the first call
3170 to select(). If more time than the timeout has elapsed, the interrupt is
3171 treated as a timeout.
3173 5. Some internal re-factoring in preparation for the addition of Sieve
3174 extensions (by MH). In particular, the "personal" test is moved to a
3175 separate function, and given an option for scanning Cc: and Bcc: (which is
3176 not set for Exim filters).
3178 6. When Exim created an email address using the login of the caller as the
3179 local part (e.g. when creating a From: or Sender: header line), it was not
3180 quoting the local part when it contained special characters such as @.
3182 7. Installed new OpenBSD configuration files.
3184 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3185 try to make them clearer.
3187 9. Callout options, other than the timeout value, were being ignored when
3188 verifying sender addresses in header lines. For example, when using
3190 verify = header_sender/callout=no_cache
3192 the cache was (incorrectly) being used.
3194 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3195 only the Cygwin environment.
3197 11. When return_path_on_delivery was set as a log selector, if different remote
3198 addresses in the same message used different return paths and parallel
3199 remote delivery occurred, the wrong values would sometimes be logged.
3200 (Whenever a remote delivery process finished, the return path value from
3201 the most recently started remote delivery process was logged.)
3203 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3204 lines when AUTH and/or TLS are in use. This is the "received protocol"
3205 field. Exim used to use "asmtp" for authenticated SMTP, without any
3206 indication (in the protocol name) for TLS use. Now it follows the RFC and
3207 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3208 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3209 names appear in log lines as well as in Received: header lines.
3211 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3212 extensions, and comparison tests, and to fix some bugs.
3214 14. Changes to the "personal" filter test:
3216 (1) The test was buggy in that it was just doing the equivalent of
3217 "contains" tests on header lines. For example, if a user's address was
3218 anne@some.where, the "personal" test would incorrectly be true for
3220 To: susanne@some.where
3222 This test is now done by extracting each address from the header in turn,
3223 and checking the entire address. Other tests that are part of "personal"
3224 are now done using regular expressions (for example, to check local parts
3225 of addresses in From: header lines).
3227 (2) The list of non-personal local parts in From: addresses has been
3228 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3229 taken from the Sieve specification recommendations.
3231 (3) If the message contains any header line starting with "List-" it is
3232 treated as non-personal.
3234 (4) The test for "circular" in the Subject: header line has been removed
3235 because it now seems ill-conceived.
3237 15. Minor typos in src/EDITME comments corrected.
3239 16. Installed latest exipick from John Jetmore.
3241 17. If headers_add on a router specified a text string that was too long for
3242 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3243 of string_sprintf() is now avoided.
3245 18. $message_body_size was not set (it was always zero) when running the DATA
3246 ACL and the local_scan() function.
3248 19. For the "mail" command in an Exim filter, no default was being set for
3249 the once_repeat time, causing a random time value to be used if "once" was
3250 specified. (If the value happened to be <= 0, no repeat happened.) The
3251 default is now 0s, meaning "never repeat". The "vacation" command was OK
3252 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3253 (I found it when inspecting the code).
3255 20. There is now an overall timeout for performing a callout verification. It
3256 defaults to 4 times the callout timeout, which applies to individual SMTP
3257 commands during the callout. The overall timeout applies when there is more
3258 than one host that can be tried. The timeout is checked before trying the
3259 next host. This prevents very long delays if there are a large number of
3260 hosts and all are timing out (e.g. when the network connections are timing
3261 out). The value of the overall timeout can be changed by specifying an
3262 additional sub-option for "callout", called "maxwait". For example:
3264 verify = sender/callout=5s,maxwait=20s
3266 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3267 to the end before writing, but this should make it even safer).
3269 22. Exim was forgetting that it had advertised PIPELINING for the second and
3270 subsequent messages on an SMTP connection. It was also not resetting its
3271 memory on STARTTLS and an internal HELO.
3273 23. When Exim logs an SMTP synchronization error within a session, it now
3274 records whether PIPELINING has been advertised or not.
3276 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3277 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3278 rather than long int.
3280 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3282 26. Added the never_mail option to autoreply.
3288 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3289 crash if the getsockname() call failed; this can happen if a connection is
3290 closed very soon after it is established. The problem was simply in the
3291 order in which certain operations were done, causing Exim to try to write
3292 to the SMTP stream before it had set up the file descriptor. The bug has
3293 been fixed by making things happen in the correct order.
3299 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3300 before the connection was closed, thus losing the rejection response.
3302 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3303 some early Solaris releases, but causes trouble in current releases where
3304 socklen_t is defined.
3306 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3309 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3312 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3313 that releasing the top of it at the end releases what was used for sub-
3314 expansions (unless the block got too big). However, discard this block if
3315 the first thing is a variable or header, so that we can use its block when
3316 it is dynamic (useful for very large $message_headers, for example).
3318 6. Lookups now cache *every* query, not just the most recent. A new, separate
3319 store pool is used for this. It can be recovered when all lookup caches are
3320 flushed. Lookups now release memory at the end of their result strings.
3321 This has involved some general refactoring of the lookup sources.
3323 7. Some code has been added to the store_xxx() functions to reduce the amount
3324 of flapping under certain conditions.
3326 8. log_incoming_interface used to affect only the <= reception log lines. Now
3327 it causes the local interface and port to be added to several more SMTP log
3328 lines, for example "SMTP connection from", and rejection lines.
3330 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3332 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3334 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3335 could be overwritten at the end of the current message (or the start of a
3336 new message if it was set in a HELO ACL). The value is now preserved for
3337 the duration of the SMTP connection.
3339 12. If a transport had a headers_rewrite setting, and a matching header line
3340 contained an unqualified address, that address was qualified, even if it
3341 did not match any rewriting rules. The underlying bug was that the values
3342 of the flags that permit the existence of unqualified sender and recipient
3343 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3344 non-local messages, and by -bnq for local messages) were not being
3345 preserved with the message after it was received.
3347 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3348 "next input=" as part of the text comprising the host identity instead of
3349 the correct text. The code was using the same buffer for two different
3350 strings. However, depending on which order the printing function evaluated
3351 its arguments, the bug did not always show up. Under Linux, for example, my
3352 test suite worked just fine.
3354 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3355 doesn't work with some older versions of Perl. It has been changed to "my",
3356 which in any case is probably the better facility to use.
3358 15. A really picky compiler found some instances of statements for creating
3359 error messages that either had too many or two few arguments for the format
3362 16. The size of the buffer for calls to the DNS resolver has been increased
3363 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3364 for addresses that have a lot of PTR records. This alleviates a problem; it
3365 does not fully solve it.
3367 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3368 buffer now truncates the list and logs the incident, which is the same
3369 action as happens when Exim is looking up a host name and its aliases.
3370 Previously in this situation something unpredictable would happen;
3371 sometimes it was "internal error: store_reset failed".
3373 18. If a server dropped the connection unexpectedly when an Exim client was
3374 using GnuTLS and trying to read a response, the client delivery process
3375 crashed while trying to generate an error log message.
3377 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3378 string, for example:
3380 warn message = H1: something\nH2: something
3382 the text was added as a single header line from Exim's point of view
3383 though it ended up OK in the delivered message. However, searching for the
3384 second and subsequent header lines using $h_h2: did not work. This has been
3385 fixed. Similarly, if a system filter added multiple headers in this way,
3386 the routers could not see them.
3388 20. Expanded the error message when iplsearch is called with an invalid key to
3389 suggest using net-iplsearch in a host list.
3391 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3392 ACLs are no longer actually imposed (and a message to that effect is
3395 22. If a "gecos" field in a passwd entry contained escaped characters, in
3396 particular, if it contained a \" sequence, Exim got it wrong when building
3397 a From: or a Sender: header from that name. A second bug also caused
3398 incorrect handling when an unquoted " was present following a character
3399 that needed quoting.
3401 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3402 was not being matched caselessly.
3404 24. Arranged for all hyphens in the exim.8 source to be escaped with
3407 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3408 properly. Recipient callout cache records were still being keyed to include
3409 the sender, even when use_sender was set false. This led to far more
3410 callouts that were necessary. The sender is no longer included in the key
3411 when use_sender is false.
3413 26. Added "control = submission" modifier to ACLs.
3415 27. Added the ${base62d: operator to decode base 62 numbers.
3417 28. dnsdb lookups can now access SRV records.
3419 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3420 the configuration file.
3422 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3423 (-v) mode. This makes the output for a verbose queue run more intelligible.
3425 31. Added a use_postmaster feature to recipient callouts.
3427 32. Added the $body_zerocount variable, containing the number of binary zero
3428 bytes in the message body.
3430 33. The time of last modification of the "new" subdirectory is now used as the
3431 "mailbox time last read" when there is a quota error for a maildir
3434 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3436 35. Added +ignore_unknown as a special item in host lists.
3438 36. Code for decoding IPv6 addresses in host lists is now included, even if
3439 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3440 address was recognized as an IP address, but was then not correctly decoded
3441 into binary, causing unexpected and incorrect effects when compared with
3448 1. Very minor rewording of debugging text in manualroute to say "list of
3449 hosts" instead of "hostlist".
3451 2. If verify=header_syntax was set, and a header line with an unqualified
3452 address (no domain) and a large number of spaces between the end of the
3453 name and the colon was received, the reception process suffered a buffer
3454 overflow, and (when I tested it) crashed. This was caused by some obsolete
3455 code that should have been removed. The fix is to remove it!
3457 3. When running in the test harness, delay a bit after writing a bounce
3458 message to get a bit more predictability in the log output.
3460 4. Added a call to search_tidyup() just before forking a reception process. In
3461 theory, someone could use a lookup in the expansion of smtp_accept_max_
3462 per_host which, without the tidyup, could leave open a database connection.
3464 5. Added the variables $recipient_data and $sender_data which get set from a
3465 lookup success in an ACL "recipients" or "senders" condition, or a router
3466 "senders" option, similar to $domain_data and $local_part_data.
3468 6. Moved the writing of debug_print from before to after the "senders" test
3471 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3472 problems for message scanning, either using a data ACL, or using
3473 local_scan() because the Received: header was not generated till after they
3474 were called (in order to set the time as the time of reception completion).
3475 I have revised the way this works. The header is now generated after the
3476 body is received, but before the ACL or local_scan() are called. After they
3477 are run, the timestamp in the header is updated.
3483 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3484 before starting a queue runner without re-exec. This happened only when
3485 deliver_drop_privilege was set or when the Exim user was set to root. The
3486 effect of the bug was that timeouts during subsequent deliveries caused
3487 crashes instead of being properly handled. The handler is now left at its
3488 default (and expected) setting.
3490 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3491 message, again when deliver_drop_privilege is set or Exim is run as root.
3492 The bug described in (1) was not present in this case, but the tidying up
3493 of the other signals was missing. I have made the two cases consistent.
3495 3. The ignore_target_hosts setting on a manualroute router was being ignored
3496 for hosts that were looked up using the /MX notation.
3498 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3501 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3502 operated on the sender address. After changing the $sender_address to <>
3503 for the sender address verify, Exim was re-instated it as the original
3504 (before rewriting) address, but remembering that it had rewritten it, so it
3505 wasn't rewriting it again. This bug also had the effect of breaking the
3506 sender address verification caching when the sender address was rewritten.
3508 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3509 This has been changed so that if the ip literal address matches
3510 ignore_target_hosts, the router declines.
3512 7. Added expansion conditions match_domain, match_address, and match_local_
3513 part (NOT match_host).
3515 8. The placeholder for the Received: header didn't have a length field set.
3517 9. Added code to Exim itself and to exim_lock to test for a specific race
3518 condition that could lead to file corruption when using MBX delivery. The
3519 issue is with the lockfile that is created in /tmp. If this file is removed
3520 after a process has opened it but before that process has acquired a lock,
3521 there is the potential for a second process to recreate the file and also
3522 acquire a lock. This could lead to two Exim processes writing to the file
3523 at the same time. The added code performs the same test as UW imapd; it
3524 checks after acquiring the lock that its file descriptor still refers to
3525 the same named file.
3527 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3528 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3531 11. Added the smtp_active_hostname option. If used, this will typically be made
3532 to depend on the incoming interface address. Because $interface_address is
3533 not set up until the daemon has forked a reception process, error responses
3534 that can happen earlier (such as "too many connections") no longer contain
3537 12. If an expansion in a condition on a "warn" statement fails because a lookup
3538 defers, the "warn" statement is abandoned, and the next ACL statement is
3539 processed. Previously this caused the whole ACL to be aborted.
3541 13. Added the iplsearch lookup type.
3543 14. Added ident_timeout as a log selector.
3545 15. Added tls_certificate_verified as a log selector.
3547 16. Added a global option tls_require_ciphers (compare the smtp transport
3548 option of the same name). This controls incoming TLS connections.
3550 17. I finally figured out how to make tls_require_ciphers do a similar thing
3551 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3552 before starting the TLS session.
3554 18. Tabs are now shown as \t in -bP output.
3556 19. If the log selector return_path_on_delivery was set, Exim crashed when
3557 bouncing a message because it had too many Received: header lines.
3559 20. If two routers both had headers_remove settings, and the first one included
3560 a superfluous trailing colon, the final name in the first list and the
3561 first name in the second list were incorrectly joined into one item (with a
3562 colon in the middle).
3568 1. Added -C and -D options to the exinext utility, mainly to make it easier
3569 to include in the automated testing, but these could be helpful when
3570 multiple configurations are in use.
3572 2. The exinext utility was not formatting the output nicely when there was
3573 an alternate port involved in the retry record key, nor when there was a
3574 message id as well (for retries that were specific to a specific message
3575 and a specific host). It was also confused by IPv6 addresses, because of
3576 the additional colons they contain. I have fixed the IPv4 problem, and
3577 patched it up to do a reasonable job for IPv6.
3579 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3580 delivery, the log line now contains "pipelined" if PIPELINING was used.
3582 4. An SMTP transport process used to panic and die if the bind() call to set
3583 an explicit outgoing interface failed. This has been changed; it is now
3584 treated in the same way as a connect() failure.
3586 5. A reference to $sender_host_name in the part of a conditional expansion
3587 that was being skipped was still causing a DNS lookup. This no longer
3590 6. The def: expansion condition was not recognizing references to header lines
3591 that used bh_ and bheader_.
3593 7. Added the _cache feature to named lists.
3595 8. The code for checking quota_filecount in the appendfile transport was
3596 allowing one more file than it should have been.
3598 9. For compatibility with Sendmail, the command line option
3606 and sets the incoming protocol and host name (for trusted callers). The
3607 host name and its colon can be omitted when only the protocol is to be set.
3608 Note the Exim already has two private options, -pd and -ps, that refer to
3609 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3610 "s", but I don't think that's a major issue.
3612 10. A number of refactoring changes to the code, none of which should affect
3615 (a) The number of logging options was getting close to filling up the
3616 32-bit word that was used as a bit map. I have split them into two classes:
3617 those that are passed in the argument to log_write(), and those that are
3618 only ever tested independently outside of that function. These are now in
3619 separate 32-bit words, so there is plenty of room for expansion again.
3620 There is no change in the user interface or the logging behaviour.
3622 (b) When building, for example, log lines, the code previously used a
3623 macro that called string_cat() twice, in order to add two strings. This is
3624 not really sufficiently general. Furthermore, there was one instance where
3625 it was actually wrong because one of the argument was used twice, and in
3626 one call a function was used. (As it happened, calling the function twice
3627 did not affect the overall behaviour.) The macro has been replaced by a
3628 function that can join an arbitrary number of extra strings onto a growing
3631 (c) The code for expansion conditions now uses a table and a binary chop
3632 instead of a serial search (which was left over from when there were very
3633 few conditions). Also, it now recognizes conditions like "pam" even when
3634 the relevant support is not compiled in: a suitably worded error message is
3635 given if an attempt is made to use such a condition.
3637 11. Added ${time_interval:xxxxx}.
3639 12. A bug was causing one of the ddress fields not to be passed back correctly
3640 from remote delivery subprocesses. The field in question was not being
3641 subsequently used, so this caused to problems in practice.
3643 13. Added new log selectors queue_time and deliver_time.
3645 14. Might have fixed a bug in maildirsizefile handling that threw up
3646 "unexpected character" debug warnings, and recalculated the data
3647 unnecessarily. In any case, I expanded the warning message to give more
3650 15. Added the message "Restricted characters in address" to the statements in
3651 the default ACL that block characters like @ and % in local parts.
3653 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3654 Three changes have been made:
3656 (a) There was a serious bug; a negative response to MAIL caused the whole
3657 recipient domain to be cached as invalid, thereby blocking all messages
3658 to all local parts at the same domain, from all senders. This bug has
3659 been fixed. The domain is no longer cached after a negative response to
3660 MAIL if the sender used is not empty.
3662 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3665 (c) A new callout option, "use_sender" has been added for people who want
3666 the modified behaviour.
3672 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3675 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3676 indeed breaks things for older releases.
3678 3. Added additional logging to the case where there is a problem reading data
3679 from a filter that is running in a subprocess using a pipe, in order to
3680 try to track down a specific problem.
3682 4. Testing facility fudge: when running in the test harness and attempting
3683 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3684 getting "No route to host". Convert this to a timeout.
3686 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3689 6. Some OS don't have socklen_t but use size_t instead. This affects the
3690 fifth argument of getsockopt() amongst other things. This is now
3691 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3692 can be set for individual OS. I have set it for SunOS5, OSF1, and
3693 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3694 some earlier ones do not.
3696 7. Change 4.30/15 was not doing the test caselessly.
3698 8. The standard form for an IPv6 address literal was being rejected by address
3699 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3700 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3701 this, as well as the form without the "IPv6" on the front (but only when
3702 address literals are enabled, of course).
3704 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3706 10. Exim crashed if a message with an empty sender address specified by -f
3707 encountered a router with an errors_to setting. This could be provoked only
3708 by a command such as
3712 where an empty string was supplied; "<>" did not hit this bug.
3714 11. Installed PCRE release 4.5.
3716 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3717 remained set. It is now erased.
3719 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3720 times from message ids (which are base 36 rather than the normal 62).
3722 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3723 were being counted as actual protocol errors, and logged if the log
3724 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3725 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3726 and DATA following a set of rejected RCPTs do not count as protocol errors.
3727 In other words, Exim assumes they were pipelined, though this may not
3728 actually be the case. Of course, in all cases the client gets an
3729 appropriate error code.
3731 15. If a lookup fails in an ACL condition, a message about the failure may
3732 be available; it is used if testing the ACL cannot continue, because most
3733 such messages specify what the cause of the deferral is. However, some
3734 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3735 that caused an old message to be retained and used if a later statement
3736 caused a defer, replacing the real cause of the deferral.
3738 16. If an IP address had so many PTR records that the DNS lookup buffer
3739 was not large enough to hold them, Exim could crash while trying to process
3740 the truncated data. It now detects and logs this case.
3742 17. Further to 4.21/58, another change has been made: if (and only if) the
3743 first line of a message (the first header line) ends with CRLF, a bare LF
3744 in a subsequent header line has a space inserted after it, so as not to
3745 terminate the header.
3747 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3748 unnecessarily, used atoi() instead of strtol(), and didn't check the
3749 termination when getting file sizes from file names by regex.
3751 19. Completely re-implemented the support for maildirsize files, in the light
3752 of a number of problems with the previous contributed implementation
3753 (4.30/29). In particular:
3755 . If the quota is zero, the maildirsize file is maintained, but no quota is
3758 . If the maildir directory does not exist, it is created before any attempt
3759 to write a maildirsize file.
3761 . The quota value in the file is just a cache; if the quota is changed in
3762 the transport, the new value overrides.
3764 . A regular expression is available for excluding directories from the
3767 20. The autoreply transport checks the characters in options that define the
3768 message's headers; it allows continued headers, but it was checking with
3769 isspace() after an embedded newline instead of explicitly looking for a
3772 21. If all the "regular" hosts to which an address was routed had passed their
3773 expiry times, and had not reached their retry times, the address was
3774 bounced, even if fallback hosts were defined. Now Exim should go on to try
3777 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3778 equivalent code in the SMTP transport. Some hosts send humungous responses
3779 to HELO/EHLO, more than 1024 it seems.
3781 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3782 gives compiler warnings in some environments. I've now done it "properly",
3785 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3786 (because of problems with the built-in one) was declared to return uschar *
3787 instead of char *, causing compiler failure.
3789 25. Fixed a file descriptor leak when processing alias/forward files.
3791 26. Fixed a minor format string issue in dbfn.c.
3793 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3795 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3796 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3799 29. When a sender address is verified, it is cached, to save repeating the test
3800 when there is more than one recipient in a message. However, when the
3801 verification involves a callout, it is possible for different callout
3802 options to be set for different recipients. It is too complicated to keep
3803 track of this in the cache, so now Exim always runs a verification when a
3804 callout is required, relying on the callout cache for the optimization.
3805 The overhead is duplication of the address routing, but this should not be
3808 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3809 to be verified with callout=postmaster, and the main callout worked but the
3810 postmaster check failed, the verification correctly failed. However, if a
3811 subsequent RCPT command asked for sender verification *without* the
3812 postmaster check, incorrect caching caused this verification also to fail,
3815 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3816 it was not caching the DNS options (qualify_single, search_parents) that
3817 were used when the lookup failed. A subsequent lookup with different
3818 options therefore always gave the same answer, though there were cases
3819 where it should not have. (Example: a "domains = !$mx_any" option on a
3820 dnslookup router: the "domains" option is always processed without any
3821 widening, but the router might have qualify_single set.) Now Exim uses the
3822 cached value only when the same options are set.
3824 32. Added John Jetmore's "exipick" utility to the distribution.
3826 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3827 than a timeout (e.g. a certificate is required, and is not provided), an
3828 Exim server now closes the connection immediately. Previously it waited for
3829 the client to close - but if the client is SSL, it seems that they each
3830 wait for each other, leading to a delay before one of them times out.
3832 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3833 maintained 0.8.x compatibility because I don't think many are using it, and
3834 it is clearly obsolete.
3836 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3839 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3840 client certificate was expired. A simple patch fixes this, though I don't
3841 understand the full logic of why the verify callback is called multiple
3844 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3845 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3846 which causes problems with some clients (such as the Certicom SSL Plus
3847 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3848 disables the coutermeasure allowing Eudora to connect."
3850 38. Exim was not checking that a write() to a log file succeeded. This could
3851 lead to Bad Things if a log got too big, in particular if it hit a file
3852 size limit. Exim now panics and dies if it cannot write to a log file, just
3853 as it does if it cannot open a log file.
3855 39. Modified OS/Makefile-Linux so that it now contains
3857 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3859 The two -D definitions ensure that Exim is compiled with large file
3860 support, which makes it possible to handle log files that are bigger than
3863 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3864 instance) a domain was checked against a named list that involved a lookup,
3865 causing $domain_data to be set, then another domain was checked against the
3866 same list, then the first domain was re-checked, the value of $domain_data
3867 after the final check could be wrong. In particular, if the second check
3868 failed, it could be set empty. This bug probably also applied to
3871 41. The strip_trailing_dot option was not being applied to the address given
3872 with the -f command-line option.
3874 42. The code for reading a message's header from the spool was incrementing
3875 $received_count, but never initializing it. This meant that the value was
3876 incorrect (doubled) while delivering a message in the same process in which
3877 it was received. In the most common configuration of Exim, this never
3878 happens - a fresh exec is done - but it can happen when
3879 deliver_drop_privilege is set.
3881 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3882 it now includes up to 150 characters of the unexpected data in the log
3885 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3886 and building data strings. The size of both of these buffers was 10 000
3887 bytes - far larger than anybody would *ever* want, thought I. Needless to
3888 say, somebody hit the limit. I have increased the maximum line length to
3889 20 000 and the maximum data length of concatenated lines to 100 000. I have
3890 also fixed two bugs, because there was no checking on these buffers. Tsk,
3891 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3892 buffer is too small.
3894 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3895 lsearch lookups. Now it does.
3897 46. When parsing a route_list item in a manualroute router, a fixed-length
3898 buffer was used for the list of hosts. I made this 1024 bytes long,
3899 thinking that nobody would ever have a list of hosts that long. Wrong.
3900 Somebody had a whole pile of complicated expansion conditions, and the
3901 string was silently truncated, leading to an expansion error. It turns out
3902 that it is easier to change to an unlimited length (owing to other changes
3903 that have happened since this code was originally written) than to build
3904 structure for giving a limitation error. The length of the item that
3905 expands into the list of hosts is now unlimited.
3907 47. The lsearch lookup could not handle data where the length of text line was
3908 more than 4095 characters. Such lines were truncated, leading to shortened
3909 data being returned. It should now handle lines of any length.
3911 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3912 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3915 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3916 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3917 no longer happens in comment lines. A list of replacements is now placed
3918 at the head of all of the source files, except those whose only change is
3919 to replace PERL_COMMAND in the very first #! line.
3921 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3922 messages on the queue, with a bottom-up merge sort, using code contributed
3923 by Michael Haardt. This should make operations like -bp somewhat faster on
3924 large queues. It won't affect queue runners, except when queue_run_in_order
3927 51. Installed eximstats 1.31 in the distribution.
3929 52. Added support for SRV lookups to the dnslookup router.
3931 53. If an ACL referred to $message_body or $message_body_end, the value was not
3932 reset for any messages that followed in the same SMTP session.
3934 54. The store-handling optimization for building very long strings was not
3935 differentiating between the different store pools. I don't think this
3936 actually made any difference in practice, but I've tidied it.
3938 55. While running the routers to verify a sender address, $sender_address
3939 was still set to the sender address. This is wrong, because when routing to
3940 send a bounce to the sender, it would be empty. Therefore, I have changed
3941 it so that, while verifying a sender address, $sender_address is set to <>.
3942 (There is no change to what happens when verifying a recipient address.)
3944 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3945 target A or AAAA records (if not already returned) without resetting the
3946 qualify_single or search_parents options of the DNS resolver. These are
3947 inappropriate in this case because the targets of MX and SRV records must
3948 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3949 target that, when qualified, matched something in the local domain. These
3950 two options are now turned off when doing these lookups.
3952 57. It seems that at least some releases of Reiserfs (which does not have the
3953 concept of a fixed number of inodes) returns zero and not -1 for the
3954 number of available inodes. This interacted badly with check_spool_inodes,
3955 which assumed that -1 was the "no such thing" setting. What I have done is
3956 to check that the total number of inodes is greater than zero before doing
3957 the test of how many are available.
3959 58. When a "warn" ACL statement has a log_message modifier, the message is
3960 remembered, and not repeated. This is to avoid a lot of repetition when a
3961 message has many recipients that cause the same warning to be written.
3962 Howewer, Exim was preserving the list of already written lines for an
3963 entire SMTP session, which doesn't seem right. The memory is now reset if a
3964 new message is started.
3966 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3967 debugging output unless log_rewrite was also set.
3969 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3970 of (int)(handle) when we know that handle contains (void *)(-1).
3972 61. The Exim daemon panic-logs an error return when it closes the incoming
3973 connection. However "connection reset by peer" seems to be common, and
3974 isn't really an error worthy of noting specially, so that particular error
3977 62. When Exim is trying to find all the local interfaces, it used to panic and
3978 die if the ioctl to get the interface flags failed. However, it seems that
3979 on at least one OS (Solaris 9) it is possible to have an interface that is
3980 included in the list of interfaces, but for which you get a failure error
3981 for this call. This happens when the interface is not "plumbed" into a
3982 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3983 failure of the "get flags" call assumes that the interface is down.
3985 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3986 makes life easier for people who are doing arithmetic on fields extracted
3987 from dates, where you often get leading zeros that should not be
3988 interpreted as octal.
3990 64. Added qualify_domain to the redirect router, to override the global
3993 65. If a pathologically long header line contained very many addresses (the
3994 report of this problem mentioned 10 000) and each of them was rewritten,
3995 Exim could use up a very large amount of memory. (It kept on making new
3996 copies of the header line as it rewrote, and never released the old ones.)
3997 At the expense of a bit more processing, the header rewriting function has
3998 been changed so that it no longer eats memory in this way.
4000 66. The generation of the Received: header has been moved from the time that a
4001 message starts to be received, to the time that it finishes. The timestamp
4002 in the Received: header should now be very close to that of the <= log
4003 line. There are two side-effects of this change:
4005 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4006 logged header lines no longer include the local Received: line, because
4007 it has not yet been created. The same applies to a copy of the message
4008 that is returned to a non-SMTP sender when a message is rejected.
4010 (b) When a filter file is tested using -bf, no additional Received: header
4011 is added to the test message. After some thought, I decided that this
4014 This change does not affect the value of $received_for. It is still set
4015 after address rewriting, but before local_scan() is called.
4017 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4019 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4020 gave an unhelpful panic error message, and a defer error. I have managed to
4021 change this behaviour so that it now rejects any supplied certificate,
4022 which seems right, as the list of acceptable certificates is empty.
4024 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4025 gave an unhelpful defer error. I have not managed to make this reject any
4026 supplied certificates, but the error message it gives is "no certificate
4027 supplied", which is not helpful.
4029 70. exigrep's output now also includes lines that are not associated with any
4030 message, but which match the given pattern. Implemented by a patch from
4031 Martin Sluka, which also tidied up the Perl a bit.
4033 71. Recipient callout verification, like sender verification, was using <> in
4034 the MAIL FROM command. This isn't really the right thing, since the actual
4035 sender may affect whether the remote host accepts the recipient or not. I
4036 have changed it to use the actual sender in the callout; this means that
4037 the cache record is now keyed on a recipient/sender pair, not just the
4038 recipient address. There doesn't seem to be a real danger of callout loops,
4039 since a callout by the remote host to check the sender would use <>.
4040 [SEE ABOVE: changed after hitting problems.]
4042 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4043 temporary errors. However, in the case of such a code being given after
4044 the end of a data transmission (i.e. after ".") Exim was failing to write
4045 a retry record for the message. (Yes, there was some broken host that was
4046 actually sending 8xx at this point.)
4048 73. An unknown lookup type in a host list could cause Exim to panic-die when
4049 the list was checked. (An example that provoked this was putting <; in the
4050 middle of a list instead of at the start.) If this happened during a DATA
4051 ACL check, a -D file could be left lying around. This kind of configuration
4052 error no longer causes Exim to die; instead it causes a defer errror. The
4053 incident is still logged to the main and panic logs.
4055 74. Buglet left over from Exim 3 conversion. The message "too many messages
4056 in one connection" was written to the rejectlog but not the mainlog, except
4057 when address rewriting (yes!) was being logged.
4059 75. Added write_rejectlog option.
4061 76. When a system filter was run not as root (that is, when system_filter_user
4062 was set), the values of the $n variables were not being returned to the
4063 main process; thus, they were not subsequently available in the $sn
4066 77. Added +return_path_on_delivery log selector.
4068 78. A connection timeout was being treated differently from recipients deferred
4069 when testing hosts_max_try with a message that was older than the host's
4070 retry timeout. (The host should not be counted, thus allowing all hosts to
4071 be tried at least once before bouncing.) This may have been the cause of an
4072 occasionally reported bug whereby a message would remain on the queue
4073 longer than the retry timeout, but would be bounced if a delivery was
4074 forced. I say "may" because I never totally pinned down the problem;
4075 setting up timeout/retry tests is difficult. See also the next item.
4077 79. The ultimate address timeout was not being applied to errors that involved
4078 a combination of host plus message (for example, a timeout on a MAIL
4079 command). When an address resolved to a number of possible hosts, and they
4080 were not all tried for each delivery (e.g. because of hosts_max_try), a
4081 message could remain on the queue longer than the retry timeout.
4083 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4086 81. Fixed an obscure SMTP outgoing bug which required at least the following
4087 conditions: (a) there was another message waiting for the same server;
4088 (b) the server returned 5xx to all RCPT commands in the first message so
4089 that the message was not completed; (c) the server dropped the connection
4090 or gave a negative response to the RSET that Exim sends to abort the
4091 transaction. The observed case was a dropped connection after DATA that had
4092 been sent in pipelining mode. That is, the server had advertised PIPELINING
4093 but was not implementing it correctly. The effect of the bug was incorrect
4094 behaviour, such as trying another host, and this could lead to a crash.
4100 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4101 and daemon.c were passed as pointers to ints; they should have been
4102 pointers to socklen_t variables (which are typically unsigned ints).
4104 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4107 3. Fixed a really odd bug that affected only the testing scheme; patching a
4108 certain fixed string in the binary changed the value of another string that
4109 happened to be identical to the end of the original first string.
4111 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4112 name", it returns that address as the IP address. On some operating
4113 systems (e.g. Solaris), it also passes back the IP address string as the
4114 "host name". However, on others (e.g. Linux), it passes back an empty
4115 string. Exim wasn't checking for this, and was changing the host name to an
4116 empty string, assuming it had been canonicized.
4118 5. Although rare, it is permitted to have more than one PTR record for a given
4119 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4120 all the names associated with an address, because they do in Solaris.
4121 However, it seems that they do not in Linux for data that comes from the
4122 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4123 I found this out when I moved to a new Linux workstation and tried to run
4124 the Exim test suite.
4126 To get round this problem I have changed the code so that it now does its
4127 own call to the DNS to look up PTR records when searching for a host name.
4128 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4129 addresses that are only in /etc/hosts are still found.
4131 This behaviour is, however, controlled by an option called host_lookup_
4132 order, which defaults to "bydns:byaddr". If people want to use the other
4133 order, or indeed, just use one or the other means of lookup, they can
4134 specify it in this variable.
4136 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4137 some operating systems, this comes back from gethostbyaddr() as an empty
4138 string, and this is what Exim used to test for. However, it seems that in
4139 other systems, "." is yielded. Exim now tests for this case too.
4141 7. The values of check_spool_space and check_log_space are now held internally
4142 as a number of kilobytes instead of an absolute number of bytes. If a
4143 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4144 kilobyte. This means that much larger values can be stored.
4146 8. Exim monitor: an attempt to get the action menu when not actually pointing
4147 at a message produces an empty menu entitled "No message selected". This
4148 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4149 no entries in it ("Shell widget menu has zero width and/or height"). So I
4150 have added a single, blank menu entry in this case.
4152 9. Added ${quote_local_part.
4154 10. MIME decoding is now applied to the contents of Subject: header lines when
4157 11. Now that a reference to $sender_host_address automatically causes a reverse
4158 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4159 host lookup before query-style lookups in lists that might use this
4160 variable. This has therefore been abolished, and the "net-" prefix is no
4161 longer necessary for query-style lookups.
4163 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4164 have been a typo for LFLAGS, so it has been changed.
4166 13. The install script calls Exim with "-C /dev/null" in order to find the
4167 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4168 to be output. Howeve, since Exim outputs its version number before the
4169 error, it didn't break the script. It just looked ugly. I fixed this by
4170 always allowing "-C /dev/null" if the caller is root.
4172 14. Ignore overlarge ACL variable number when reading spool file - insurance
4173 against a later release with more variables having written the file.
4175 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4176 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4177 this, as well as the form without the "IPv6" on the front.
4179 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4180 OS/Makefile-Darwin file.
4182 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4183 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4185 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4186 variable. However, some OS use ints for inodes. I've added cast to long int
4187 to get rid of the compiler warning.
4189 19. I had forgotten to lock out "/../" in configuration file names when
4190 ALT_CONFIG_PREFIX was set.
4192 20. Routers used for verification do not need to specify transports. However,
4193 if such a router generated a host list, and callout was configured, Exim
4194 crashed, because it could not find a port number from the (non-existent)
4195 transport. It now assumes port 25 in this circumstance.
4197 21. Added the -t option to exigrep.
4199 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4200 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4201 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4202 them is set, LOOKUP_LSEARCH is forced.
4204 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4206 24. Added sender and host information to the "rejected by local_scan()" log
4207 line; previously there was no indication of these.
4209 25. Added .include_if_exists.
4211 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4212 Linux. It turns out that not all file systems support this. Apparently some
4213 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4214 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4215 not supported on the file descriptor, is now ignored when Exim is trying to
4216 sync a directory. This applies only to Linux.
4218 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4220 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4221 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4224 29. Added support for maildirsize files from supplied patch (modified a bit).
4226 30. The use of :fail: followed by an empty string could lead Exim to respond to
4227 sender verification failures with (e.g.):
4229 550 Verification failed for <xxx>
4230 550 Sender verify failed
4232 where the first response line was missing the '-' that indicates it is not
4233 the final line of the response.
4235 31. The loop for finding the name of the user that called Exim had a hardwired
4236 limit of 10; it now uses the value of finduser_retries, which is used for
4237 all other user lookups.
4239 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4242 33. Exim was neglecting to zero errno before one call of strtol() when
4243 expanding a string and expecting an integer value. On some systems this
4244 resulted in spurious "integer overflow" errors. Also, it was casting the
4245 result into an int without checking.
4247 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4248 did not work. The code looks as if it has *never* worked, though it appears
4249 to have been documented since at least releast 1.62. I have made it work.
4251 35. The "timeout_DNS" error in retry rules, also documented since at least
4252 1.62, also never worked. As it isn't clear exactly what this means, and
4253 clearly it isn't a major issue, I have abolished the feature by treating it
4254 as "timeout", and writing a warning to the main and panic logs.
4256 36. The display of retry rules for -brt wasn't always showing the error code
4259 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4260 timeout_connect_A, timeout_connect_MX.
4262 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4263 to the empty sender.
4265 39. The daemon was not analysing the content of -oX till after it had closed
4266 stderr and disconnected from the controlling terminal. This meant that any
4267 syntax errors were only noted on the panic log, and the return code from
4268 the command was 0. By re-arranging the code a little, I've made the
4269 decoding happen first, so such errors now appear on stderr, and the return
4270 code is 1. However, the actual setting up of the sockets still happens in
4271 the disconnected process, so errors there are still only recorded on the
4274 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4275 connections (as happens on some IP stacks) was logged at start up time as
4276 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4277 it from "IPv6 and IPv4", which means that two separate sockets are being
4280 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4281 says whether AF_INET or AF_INET6 was passed as an argument.
4283 42. Exiwhat output was messed up when time zones were included in log
4286 43. Exiwhat now gives more information about the daemon's listening ports,
4287 and whether -tls-on-connect was used.
4289 44. The "port" option of the smtp transport is now expanded.
4291 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4292 silently ignored. Now an error message is written to the main and panic
4295 46. There's a new ACL modifier called "logwrite" which writes to a log file
4296 as soon as it is encountered.
4298 47. Added $local_user_uid and $local_user_gid at routing time.
4300 48. Exim crashed when trying to verify a sender address that was being
4303 49. Exim was recognizing only a space character after ".include". It now also
4304 recognizes a tab character.
4306 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4307 extracting the relevant information from the specification. The man page no
4308 longer contains scrambled data for the -d option, and I've added a section
4309 at the front about calling Exim under different names.
4311 51. Added "extra_headers" argument to the "mail" command in filter files.
4313 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4316 53. Installed eximstats 1.29.
4318 54. Added transport_filter_timeout as a generic transport option.
4320 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4321 Cc: header lines. This was required by RFC 822, but it not required by RFC
4324 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4325 incoming messages that did not have them. Now it does so only if the
4326 message originates locally, that is, if there is no associated remote host
4327 address. When Resent- header lines are present, this applies to the Resent-
4328 lines rather than the non-Resent- lines.
4330 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4331 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4333 58. Messages for configuration errors now include the name of the main
4334 configuration file - useful now that there may be more than one file in a
4335 list (.included file names were always shown).
4337 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4338 for those rare installations that do not start the daemon as root or run it
4339 setuid root. I've cut out the call to initgroups() if the daemon is not
4342 60. The Exim user and group can now be bound into the binary as text strings
4343 that are looked up at the start of Exim's processing.
4345 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4347 62. Added $mailstore_basename variable.
4349 63. Installed patch to sieve.c from Michael Haardt.
4351 64. When Exim failed to open the panic log after failing to open the main log,
4352 the original message it was trying to log was written to stderr and debug
4353 output, but if they were not available (the usual case in production), it
4354 was lost. Now it is written to syslog before the two lines that record the
4355 failures to open the logs.
4357 65. Users' Exim filters run in subprocesses under the user's uid. It is
4358 possible for a "deliver" command or an alias in a "personal" command to
4359 provoke an address rewrite. If logging of address rewriting is configured,
4360 this fails because the process is not running as root or exim. There may be
4361 a better way of dealing with this, but for the moment (because 4.30 needs
4362 to be released), I have disabled address rewrite logging when running a
4363 filter in a non-root, non-exim process.
4369 1. The buildconfig auxiliary program wasn't quoting the value set for
4370 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4371 not defined. This bug was masked in 4.22 by the effect that was fixed in
4374 2. Some messages that were rejected after a message id was allocated were
4375 shown as "incomplete" by exigrep. It no longer does this for messages that
4376 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4378 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4379 have allow_domain_literals set, the ID did not get logged in the <= line.
4380 Domain literals are now always recognized in Message-ID: header lines.
4382 4. The first argument for a ${extract expansion item is the key name or field
4383 number. Leading and trailing spaces in this item were not being ignored,
4384 causing some misleading effects.
4386 5. When deliver_drop_privilege was set, single queue runner processes started
4387 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4388 same command in the process it spins off) were not dropping privilege.
4390 6. When the daemon running as "exim" started a queue runner, it always
4391 re-executed Exim in the spun-off process. This is a waste of effort when
4392 deliver_drop_privilege is set. The new process now just calls the
4393 queue-runner function directly.
4399 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4402 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4403 ignored. Though the use of -oP was forcing the writing of a pid file, it
4404 was always written to the default place.
4406 3. If the message "no IP address found for host xxxx" is generated during
4407 incoming verification, it is now followed by identification of the incoming
4408 connection (so you can more easily find what provoked it).
4410 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4412 5. Added some features to "harden" Exim a bit more against certain attacks:
4414 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4415 be put in Local/Makefile. This is like the never_users runtime option,
4416 but it cannot be overridden. The default setting is "root".
4418 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4419 prefix string with which any file named in a -C command line option
4422 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4423 is retained for -C and -D only if the caller of Exim is root. Without
4424 it, the exim user may also use -C and -D and retain privilege.
4426 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4427 command line option is disabled.
4429 6. Macro names set by the -D option must start with an upper case letter, just
4430 like macro names defined in the configuration file.
4432 7. Added "dereference=" facility to LDAP.
4434 8. Two instances of the typo "uknown" in the source files are fixed.
4436 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4437 the Configure-Makefile script screwed up while processing it.
4439 10. Incorporated PCRE 4.4.
4441 11. The SMTP synchronization check was not operating right at the start of an
4442 SMTP session. For example, it could not catch a HELO sent before the client
4443 waited for the greeting. There is now a check for outstanding input at the
4444 point when the greeting is written. Because of the duplex, asynchronous
4445 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4446 way, but not yet received, when the check is performed.
4448 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4449 on TCP/IP sockets, because this apparently causes some broken clients to
4452 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4453 unchanged) from the Cygwin maintainer.
4455 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4456 was supported instead of testing for mbx. Effectively a typo.
4458 15. The spa authenticator server code was not checking that the input it
4459 received was valid base64.
4461 16. The debug output line for the "set" modifier in ACLs was not showing the
4462 name of the variable that was being set.
4464 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4466 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4467 lookup on its own. Something else was needed to trigger the lookup. For
4468 example, a match in host_lookup or the need for a host name in a host list.
4469 Now, if $sender_host_name is referenced and the host name has not yet been
4470 looked up, a lookup is performed. If the lookup fails, the variable remains
4471 empty, and $host_lookup_failed is set to "1".
4473 19. Added "eqi" as a case-independent comparison operator.
4475 20. The saslauthd authentication condition could segfault if neither service
4476 nor realm was specified.
4478 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4479 error message that was logged was misleading, and incoming SMTP
4480 connections were dropped. The message is now more accurate, and temporary
4481 errors are given to SMTP connections.
4483 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4484 (except RSET) with a 421 error, until QUIT is received. However, it was
4485 failing to send a response to QUIT.
4487 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4488 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4489 ACL is now run after the helo_try_verify_hosts code.
4491 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4492 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4493 case-independent, so other case variants are also recognized). Apparently
4494 some systems use these upper case variants.
4496 25. If more than two messages were waiting for the same host, and a transport
4497 filter was specified for the transport, Exim sent two messages over the
4498 same TCP/IP connection, and then failed with "socket operation on non-
4499 socket" when it tried to send the third.
4501 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4503 27. The extern definition of crypt16() in expand.c was not being excluded when
4504 the OS had its own crypt16() function.
4506 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4507 as a preferred synonym for return_size_limit, both as an option and as an
4510 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4512 30. Changed the default configuration ACL to relax the local part checking rule
4513 for addresses that are not in any local domains. For these addresses,
4514 slashes and pipe symbols are allowed within local parts, but the sequence
4515 /../ is explicitly forbidden.
4517 31. SPA server authentication was not clearing the challenge buffer before
4520 32. log_message in a "warn" ACL statement was writing to the reject log as
4521 well as to the main log, which contradicts the documentation and doesn't
4522 seem right (because no rejection is happening). So I have stopped it.
4524 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4525 However, I am unable to do any testing of this.
4527 34. Fixed an infelicity in the appendfile transport. When checking directories
4528 for a mailbox, to see if any needed to be created, it was accidentally
4529 using path names with one or more superfluous leading slashes; tracing
4530 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4532 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4533 contents are added to the log line that is written for every discarded
4534 recipient. (Previously a log_message setting was ignored.)
4536 36. The ${quote: operator now quotes the string if it is empty.
4538 37. The install script runs exim in order to find its version number. If for
4539 some reason other than non-existence or emptiness, which it checks, it
4540 could not run './exim', it was installing it with an empty version number,
4541 i.e. as "exim-". This error state is now caught, and the installation is
4544 38. An argument was missing from the function that creates an error message
4545 when Exim fails to connect to the socket for saslauthd authentication.
4546 This could cause Exim to crash, or give a corrupted message.
4548 39. Added isip, isip4, and isip6 to ${if conditions.
4550 40. The ACL variables $acl_xx are now saved with the message, and can be
4551 accessed later in routers, transports, and filters.
4553 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4554 strings in the file are not string-expanded.
4556 42. If a MAIL command specified a SIZE value that was too large to fit into an
4557 int variable, the check against message_size_limit failed. Such values are
4558 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4559 day this will have to be increased, but I don't think I want to be around
4560 when emails are that large.
4567 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4568 iconv() is not standard in FreeBSD.
4570 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4571 IPv6 enabled. The observed symptom was a segmentation fault on return from
4572 the function os_common_find_running_interfaces() in src/os.c.
4574 3. In the check_special_case() function in daemon.c I had used "errno" as an
4575 argument name, which causes warnings on some systems. This was basically a
4576 typo, since it was named "eno" in the comments!
4578 4. The code that waits for the clock to tick (at a resolution of some fraction
4579 of a second) so as to ensure message-id uniqueness was always waiting for
4580 at least one whole tick, when it could have waited for less. [This is
4581 almost certainly not relevant at current processor speeds, where it is
4582 unlikely to ever wait at all. But we try to future-proof.]
4584 5. The function that sleeps for a time interval that includes fractions of a
4585 second contained a race. It did not block SIGALRM between setting the
4586 timer, and suspending (a couple of lines later). If the interval was short
4587 and the sigsuspend() was delayed until after it had expired, the suspension
4588 never ended. On busy systems this could lead to processes getting stuck for
4591 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4592 process, before it forks any delivery processes. The open lookup caching
4593 mechanism meant that the open file or database connection was passed into
4594 the delivery process. The problem was that delivery processes always tidy
4595 up cached lookup data. This could cause a problem for the next delivery
4596 process started by the queue runner, because the external queue runner
4597 process does not know about the closure. So the next delivery process
4598 still has data in the lookup cache. In the case of a file lookup, there was
4599 no problem because closing a file descriptor in a subprocess doesn't affect
4600 the parent. However, if the lookup was caching a connection to a database,
4601 the connection was closed, and the second delivery process was likely to
4602 see errors such as "PGSQL: query failed: server closed the connection
4603 unexpectedly". The problem has been fixed by closing all cached lookups
4604 in a queue runner before running a delivery process.
4606 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4607 seem to have the "const" qualifier which it has on other OS. I've
4610 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4611 *of the same type* (client or server) with the same public name that an
4612 error should be diagnosed.
4614 9. When Exim looked up a host name for an IP address, but failed to find the
4615 original IP address when looking up the host name (a safety check), it
4616 output the message "<ip address> does not match any IP for NULL", which was
4617 confusing, to say the least. The bug was that the host name should have
4618 appeared instead of "NULL".
4620 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4621 user that is built into the binary, and the -C or -D options is used, root
4622 privilege is dropped before the configuration file is read. In addition,
4623 logging is switched to stderr instead of the normal log files. If the
4624 configuration then re-defines the Exim user, the unprivileged environment
4625 is probably not what is expected, so Exim logs a panic warning message (but
4628 However, if deliver_drop_privilege is set, the unprivileged state may well
4629 be exactly what is intended, so the warning has been cut out in that case,
4630 and Exim is allowed to try to write to its normal log files.
4636 1. smtp_return_error_details was not giving details for temporary sender
4637 or receiver verification errors.
4639 2. Diagnose a configuration error if two authenticators have the same public
4642 3. Exim used not to create the message log file for a message until the first
4643 delivery attempt. This could be confusing when incoming messages were held
4644 for policy or load reasons. The message log file is now created at the time
4645 the message is received, and an initial "Received" line is written to it.
4647 4. The automatically generated man page for command line options had a minor
4648 bug that caused no ill effects; however, a more serious problem was that
4649 the procedure for building the man page automatically didn't always
4650 operate. Consequently, release 4.20 contains an out-of-date version. This
4651 shouldn't happen again.
4653 5. When building Exim with embedded Perl support, the script that builds the
4654 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4655 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4657 6. The freeze_tell option was not being used for messages that were frozen on
4658 arrival, either by an ACL or by local_scan().
4660 7. Added the smtp_incomplete_transaction log selector.
4662 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4663 was accepting AUTH without a new EHLO.
4665 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4666 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4669 10. Logging of TCP/IP connections (when configured) now happens in the main
4670 daemon process instead of the child process, so that the TCP/IP connection
4671 count is more accurate (but it can never be perfect).
4673 11. The use of "drop" in a nested ACL was not being handled correctly in the
4674 outer ACL. Now, if condition failure induced by the nested "drop" causes
4675 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4676 or "require"), the connection is dropped.
4678 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4679 that yield "discard" can now be used with an "accept" or a "discard" verb,
4680 but an error is generated for any others (because I can't see a useful way
4681 to define what should happen).
4683 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4684 are now processed in the same way as lines in the Exim configuration file.
4685 In particular, continuation lines are supported.
4687 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4689 15. Added -ti meaning -t -i.
4691 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4692 domains, and warn by logging if others are found.
4694 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4695 returned from GIFCONF when Exim is trying to find the list of interfaces on
4696 a host. The code in os.c has been modified to copy each ifreq to an aligned
4697 structure in all cases.
4699 Also, in some cases, the returned ifreq's were being copied to a 'struct
4700 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4701 means the last couple of bytes of an IPv6 address could be chopped if the
4702 ifreq contained only a normal sockaddr (14 bytes storage).
4704 18. Named domain lists were not supported in the hosts_treat_as_local option.
4705 An entry such as +xxxx was not recognized, and was treated as a literal
4708 19. Ensure that header lines added by a DATA ACL are included in the reject log
4709 if the ACL subsequently rejects the message.
4711 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4712 MD5 (which is deprecated).
4714 21. When testing a filter file using -bf, Exim was writing a message when it
4715 took the sender from a "From " line in the message, but it was not doing so
4716 when it took $return_path from a Return-Path: header line. It now does.
4718 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4719 with a valid header line field name (a series of printing characters
4720 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4722 23. Changed "disc" in the source to "disk" to conform to the documentation and
4723 the book and for uniformity.
4725 24. Ignore Sendmail's -Ooption=value command line item.
4727 25. When execve() failed while trying to run a command in a pipe transport,
4728 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4729 could be confused with a return value of 69 from the command itself. This
4730 has been changed to 127, the value the shell returns if it is asked to run
4731 a non-existent command. The wording for the related log line suggests a
4732 non-existent command as the problem.
4734 26. If received_header_text expands to an empty string, do not add a Received:
4735 header line to the message. (Well, it adds a token one on the spool, but
4736 marks it "old" so that it doesn't get used or transmitted.)
4738 27. Installed eximstats 1.28 (addition of -nt option).
4740 28. There was no check for failure on the call to getsockname() in the daemon
4741 code. This can fail if there is a shortage of resources on the system, with
4742 ENOMEM, for example. A temporary error is now given on failure.
4744 29. Contrary to the C standard, it seems that in some environments, the
4745 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4746 program. Exim now does this explicitly; it affects the formatting of
4747 timestamps using strftime().
4749 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4750 complaints. I've now initialized all the variables, to avoid this.
4752 32. Header lines added by a system filter were not being "seen" during
4753 transport-time rewrites.
4755 33. The info_callback() function passed to OpenSSL is set up with type void
4756 (*)(SSL *, int, int), as described somewhere. However, when calling the
4757 function (actually a macro) that sets it up, the type void(*)() is
4758 expected. I've put in a cast to prevent warnings from picky compilers.
4760 34. If a DNS black list lookup found a CNAME record, but there were no A
4761 records associated with the domain it pointed at, Exim crashed.
4763 35. If a DNS black list lookup returned more than one A record, Exim ignored
4764 all but the first. It now scans all returned addresses if a particular IP
4765 value is being sought. In this situation, the contents of the
4766 $dnslist_value variable are a list of all the addresses, separated by a
4769 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4770 to accept a host name and all its aliases if the forward lookup for any of
4771 them yielded the IP address of the incoming connection. Now it accepts only
4772 those names whose forward lookup yields the correct IP address. Any other
4773 names are discarded. This closes a loophole whereby a rogue DNS
4774 administrator could create reverse DNS records to break through a
4775 wildcarded host restriction in an ACL.
4777 37. If a user filter or a system filter that ran in a subprocess used any of
4778 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4779 the wrong values were passed to the pipe command ($thisaddress had the
4780 value of $0, $0 had the value of $1, etc). This bug was introduced by
4781 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4783 38. Improved the line breaking for long SMTP error messages from ACLs.
4784 Previously, if there was no break point between 40 and 75 characters, Exim
4785 left the rest of the message alone. Two changes have been made: (a) I've
4786 reduced the minimum length to 35 characters; (b) if it can't find a break
4787 point between 35 and 75 characters, it looks ahead and uses the first one
4788 that it finds. This may give the occasional overlong line, but at least the
4789 remaining text gets split now.
4791 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4792 file descriptors might be low, and that setting 1000 would always raise it.
4793 It turns out that in some environments, the limit is already over 1000 and
4794 that lowering it causes trouble. So now Exim takes care not to decrease it.
4796 40. When delivering a message, the value of $return_path is set to $sender_
4797 address at the start of routing (routers may change the value). By an
4798 oversight, this default was not being set up when an address was tested by
4799 -bt or -bv, which affected the outcome if any router or filter referred to
4802 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4803 log only when "message" or "log_message" are set. However, if one of the
4804 conditions was an address verification, or a call to a nested ACL, the
4805 messages generated by the underlying test were being passed through. This
4806 no longer happens. The underlying message is available in $acl_verify_
4807 message for both "message" and "log_message" expansions, so it can be
4808 passed through if needed.
4810 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4811 new expansion $bh_ to give the encoded byte string without charset
4812 translation. Translation happens only if iconv() is available; HAVE_ICONV
4813 indicates this at build time. HEADERS_CHARSET gives the charset to
4814 translate to; headers_charset can change it in the configuration, and
4815 "headers charset" can change it in an individual filter file.
4817 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4818 that creates RFC 2047 encoded "words" labels them as that charset instead
4819 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4820 expansion operator; (ii) when Exim creates a From: line for a local
4821 message; (iii) when a header line is rewritten to include a "phrase" part.
4823 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4824 buggy, causing it to skip the first lines of messages whose message ID
4825 ended in 'D'. This would not have bitten before Exim release 4.14, because
4826 message IDs were unlikely to end in 'D' before then. The effect was to have
4827 incorrect size information for certain domains.
4829 45. #include "config.h" was missing at the start of the crypt16.c module. This
4830 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4833 46. If there was a timeout during a "random" callout check, Exim treated it as
4834 a failure of the random address, and carried on sending RSET and the real
4835 address. If the delay was just some slowness somewhere, the response to the
4836 original RCPT would be taken as a response to RSET and so on, causing
4837 mayhem of various kinds.
4839 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4840 when I implemented it. It didn't allow for the fact that some option values
4841 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4842 the right test for positive values.
4844 48. Domain names in DNS records are case-independent. Exim always looks them up
4845 in lower case. Some resolvers return domain names in exactly the case they
4846 appear in the zone file, that is, they may contain uppercase letters. Not
4847 all resolvers do this - some return always lower case. Exim was treating a
4848 change of case by a resolver as a change of domain, similar to a widening
4849 of a domain abbreviation. This triggered its re-routing code and so it was
4850 trying to route what was effectively the same domain again. This normally
4851 caused routing to fail (because the router wouldn't handle the domain
4852 twice). Now Exim checks for this case specially, and just changes the
4853 casing of the domain that it ultimately uses when it transmits the message
4856 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4859 50. If a filter generated a file delivery with a non-absolute name (possible if
4860 no home directory exists for the router), the forbid_file option was not
4863 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4864 the existing equality matching.
4866 52. Exim was using ints instead of ino_t variables in some places where it was
4867 dealing with inode numbers.
4869 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4870 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4871 called TMPDIR, and if it finds it is different, it changes its value.
4873 54. The smtp_printf() function is now made available to local_scan() so
4874 additional output lines can be written before returning. There is also an
4875 smtp_fflush() function to enable the detection of a dropped connection.
4876 The variables smtp_input and smtp_batched_input are exported to
4879 55. Changed the default runtime configuration: the message "Unknown user"
4880 has been removed from the ACL, and instead placed on the localuser router,
4881 using the cannot_route_message feature. This means that any verification
4882 failures that generate their own messages won't get overridden. Similarly,
4883 the "Unrouteable address" message that was in the ACL for unverifiable
4884 relay addresses has also been removed.
4886 56. Added hosts_avoid_esmtp to the smtp transport.
4888 57. The exicyclog script was not checking for the esoteric option
4889 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4890 will work only if exicyclog is run under the appropriate euid.
4892 58. Following a discussion on the list, the rules by which Exim recognises line
4893 endings on incoming messages have been changed. The -dropcr and drop_cr
4894 options are now no-ops, retained only for backwards compatibility. The
4895 following line terminators are recognized: LF CRLF CR. However, special
4896 processing applies to CR:
4898 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4899 nor a local message in the state where . is a terminator.
4901 (ii) If a bare CR is encountered in a header line, an extra space is added
4902 after the line terminator so as not to end the header. The reasoning
4903 behind this is that bare CRs in header lines are most likely either
4904 to be mistakes, or people trying to play silly games.
4906 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4907 was being incorrectly given as 18 bytes larger than it should have been.
4908 This is a VOB (very old bug).
4910 60. This may never have affected anything current, but just in case it has:
4911 When the local host is found other than at the start of a list of hosts,
4912 the local host, those with the same MX, and any that follow, are discarded.
4913 When the list in question was part of a longer list of hosts, the following
4914 hosts (not currently being processed) were also being discarded. This no
4915 longer happens. I'm not sure if this situation could ever has previously
4918 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4921 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4922 header. This is something that is recommended in a new Internet Draft, and
4923 is something that is documented as being done by Sendmail. There are two
4924 possible values. For messages generated by the autoreply transport, Exim
4927 Auto-Submitted: auto-replied
4929 whereas for all other generated messages (e.g. bounces) it adds
4931 Auto-Submitted: auto-generated
4933 63. The "personal" condition in filters now includes a test for the
4934 Auto-Submitted: header. If it contains the string "auto-" the message it
4935 not considered personal.
4937 64. Added rcpt_include_affixes as a generic transport option.
4939 65. Added queue_only_override (default true).
4941 66. Added the syslog_duplication option.
4943 67. If what should have been the first header line of a message consisted of
4944 a space followed by a colon, Exim was mis-interpreting it as a header line.
4945 It isn't of course - it is syntactically invalid and should therefore be
4946 treated as the start of the message body. The misbehaviour could have
4947 caused a number of strange effects, including loss of data in subsequent
4948 header lines, and spool format errors.
4950 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4951 client host had authenticated. This control can now be exercised by an ACL
4952 for more flexibility.
4954 69. By default, callouts do not happen when testing with -bh. There is now a
4955 variant, -bhc, which does actually run the callout code, including
4956 consulting and updating the callout cache.
4958 70. Added support for saslauthd authentication, courtesy of Alexander
4961 71. If statvfs() failed on the spool or log directories while checking their
4962 size for availability, Exim confusingly gave the error "space shortage".
4963 Furthermore, in debugging mode it crashed with a floating point exception.
4964 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4965 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4966 serious problem, Exim now writes to the main and panic logs when this
4967 happens, with details of the failure. It then refuses to accept the
4968 incoming message, giving the message "spool directory problem" or "log
4969 directory problem" with a 421 code for SMTP messages.
4971 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4972 0, 1, and 2 exist, because some OS complain for execs without them (see
4973 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4974 descriptors. However, the code omitted to check that the open succeeded,
4975 causing mysterious errors if for some reason the permissions on /dev/null
4976 got screwed. Now Exim writes a message to the main and panic logs, and
4977 bombs out if it can't open /dev/null.
4979 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4980 interact so that it is all more flexible. It is supposed to remain
4981 backwards compatible. Also added extra_local_interfaces.
4983 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4984 to bomb out with an assertion failure - to the client this appears as a
4985 connection drop. This problem occurs in the part of the code that was taken
4986 from the Samba project. Fortunately, the assertion is in a very simple
4987 function, so I have fixed this by reproducing the function inline in the
4988 one place where it is called, and arranging for authentication to fail
4989 instead of killing the process with assert().
4991 75. The SPA client code was not working when the server requested OEM rather
4992 than Unicode encoding.
4994 76. Added code to make require_files with a specific uid setting more usable in
4995 the case where statting the file as root fails - usually a non-root-mounted
4996 NFS file system. When this happens and the failure is EACCES, Exim now
4997 forks a subprocess and does the per-uid checking as the relevant uid.
4999 77. Added process_log_path.
5001 78. If log_file_path was not explicitly set, a setting of check_log_space or
5002 check_log_inodes was ignored.
5004 79. If a space check for the spool or log partitions fails, the incident is now
5005 logged. Of course, in the latter case the data may get lost...
5007 80. Added the %p formatting code to string_format() so that it can be used to
5008 print addresses in debug_print(). Adjusted all the address printing in the
5009 debugging in store.c to use %p rather than %d.
5011 81. There was a concern that a line of code in smtp_in.c could overflow a
5012 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5013 initially expressed, the concern was not well-founded, because trailing
5014 spaces are removed early. However, if the trailing spaces were followed by
5015 a NULL, they did not get removed, so the overflow was possible. Two fixes
5018 (a) I re-wrote the offending code in a cleaner fashion.
5019 (b) If an incoming SMTP command contains a NULL character, it is rejected
5022 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5023 runs initgroups(), so that if the Exim user is in any additional groups,
5024 they will be used during message reception.
5030 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob