1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.597 2009/11/20 12:18:19 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.
13 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
15 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
21 TK/01 Added patch by Johannes Berg that expands the main option
22 "spamd_servers" if it starts with a dollar sign.
24 TK/02 Write list of recipients to X-Envelope-Sender header when building
25 the mbox-format spool file for content scanning (suggested by Jakob
28 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
29 (http://www.dcc-servers.net/) support via dccifd. Activated by
30 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
31 experimental_spec.txt for more documentation.
33 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
34 by Mark Daniel Reidel <mr@df.eu>.
36 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
37 When building exim an external PCRE library is now needed -
38 PCRE is a system library on the majority of modern systems.
39 See entry on PCRE_LIBS in EDITME file.
41 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
42 conversation. Added nologin parameter to request.
43 Patch contributed by Kirill Miazine
45 TF/01 Do not log submission mode rewrites if they do not change the address.
47 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
49 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
50 log files in place. Contributed by Roberto Lima
52 NM/04 Bugzilla 667: close socket used by dovecot authenticator
54 TF/03 Bugzilla 615: When checking the local_parts router precondition
55 after a local_part_suffix or local_part_prefix option, Exim now
56 does not use the address's named list lookup cache, since this
57 contains cached lookups for the whole local part.
59 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
60 Robert Millan. Documentation is in experimental-spec.txt
62 TF/04 Bugzilla 668: Fix parallel build (make -j).
64 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
66 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
67 Patch provided by Jan Srzednicki
69 TF/05 Leading white space used to be stripped from $spam_report which
70 wrecked the formatting. Now it is preserved.
72 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
73 that they are available at delivery time.
75 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
77 TF/08 TLS error reporting now respects the incoming_interface and
78 incoming_port log selectors.
80 TF/09 Produce a more useful error message if an SMTP transport's hosts
81 setting expands to an empty string.
83 NM/06 Bugzilla 744: EXPN did not work under TLS.
84 Patch provided by Phil Pennock
86 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
87 Patch provided by Richard Godbee
89 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
90 acl_smtp_notquit, added index entry.
92 NM/09 Bugzilla 787: Potential buffer overflow in string_format
93 Patch provided by Eugene Bujak
95 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
96 Patch provided by Maxim Dounin
98 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
99 Patch provided by Phil Pennock
101 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
103 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
104 Patch provided by Brad "anomie" Jorsch
106 NM/14 Bugzilla 622: Added timeout setting to transport filter.
107 Patch provided by Dean Brooks
109 TK/05 Add native DKIM support (does not depend on external libraries).
111 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
112 Patch provided by Graeme Fowler
114 NM/16 Bugzilla 851: Documentation example syntax fix.
116 NM/17 Changed NOTICE file to remove references to embedded PCRE.
118 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
120 NM/19 Bugzilla 745: TLS version reporting
121 Patch provided by Phil Pennock
123 NM/20 Bugzilla 167: bool: condition support
124 Patch provided by Phil Pennock
126 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
127 Patch provided by Phil Pennock
129 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
130 Patch provided by Brad "anomie" Jorsch
132 NM/23 Bugzilla 687: Fix misparses in eximstats
133 Patch provided by Heiko Schlittermann
135 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
136 Patch provided by Heiko Schlittermann
138 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
139 plus update to original patch.
141 NM/26 Bugzilla 799: Documentation correction for ratelimit
143 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
144 Patch provided by David Brownlee
146 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
148 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
150 NM/30 Bugzilla 888: TLS documentation bugfixes
152 NM/31 Bugzilla 896: Dovecot buffer overrun fix
154 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
155 Unlike the original bugzilla I have changed all shell scripts in src tree
157 NM/33 Bugzilla 898: Transport filter timeout fix
158 Patch by Todd Rinaldo
160 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
161 Patch by Serge Demonchaux
163 NM/35 Bugzilla 39: Base64 decode bug fixes
164 Patch by Jakob Hirsch
166 NM/36 Bugzilla 909: Correct connect() call in dcc code
168 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
170 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
172 NM/39 Bugzilla 911: Fixed MakeLinks build script
178 TK/01 Add preliminary DKIM support. Currently requires a forked version of
179 ALT-N's libdkim that I have put here:
180 http://duncanthrax.net/exim-experimental/
182 Note to Michael Haardt: I had to rename some vars in sieve.c. They
183 were called 'true' and it seems that C99 defines that as a reserved
184 keyword to be used with 'bool' variable types. That means you could
185 not include C99-style headers which use bools without triggering
186 build errors in sieve.c.
188 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
189 as mailq or other aliases. Changed the --help handling significantly
190 to do whats expected. exim_usage() emits usage/help information.
192 SC/01 Added the -bylocaldomain option to eximstats.
194 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
196 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
198 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
204 PH/01 Another patch from the Sieve maintainer.
206 PH/02 When an IPv6 address is converted to a string for single-key lookup
207 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
208 dots are used instead of colons so that keys in lsearch files need not
209 contain colons. This was done some time before quoting was made available
210 in lsearch files. However, iplsearch files do require colons in IPv6 keys
211 (notated using the quote facility) so as to distinguish them from IPv4
212 keys. This meant that lookups for IP addresses in host lists did not work
213 for iplsearch lookups.
215 This has been fixed by arranging for IPv6 addresses to be expressed with
216 colons if the lookup type is iplsearch. This is not incompatible, because
217 previously such lookups could never work.
219 The situation is now rather anomolous, since one *can* have colons in
220 ordinary lsearch keys. However, making the change in all cases is
221 incompatible and would probably break a number of configurations.
223 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
226 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
227 conversion specification without a maximum field width, thereby enabling
228 a rogue spamd server to cause a buffer overflow. While nobody in their
229 right mind would setup Exim to query an untrusted spamd server, an
230 attacker that gains access to a server running spamd could potentially
231 exploit this vulnerability to run arbitrary code as the Exim user.
233 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
234 $primary_hostname instead of what libspf2 thinks the hosts name is.
236 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
237 a directory entry by the name of the lookup key. Previously, if a
238 symlink pointed to a non-existing file or a file in a directory that
239 Exim lacked permissions to read, a lookup for a key matching that
240 symlink would fail. Now it is enough that a matching directory entry
241 exists, symlink or not. (Bugzilla 503.)
243 PH/03 The body_linecount and body_zerocount variables are now exported in the
246 PH/04 Added the $dnslist_matched variable.
248 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
249 This means they are set thereafter only if the connection becomes
252 PH/06 Added the client_condition to authenticators so that some can be skipped
253 by clients under certain conditions.
255 PH/07 The error message for a badly-placed control=no_multiline_responses left
256 "_responses" off the end of the name.
258 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
260 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
261 (without spaces) instead of just copying the configuration text.
263 PH/10 Added the /noupdate option to the ratelimit ACL condition.
265 PH/11 Added $max_received_linelength.
267 PH/12 Added +ignore_defer and +include_defer to host lists.
269 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
270 way in which PCRE > 7.0 is built.
272 PH/14 Implemented queue_only_load_latch.
274 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
275 MAIL command. The effect was to mangle the value on 64-bit systems.
277 PH/16 Another patch from the Sieve maintainer.
279 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
281 PH/18 If a system quota error occurred while trying to create the file for
282 a maildir delivery, the message "Mailbox is full" was not appended to the
283 bounce if the delivery eventually timed out. Change 4.67/27 below applied
284 only to a quota excession during the actual writing of the file.
286 PH/19 It seems that peer DN values may contain newlines (and other non-printing
287 characters?) which causes problems in log lines. The DN values are now
288 passed through string_printing() before being added to log lines.
290 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
291 and InterBase are left for another time.)
293 PH/21 Added message_body_newlines option.
295 PH/22 Guard against possible overflow in moan_check_errorcopy().
297 PH/23 POSIX allows open() to be a macro; guard against that.
299 PH/24 If the recipient of an error message contained an @ in the local part
300 (suitably quoted, of course), incorrect values were put in $domain and
301 $local_part during the evaluation of errors_copy.
307 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
308 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
311 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
312 issue a MAIL command.
314 PH/02 In an ACL statement such as
316 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
318 if a client was not listed at all, or was listed with a value other than
319 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
320 the condition was not true (as it should be), so access was not denied.
321 The bug was that the ! inversion was incorrectly passed on to the second
322 item. This has been fixed.
324 PH/03 Added additional dnslists conditions == and =& which are different from
325 = and & when the dns lookup returns more than one IP address.
327 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
328 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
330 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
331 FSYNC, which compiles an option called disable_fsync that allows for
332 bypassing fsync(). The documentation is heavily laced with warnings.
334 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
336 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
337 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
338 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
339 including adding "make clean"; (3) Added -fPIC when compiling the test
340 dynamically loaded module, to get rid of a warning.
342 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
343 message fails, move_frozen_messages = true and ignore_bounce_errors_after
344 = 0s. The bug is otherwise harmless.
346 PH/07 There was a bug in the dovecot authenticator such that the value of
347 $auth1 could be overwritten, and so not correctly preserved, after a
348 successful authentication. This usually meant that the value preserved by
349 the server_setid option was incorrect.
351 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
353 PH/09 Installed PCRE release 7.0.
355 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
356 run for batched SMTP input. It is now run at the start of every message
357 in the batch. While fixing this I discovered that the process information
358 (output by running exiwhat) was not always getting set for -bs and -bS
359 input. This is fixed, and it now also says "batched" for BSMTP.
361 PH/11 Added control=no_pipelining.
363 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
364 patch, slightly modified), and move the expansion of helo_data till after
365 the connection is made in the smtp transport (so it can use these
368 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
370 PH/14 Added log_selector = +pid.
372 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
374 PH/16 Add ${if forany and ${if forall.
376 PH/17 Added dsn_from option to vary the From: line in DSNs.
378 PH/18 Flush SMTP output before performing a callout, unless control =
379 no_callout_flush is set.
381 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
382 was true (the default) a successful delivery failed to delete the retry
383 item, thus causing premature timeout of the address. The bug is now
386 PH/20 Added hosts_avoid_pipelining to the smtp transport.
388 PH/21 Long custom messages for fakedefer and fakereject are now split up
389 into multiline reponses in the same way that messages for "deny" and
390 other ACL rejections are.
392 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
393 with slight modification.
395 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
396 draft, changing the syntax and factoring some duplicate code".
398 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
399 for deliveries of the second and subsequent messages over the same SMTP
402 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
403 ${reduce, with only minor "tidies".
405 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
407 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
408 expansion side effects.
410 PH/27 When a message times out after an over-quota error from an Exim-imposed
411 quota, the bounce message says "mailbox is full". This message was not
412 being given when it was a system quota that was exceeded. It now should
415 MH/03 Made $recipients available in local_scan(). local_scan() already has
416 better access to the recipient list through recipients_list[], but
417 $recipients can be useful in postmaster-provided expansion strings.
419 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
420 in the case of a MAIL command with additional options following the
421 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
422 were accidentally chopped off.
424 PH/29 SMTP synchronization checks are implemented when a command is read -
425 there is a check that no more input is waiting when there shouldn't be
426 any. However, for some commands, a delay in an ACL can mean that it is
427 some time before the response is written. In this time, more input might
428 arrive, invalidly. So now there are extra checks after an ACL has run for
429 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
430 pipelining has not been advertised.
432 PH/30 MH's patch to allow iscntrl() characters to be list separators.
434 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
435 returned in the SMTP response when smtp_return_error_details was false.
438 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
439 instead of the C I/O that was originally supplied, because problems were
442 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
443 Exim which did not show up earlier: it was assuming that a call to
444 SSL_CTX_set_info_callback() might give an error value. In fact, there is
445 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
446 was a macro that became an assignment, so it seemed to work. This has
447 changed to a proper function call with a void return, hence the compile
448 error. Exim's code has been fixed.
450 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
453 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
455 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
457 PH/37 If a message is not accepted after it has had an id assigned (e.g.
458 because it turns out to be too big or there is a timeout) there is no
459 "Completed" line in the log. When some messages of this type were
460 selected by exigrep, they were listed as "not completed". Others were
461 picked up by some special patterns. I have improved the selection
462 criteria to be more general.
464 PH/38 The host_find_failed option in the manualroute router can now be set
465 to "ignore", to completely ignore a host whose IP address cannot be
466 found. If all hosts are ignored, the behaviour is controlled by the new
467 host_all_ignored option.
469 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
470 homing or because of multiple MX records with /mx) generated more than
471 one IP address, and the following item turned out to be the local host,
472 all the secondary addresses of the first item were incorrectly removed
473 from the list, along with the local host and any following hosts (which
474 is what is supposed to happen).
476 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
477 whoever called Exim into the -H file. In the case of the daemon it was
478 behaving confusingly. When first started, it used values for whoever
479 started the daemon, but after a SIGHUP it used the Exim user (because it
480 calls itself on a restart). I have changed the code so that it now always
483 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
484 message are rejected with the same error (e.g. no authentication or bad
485 sender address), and a DATA command is nevertheless sent (as can happen
486 with PIPELINING or a stupid MUA), the error message that was given to the
487 RCPT commands is included in the rejection of the DATA command. This is
488 intended to be helpful for MUAs that show only the final error to their
491 PH/42 Another patch from the Sieve maintainer.
493 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
494 Eximstats - Fixed some broken HTML links and added missing column headers
496 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
497 columns for Rejects, Temp Rejects, Ham, and Spam rows.
499 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
501 PH/43 Yet another patch from the Sieve maintainer.
503 PH/44 I found a way to check for a TCP/IP connection going away before sending
504 the response to the final '.' that terminates a message, but only in the
505 case where the client has not sent further data following the '.'
506 (unfortunately, this is allowed). However, in many cases there won't be
507 any further data because there won't be any more messages to send. A call
508 to select() can be used: if it shows that the input is "ready", there is
509 either input waiting, or the socket has been closed. An attempt to read
510 the next input character can distinguish the two cases. Previously, Exim
511 would have sent an OK response which the client would never have see.
512 This could lead to message repetition. This fix should cure that, at
513 least in a lot of common cases.
515 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
516 advertised in response to EHLO.
522 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
523 fixed by 4.65/MH/01 (is this a record?) are fixed:
525 (i) An empty string was always treated as zero by the numeric comparison
526 operators. This behaviour has been restored.
528 (ii) It is documented that the numeric comparison operators always treat
529 their arguments as decimal numbers. This was broken in that numbers
530 starting with 0 were being interpreted as octal.
532 While fixing these problems I realized that there was another issue that
533 hadn't been noticed. Values of message_size_limit (both the global option
534 and the transport option) were treated as octal if they started with 0.
535 The documentation was vague. These values are now always treated as
536 decimal, and I will make that clear in the documentation.
542 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
543 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
546 MH/01 Don't check that the operands of numeric comparison operators are
547 integers when their expansion is in "skipping" mode (fixes bug
548 introduced by 4.64-PH/07).
550 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
551 child addresses, Exim now panics and dies. Previously, because the count
552 is held in a short int, deliveries were likely to be lost. As such a
553 large number of recipients for a single message is ridiculous
554 (performance will be very, very poor), I have chosen to impose a limit
555 rather than extend the field.
561 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
562 leftover -K file (the existence of which was triggered by #402).
563 While we were at it, introduced process PID as part of the -K
564 filename. This should rule out race conditions when creating
567 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
568 processing considerably. Previous code took too long for large mails,
569 triggering a timeout which in turn triggers #401.
571 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
572 in the DK code in transports.c. sendfile() is not really portable,
573 hence the _LINUX specificness.
575 TF/01 In the add_headers option to the mail command in an Exim filter,
576 there was a bug that Exim would claim a syntax error in any
577 header after the first one which had an odd number of characters
580 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
581 callout verification, Exim cached a "reject" for the entire domain. This
582 is correct for most verifications, but it is not correct for a recipient
583 verification with use_sender or use_postmaster set, because in that case
584 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
585 case of MAIL FROM:<> rejection from other early rejections (e.g.
586 rejection of HELO). When verifying a recipient using a non-null MAIL
587 address, the cache is ignored if it shows MAIL FROM:<> rejection.
588 Whatever the result of the callout, the value of the domain cache is
589 left unchanged (for any other kind of callout, getting as far as trying
590 RCPT means that the domain itself is ok).
592 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
595 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
596 manifest itself as EPIPE rather than ECONNECT. When tidying away a
597 session, the daemon ignores ECONNECT errors and logs others; it now
598 ignores EPIPE as well.
600 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
601 (quoted-printable decoding).
603 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
604 later the small subsequent patch to fix an introduced bug.
606 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
608 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
610 PH/08 An error is now given if message_size_limit is specified negative.
612 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
613 to be given (somewhat) arbitrary names.
615 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
618 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
619 miscellaneous code fixes
621 PH/10 Added the log_reject_target ACL modifier to specify where to log
624 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
625 hostname. This is wrong, because it relates to the incoming message (and
626 probably the interface on which it is arriving) and not to the outgoing
627 callout (which could be using a different interface). This has been
628 changed to use the value of the helo_data option from the smtp transport
629 instead - this is what is used when a message is actually being sent. If
630 there is no remote transport (possible with a router that sets up host
631 addresses), $smtp_active_hostname is used.
633 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
634 tweaks were necessary in order to get it to work (see also 21 below):
635 (a) The code assumed that strncpy() returns a negative number on buffer
636 overflow, which isn't the case. Replaced with Exim's string_format()
638 (b) There were several signed/unsigned issues. I just did the minimum
639 hacking in of casts. There is scope for a larger refactoring.
640 (c) The code used strcasecmp() which is not a standard C function.
641 Replaced with Exim's strcmpic() function.
642 (d) The code set only $1; it now sets $auth1 as well.
643 (e) A simple test gave the error "authentication client didn't specify
644 service in request". It would seem that Dovecot has changed its
645 interface. Fortunately there's a specification; I followed it and
646 changed what the client sends and it appears to be working now.
648 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
651 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
652 address A is aliased to B and C, where B exists and C does not. Without
653 -v the output is "A verified" because verification stops after a
654 successful redirection if more than one address is generated. However,
655 with -v the child addresses are also verified. Exim was outputting "A
656 failed to verify" and then showing the successful verification for C,
657 with its parentage. It now outputs "B failed to verify", showing B's
658 parentage before showing the successful verification of C.
660 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
661 look up a TXT record in a specific list after matching in a combined
664 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
665 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
666 they consult the DNS. I had assumed they would set it the way they
667 wanted; and indeed my experiments on Linux seem to show that in some
668 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
669 To be on the safe side, however, I have now made the interface to
670 host_find_byname() similar to host_find_bydns(), with an argument
671 containing the DNS resolver options. The host_find_byname() function now
672 sets these options at its start, just as host_find_bydns() does. The smtp
673 transport options dns_qualify_single and dns_search_parents are passed to
674 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
675 of host_find_byname() use the default settings of RES_DEFNAMES
676 (qualify_single) but not RES_DNSRCH (search_parents).
678 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
679 spool_read_header() do less string testing, by means of a preliminary
680 switch on the second character of optional "-foo" lines. (This is
681 overdue, caused by the large number of possibilities that now exist.
682 Originally there were few.) While I was there, I also converted the
683 str(n)cmp tests so they don't re-test the leading "-" and the first
684 character, in the hope this might squeeze out yet more improvement.
686 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
687 flag allowing group syntax was set by the header_syntax check but not
688 turned off, possible causing trouble later; (2) The flag was not being
689 set at all for the header_verify test, causing "group"-style headers to
690 be rejected. I have now set it in this case, and also caused header_
691 verify to ignore an empty address taken from a group. While doing this, I
692 came across some other cases where the code for allowing group syntax
693 while scanning a header line wasn't quite right (mostly, not resetting
694 the flag correctly in the right place). These bugs could have caused
695 trouble for malformed header lines. I hope it is now all correct.
697 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
698 with the "reply" argument non-NULL. The code, however (which originally
699 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
700 but it didn't always do it. This confused somebody who was copying the
701 code for some other use. I have removed all the tests.
703 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
704 feature that was used to support insecure browsers during the U.S. crypto
705 embargo. It requires special client support, and Exim is probably the
706 only MTA that supported it -- and would never use it because real RSA is
707 always available. This code has been removed, because it had the bad
708 effect of slowing Exim down by computing (never used) parameters for the
709 RSA_EXPORT functionality.
711 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
712 authenticator to fail if there's a tab character in the incoming data
713 (there should never be unless someone is messing about, as it's supposed
714 to be base64-encoded). Also added, on Timo's advice, the "secured" option
715 if the connection is using TLS or if the remote IP is the same as the
716 local IP, and the "valid-client-cert option" if a client certificate has
719 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
720 authenticators. This can be used for authorization after authentication
721 succeeds. (In the case of plaintext, it servers for both authentication
724 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
725 if any retry times were supplied.
727 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
728 connection, where there is no client IP address to check. In this
729 situation, the verify now always succeeds.
731 PH/25 Applied John Jetmore's -Mset patch.
733 PH/26 Added -bem to be like -Mset, but loading a message from a file.
735 PH/27 In a string expansion for a processed (not raw) header when multiple
736 headers of the same name were present, leading whitespace was being
737 removed from all of them, but trailing whitespace was being removed only
738 from the last one. Now trailing whitespace is removed from each header
739 before concatenation. Completely empty headers in a concatenation (as
742 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
743 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
745 PH/29 [Removed. This was a change that I later backed out, and forgot to
746 correct the ChangeLog entry (that I had efficiently created) before
747 committing the later change.]
749 PH/30 Exim was sometimes attempting to deliver messages that had suffered
750 address errors (4xx response to RCPT) over the same connection as other
751 messages routed to the same hosts. Such deliveries are always "forced",
752 so retry times are not inspected. This resulted in far too many retries
753 for the affected addresses. The effect occurred only when there were more
754 hosts than the hosts_max_try setting in the smtp transport when it had
755 the 4xx errors. Those hosts that it had tried were not added to the list
756 of hosts for which the message was waiting, so if all were tried, there
757 was no problem. Two fixes have been applied:
759 (i) If there are any address or message errors in an SMTP delivery, none
760 of the hosts (tried or untried) are now added to the list of hosts
761 for which the message is waiting, so the message should not be a
762 candidate for sending over the same connection that was used for a
763 successful delivery of some other message. This seems entirely
764 reasonable: after all the message is NOT "waiting for some host".
765 This is so "obvious" that I'm not sure why it wasn't done
766 previously. Hope I haven't missed anything, but it can't do any
767 harm, as the worst effect is to miss an optimization.
769 (ii) If, despite (i), such a delivery is accidentally attempted, the
770 routing retry time is respected, so at least it doesn't keep
771 hammering the server.
773 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
774 in ${readsocket because some servers need this prod.
776 PH/32 Added some extra debug output when updating a wait-xxx database.
778 PH/33 The hint "could be header name not terminated by colon", which has been
779 given for certain expansion errors for a long time, was not being given
780 for the ${if def:h_colon_omitted{... case.
782 PH/34 The spec says: "With one important exception, whenever a domain list is
783 being scanned, $domain contains the subject domain." There was at least
784 one case where this was not true.
786 PH/35 The error "getsockname() failed: connection reset by peer" was being
787 written to the panic log as well as the main log, but it isn't really
788 panic-worthy as it just means the connection died rather early on. I have
789 removed the panic log writing for the ECONNRESET error when getsockname()
792 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
793 runs only) independently of the message's sender address. This meant
794 that, if the 4xx error was in fact related to the sender, a different
795 message to the same recipient with a different sender could confuse
796 things. In particualar, this can happen when sending to a greylisting
797 server, but other circumstances could also provoke similar problems.
798 I have changed the default so that the retry time for these errors is now
799 based a combination of the sender and recipient addresses. This change
800 can be overridden by setting address_retry_include_sender=false in the
803 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
804 remote server are returned as part of bounce messages. This was not
805 happening for LMTP over a pipe (the lmtp transport), but now it is the
806 same for both kinds of LMTP.
808 PH/38 Despite being documented as not happening, Exim was rewriting addresses
809 in header lines that were in fact CNAMEs. This is no longer the case.
811 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
812 and queue runs started by the daemon processed all messages. This has
813 been fixed so that -R and -S can now usefully be given with -q<time>.
815 PH/40 Import PCRE release 6.7 (fixes some bugs).
817 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
819 PH/42 Give an error if -q is specified more than once.
821 PH/43 Renamed the variables $interface_address and $interface_port as
822 $received_ip_address and $received_port, to make it clear that these
823 values apply to message reception, and not to the outgoing interface when
824 a message is delivered. (The old names remain recognized, of course.)
826 PH/44 There was no timeout on the connect() call when using a Unix domain
827 socket in the ${readsocket expansion. There now is.
829 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
830 be meaningful with "accept".
832 SC/01 Eximstats V1.43
833 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
835 SC/02 Eximstats V1.44
836 Use a glob alias rather than an array ref in the generated
837 parser. This improves both readability and performance.
839 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
840 Collect SpamAssassin and rejection statistics.
841 Don't display local sender or destination tables unless
842 there is data to show.
843 Added average volumes into the top table text output.
845 SC/04 Eximstats V1.46
846 Collect data on the number of addresses (recipients)
847 as well as the number of messages.
849 SC/05 Eximstats V1.47
850 Added 'Message too big' to the list of mail rejection
851 reasons (thanks to Marco Gaiarin).
853 SC/06 Eximstats V1.48
854 Mainlog lines which have GMT offsets and are too short to
855 have a flag are now skipped.
857 SC/07 Eximstats V1.49 (Alain Williams)
858 Added the -emptyok flag.
860 SC/08 Eximstats V1.50
861 Fixes for obtaining the IP address from reject messages.
863 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
864 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
865 whitesspace changes from 4.64-PH/27
867 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
870 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
871 are found, allow negative numbers in numeric criteria)
873 JJ/06 exipick.20061117.2, added new $message_body_missing variable
875 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
876 to match changes made in 4.64-PH/43
878 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
880 PH/47 Put in an explicit test for a DNS lookup of an address record where the
881 "domain" is actually an IP address, and force a failure. This locks out
882 those revolvers/nameservers that support "A-for-A" lookups, in
883 contravention of the specifications.
885 PH/48 When a host name was looked up from an IP address, and the subsequent
886 forward lookup of the name timed out, the host name was left in
887 $sender_host_name, contrary to the specification.
889 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
890 restricted to single-key lookups, Exim was not diagnosing an error if
891 * or *@ was used with a query-style lookup.
893 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
895 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
896 long ago, but noone interested enough thought of it. Let's just say that
897 the "1.1" means that there are some new functions that weren't there at
898 some point in the past.
900 PH/51 Error processing for expansion failure of helo_data from an smtp
901 transport during callout processing was broken.
903 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
904 tested/used via the -bh/-bhc/-bs options.
906 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
907 bug, fixed in subsequent PCRE releases).
909 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
910 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
912 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
918 SC/01 Use a glob alias rather than an array ref in eximstats generated
919 parser. This improves both readability and performance.
921 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
922 Don't display local sender or destination tables in eximstats unless
923 there is data to show.
924 Added average volumes into the eximstats top table text output.
926 SC/03 Collect data on the number of addresses (recipients) as well
927 as the number of messages in eximstats.
929 TF/01 Correct an error in the documentation for the redirect router. Exim
930 does (usually) call initgroups() when daemonizing.
932 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
933 with consistent privilege compared to when running as a daemon.
935 TF/03 Note in the spec that $authenticated_id is not set for local
936 submissions from trusted users.
938 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
939 Thanks to Dean Brooks <dean@iglou.com> for the patch.
941 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
942 by adding some example configuration directives to the default
943 configuration file. A little bit of work is required to uncomment the
944 directives and define how usernames and passwords are checked, but
945 there is now a framework to start from.
947 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
948 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
949 without this. I don't know how relevant this is to other LDAP libraries.
951 PH/02 Add the verb name to the "unknown ACL verb" error.
953 PH/03 Magnus Holmgren's patch for filter_prepend_home.
955 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
957 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
958 directory not expanded when it should be if an expanded home directory
959 was set for the address (which is overridden by the transport).
961 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
964 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
965 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
966 because it is too late at that time, and has no effect.
968 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
969 security issue with \' (bugzilla #107). I could not use the
970 PQescapeStringConn() function, because it needs a PGconn value as one of
973 PH/08 When testing addresses using -bt, indicate those final addresses that
974 are duplicates that would not cause an additional delivery. At least one
975 person was confused, thinking that -bt output corresponded to deliveries.
976 (Suppressing duplicates isn't a good idea as you lose the information
977 about possibly different redirections that led to the duplicates.)
979 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
980 systems where poll() doesn't work, in particular OS X.
982 PH/10 Added more information to debugging output for retry time not reached.
984 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
985 operations in malware.c.
987 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
990 PH/13 If write_rejectlog was set false when logging was sent to syslog with
991 syslog_duplication set false, log lines that would normally be written
992 both the the main log and to the reject log were not written to syslog at
995 PH/14 In the default configuration, change the use of "message" in ACL warn
996 statements to "add_header".
998 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
999 not followed by a command (e.g. "seen endif").
1001 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1002 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1005 PH/17 Added extra conditions to the default value of delay_warning_condition
1009 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1010 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1011 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1014 The Auto-Submitted: and various List- headers are standardised, whereas I
1015 don't think Precedence: ever was.
1017 PH/18 Refactored debugging code in route_finduser() to show more information,
1018 in particular, the error code if getpwnam() issues one.
1020 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1021 This is apparently needed in addition to the PH/07 change above to avoid
1022 any possible encoding problems.
1024 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1025 but not after initializing Perl.
1027 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1028 output them only if debugging. By default they are written stderr,
1029 apparently, which is not desirable.
1031 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1034 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1037 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1039 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1040 authenticated or an ident call has been made. Suppress the default
1041 values for $authenticated_id and $authenticated_sender (but permit -oMai
1042 and -oMas) when testing with -bh.
1044 PH/24 Re-jigged the order of the tests in the default configuration so that the
1045 tests for valid domains and recipients precede the DNS black list and CSA
1046 tests, on the grounds that those ones are more expensive.
1048 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1049 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1050 This bug exists in every version of Exim that I still have, right back to
1053 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1054 However, an attempt to turn on case-sensitivity in a regex key by
1055 including (?-i) didn't work because the subject string was already
1056 lowercased, and the effects were non-intuitive. It turns out that a
1057 one-line patch can be used to allow (?-i) to work as expected.
1063 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1064 other effects) broke the use of negated acl sub-conditions.
1066 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1069 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1070 "Deny" causes Exim to reject the incoming connection with a 554 error.
1071 Unfortunately, if there is a major crisis, such as a disk failure,
1072 tcp-wrappers gives "deny", whereas what one would like would be some
1073 kind of temporary error. A kludge has been added to help with this.
1074 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1075 554 error is used if errno is still zero or contains ENOENT (which occurs
1076 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1079 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1081 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1082 errors. Otherwise a message that provokes a temporary error (when other
1083 messages do not) can cause a whole host to time out.
1085 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1086 addresses were routed directly to files or pipes from a redirect router.
1087 File deliveries just didn't batch; pipe deliveries might have suffered
1090 PH/06 A failure to get a lock for a hints database would erroneously always say
1091 "Failed to get write lock", even when it was really a read lock.
1093 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1094 of 0600. This has been changed to use the value of the lockfile_mode
1095 option (which defaults to 0600).
1097 PH/08 Applied small patch from the Sieve maintainer.
1099 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1100 folder from quota calculations, a direct delivery into this folder messed
1101 up the contents of the maildirsize file. This was because the regex was
1102 used only to exclude .Trash (or whatever) when the size of the mailbox
1103 was calculated. There was no check that a delivery was happening into an
1104 excluded directory. This bug has been fixed by ignoring all quota
1105 processing for deliveries into excluded directories.
1107 PH/10 Added the maildirfolder_create_regex option to appendfile.
1113 PH/01 The code for finding all the local interface addresses on a FreeBSD
1114 system running IPv6 was broken. This may well have applied to all BSD
1115 systems, as well as to others that have similar system calls. The broken
1116 code found IPv4 interfaces correctly, but gave incorrect values for the
1117 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1118 that it would not match correctly against @[] and not recognize the IPv6
1121 PH/02 The ipliteral router was not recognizing addresses of the form user@
1122 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1124 PH/03 Added disable_ipv6.
1126 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1127 decoded form, because it is most often used to construct To: headers
1128 lines in autoreplies, and the decoded form may well be syntactically
1129 invalid. However, $reply_address has leading white space removed, and all
1130 newlines turned into spaces so that the autoreply transport does not
1133 PH/05 If group was specified without a user on a router, and no group or user
1134 was specified on a transport, the group from the router was ignored.
1136 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1137 for visible compile-time settings that can be used to change these
1138 numbers, for those that want even more. Backwards compatibility with old
1139 spool files has been maintained. However, going back to a previous Exim
1140 release will lost any variables that are in spool files.
1142 PH/07 Two small changes when running in the test harness: increase delay when
1143 passing a TCP/IP connection to a new process, in case the original
1144 process has to generate a bounce, and remove special handling of
1145 127.0.0.2 (sic), which is no longer necessary.
1147 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1148 be the same on different OS.
1150 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1153 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1154 whether --show-vars was specified or not
1156 JJ/02 exipick: Added support for new ACL variable spool format introduced
1159 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1160 syntactically invalid From: or Reply-to: line, and a filter used this to
1161 generate an autoreply, and therefore failed to obtain an address for the
1162 autoreply, Exim could try to deliver to a non-existent relative file
1163 name, causing unrelated and misleading errors. What now happens is that
1164 it logs this as a hard delivery error, but does not attempt to create a
1167 PH/11 The exinext utility has a -C option for testing purposes, but although
1168 the given file was scanned by exinext itself; it wasn't being passed on
1169 when Exim was called.
1171 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1172 an end-of-file indication when reading a command response.
1174 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1175 compiled. In many other places in Exim, IPv6 addresses are always
1176 recognized, so I have changed this. It also means that IPv4 domain
1177 literals of the form [IPV4:n.n.n.n] are now always recognized.
1179 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1180 used if the router is not running as root, for example, when verifying at
1181 ACL time, or when using -bh. The debugging output from this situation was
1182 non-existent - all you got was a failure to exec. I have made two
1185 (a) Failures to set uid/gid, the current directory, or a process leader
1186 in a subprocess such as that created by queryprogram now generate
1187 suitable debugging ouput when -d is set.
1189 (b) The queryprogram router detects when it is not running as root,
1190 outputs suitable debugging information if -d is set, and then runs
1191 the subprocess without attempting to change uid/gid.
1193 PH/15 Minor change to Makefile for building test_host (undocumented testing
1196 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1197 additional section of a DNS packet that returns MX or SRV records.
1198 Instead, it always explicitly searches for A/AAAA records. This avoids
1199 major problems that occur when a DNS server includes only records of one
1200 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1201 fixed another bug: if SRV records were looked up and the corresponding
1202 address records were *not* found in the additional section, the port
1203 values from the SRV records were lost.
1205 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1206 using the correct key (the original address) when searching the retry
1207 rules in order to find which one to use for generating the retry hint.
1209 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1210 adding the default one. Similarly, if it contains a Reply-To: header, the
1211 errors_reply_to option, if set, is not used.
1213 PH/19 When calculating a retry time, Exim used to measure the "time since
1214 failure" by looking at the "first failed" field in the retry record. Now
1215 it does not use this if it is later than than the arrival time of the
1216 message. Instead it uses the arrival time. This makes for better
1217 behaviour in cases where some deliveries succeed, thus re-setting the
1218 "first failed" field. An example is a quota failure for a huge message
1219 when small messages continue to be delivered. Without this change, the
1220 "time since failure" will always be short, possible causing more frequent
1221 delivery attempts for the huge message than are intended.
1222 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1224 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1225 $1, $2, $3) because the numerical variables can be reset during some
1226 expansion items (e.g. "match"), thereby losing the authentication data.
1228 PH/21 Make -bV show the size of off_t variables so that the test suite can
1229 decide whether to run tests for quotas > 2G.
1231 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1232 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1233 filecount value is greater than 2G or if a quota value is greater than 2G
1234 on a system where the size of off_t is not greater than 4, a panic error
1237 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1238 never match. The debug and -bh output now contains an explicit error
1239 message indicating a malformed IPv4 address or mask.
1241 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1242 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1243 PH/23 above applies.
1245 PH/25 Do not write to syslog when running in the test harness. The only
1246 occasion when this arises is a failure to open the main or panic logs
1247 (for which there is an explicit test).
1249 PH/26 Added the /no_tell option to "control=freeze".
1251 PH/27 If a host name lookup failed very early in a connection, for example, if
1252 the IP address matched host_lookup and the reverse lookup yielded a name
1253 that did not have a forward lookup, an error message of the form "no IP
1254 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1255 could be logged. Now it outputs the IP address instead of "NULL".
1257 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1258 allows the sender and the authenticated sender to be set when
1259 submitting a message from within Exim. Since child_open_exim() is
1260 documented for local_scan(), the new function should be too.
1262 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1263 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1264 results in an empty string is now treated as unset.
1266 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1268 PH/31 Added sender_verify_fail logging option.
1270 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1271 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1272 tidied the source and removed it altogether.
1274 PH/33 When a queue run was abandoned because the load average was too high, a
1275 log line was always written; now it is written only if the queue_run log
1276 selector is set. In addition, the log line for abandonment now contains
1277 information about the queue run such as the pid. This is always present
1278 in "start" and "stop" lines but was omitted from the "abandon" line.
1280 PH/34 Omit spaces between a header name and the colon in the error message that
1281 is given when verify = headers_syntax fails (if there are lots of them,
1282 the message gets confusing).
1284 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1285 names, as there are now some PTR records that contain slashes. This check
1286 is only to protect against broken name servers that fall over on strange
1287 characters, so the fact that it applies to all lookups doesn't matter.
1289 PH/36 Now that the new test suite is complete, we can remove some of the
1290 special code in Exim that was needed for the old test suite. For example,
1291 sorting DNS records because real resolvers return them in an arbitrary
1292 order. The new test suite's fake resolver always returns records in the
1295 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1296 bounces) except when queue_only is set, to avoid logging races between
1297 the different processes.
1299 PH/38 Panic-die if .include specifies a non-absolute path.
1301 PH/39 A tweak to the "H" retry rule from its user.
1303 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1304 a label. They prevented compilation on older perls.
1306 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1307 a warning to be raised on newish perls.
1309 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1310 on queue. Changes to match documented behaviour of showing count of
1311 messages matching specified criteria.
1313 PH/40 Changed the default ident timeout from 30s to 5s.
1315 PH/41 Added support for the use of login_cap features, on those BSD systems
1316 that have them, for controlling the resources used by pipe deliveries.
1318 PH/42 The content-scanning code uses fopen() to create files in which to put
1319 message data. Previously it was not paying any attention to the mode of
1320 the files. Exim runs with umask(0) because the rest of the code creates
1321 files with open(), and sets the required mode explicitly. Thus, these
1322 files were ending up world-writeable. This was not a big issue, because,
1323 being within the spool directory, they were not world-accessible. I have
1324 created a function called modefopen, which takes an additional mode
1325 argument. It sets umask(777), creates the file, chmods it to the required
1326 mode, then resets the umask. All the relevant calls to fopen() in the
1327 content scanning code have been changed to use this function.
1329 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1330 to 24 hours. This avoids potential overflow problems when processing G
1331 and H retry rules. I suspect nobody ever tinkers with this value.
1333 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1335 PH/45 When the plaintext authenticator is running as a client, the server's
1336 challenges are checked to ensure they are valid base64 strings. By
1337 default, the authentication attempt is cancelled if an invalid string is
1338 received. Setting client_ignore_invalid_base64 true ignores these errors.
1339 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1340 they are received. Thus, the responses can be made to depend on the
1341 challenges. If an invalid string is ignored, an empty string is placed in
1344 PH/46 Messages that are created by the autoreply transport now contains a
1345 References: header, in accordance with RFCs 2822 and 3834.
1347 PH/47 Added authenticated_sender_force to the smtp transport.
1349 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1351 PH/49 Installed latest patch from the Sieve maintainer.
1353 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1354 was also set, the appendfile transport was unnecessarily scanning a
1355 directory of message files (e.g. for maildir delivery) to find the count
1356 of files (along with the size), even though it did not need this
1357 information. It now does the scan only if it needs to find either the
1358 size of the count of files.
1360 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1362 PH/52 Two bugs concerned with error handling when the smtp transport is
1365 (i) Exim was not creating retry information for temporary errors given
1366 for individual recipients after the DATA command when the smtp transport
1367 was used in LMTP mode. This meant that they could be retried too
1368 frequently, and not timed out correctly.
1370 (ii) Exim was setting the flag that allows error details to be returned
1371 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1372 recipients that were returned after the DATA command.
1374 PH/53 This is related to PH/52, but is more general: for any failing address,
1375 when detailed error information was permitted to be returned to the
1376 sender, but the error was temporary, then after the final timeout, only
1377 "retry timeout exceeded" was returned. Now it returns the full error as
1378 well as "retry timeout exceeded".
1380 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1381 do this, and (what is worse) MTAs that accept it.
1383 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1384 will now be deprecated.
1386 PH/56 New os.c-cygwin from the Cygwin maintainer.
1388 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1389 formats (previously only available in exim formats via -bpr, -bpru,
1390 and -bpra. Now also available in native and exiqgrep formats)
1392 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1393 with very large, slow to parse queues
1395 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1397 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1399 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1400 responses to authentication challenges, though it was showing the
1401 challenges; (ii) I've removed the CR characters from the debug output for
1404 PH/58 Allow for the insertion of a newline as well as a space when a string
1405 is turned into more than one encoded-word during RFC 2047 encoding. The
1406 Sieve code now uses this.
1408 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1409 data_4xx, lost_connection, tls_required.
1411 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1412 message was being sent as an SMTP response.
1414 PH/61 Add -l and -k options to exicyclog.
1416 PH/62 When verifying, if an address was redirected to one new address, so that
1417 verification continued, and the new address failed or deferred after
1418 having set something in $address_data, the value of $address_data was not
1419 passed back to the ACL. This was different to the case when no
1420 redirection occurred. The value is now passed back in both cases.
1422 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1423 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1424 use login_cap.h, so on its own it isn't the distinguishing feature. The
1425 new name refers directly to the setclassresources() function.
1427 PH/65 Added configuration files for NetBSD3.
1429 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1431 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1432 is preferred over IPv4.
1434 PH/68 The bounce_return_message and bounce_return_body options were not being
1435 honoured for bounces generated during the reception of non-SMTP messages.
1436 In particular, this applied to messages rejected by the ACL. This bug has
1437 been fixed. However, if bounce_return_message is true and bounce_return_
1438 body is false, the headers that are returned for a non-SMTP message
1439 include only those that have been read before the error was detected.
1440 (In the case of an ACL rejection, they have all been read.)
1442 PH/69 The HTML version of the specification is now built in a directory called
1443 spec_html instead of spec.html, because the latter looks like a path with
1444 a MIME-type, and this confuses some software.
1446 PH/70 Catch two compiler warnings in sieve.c.
1448 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1449 function verify_get_ident() calls ip_connect() to connect a socket, but
1450 if the "connect()" function timed out, ip_connect() used to close the
1451 socket. However, verify_get_ident() also closes the socket later, and in
1452 between Exim writes to the log, which may get opened at this point. When
1453 the socket was closed in ip_connect(), the log could get the same file
1454 descriptor number as the socket. This naturally causes chaos. The fix is
1455 not to close the socket in ip_connect(); the socket should be closed by
1456 the function that creates it. There was only one place in the code where
1457 this was missing, in the iplookup router, which I don't think anybody now
1458 uses, but I've fixed it anyway.
1460 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1461 well as to direct DNS lookups. Otherwise the handling of names in host
1462 lists is inconsistent and therefore confusing.
1468 PH/01 Two changes to the default runtime configuration:
1470 (1) Move the checks for relay_from_hosts and authenticated clients from
1471 after to before the (commented out) DNS black list checks.
1473 (2) Add control=submission to the relay_from_hosts and authenticated
1474 clients checks, on the grounds that messages accepted by these
1475 statements are most likely to be submissions.
1477 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1479 (1) Generate an error if the third argument for the ${prvs expansion is
1482 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1485 (3) Reset the variables that are obtained from the first argument of
1486 ${prvscheck and used in the second argument before leaving the code,
1487 because their memory is reclaimed, so using them afterwards may do
1490 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1491 one (it's much easier than Tom thought :-).
1493 (5) Because of (4), we can now allow for the use of $prvscheck_result
1494 inside the third argument.
1496 PH/03 For some reason, the default setting of PATH when running a command from
1497 a pipe transport was just "/usr/bin". I have changed it to
1500 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1501 anything to be listed in the output from -bV.
1503 PH/05 When a filter generated an autoreply, the entire To: header line was
1504 quoted in the delivery log line, like this:
1506 => >A.N.Other <ano@some.domain> <original@ddress> ...
1508 This has been changed so that it extracts the operative address. There
1509 may be more than one such address. If so, they are comma-separated, like
1512 => >ano@some.domain,ona@other.domain <original@ddress> ...
1514 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1515 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1516 not being looked up in the rDNS to get a host name, Exim was showing the
1517 IP address twice in Received: lines, even though the IP addresses were
1518 identical. For example:
1520 Received: from [1.2.3.4] (helo=[1.2.3.4])
1522 However, if the real host name was known, it was omitting the HELO data
1523 if it matched the actual IP address. This has been tidied up so that it
1524 doesn't show the same IP address twice.
1526 PH/07 When both +timestamp and +memory debugging was on, the value given by
1527 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1528 called by the debug printing, thereby overwriting the timestamp buffer.
1529 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1531 PH/08 When the original message was included in an autoreply transport, it
1532 always said "this is a copy of the message, including all the headers",
1533 even if body_only or headers_only was set. It now gives an appropriate
1536 PH/09 Applied a patch from the Sieve maintainer which:
1538 o fixes some comments
1539 o adds the (disabled) notify extension core
1540 o adds some debug output for the result of if/elsif tests
1541 o points to the current vacation draft in the documentation
1542 and documents the missing references header update
1546 o fixes a bug in processing the envelope test (when testing
1547 multiple envelope elements, the last element determinted the
1550 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1551 Electronic Mail") by including:
1553 Auto-submitted: auto-generated
1555 in the messages that it generates (bounce messages and others, such as
1556 warnings). In the case of bounce messages for non-SMTP mesages, there was
1557 also a typo: it was using "Auto_submitted" (underscore instead of
1558 hyphen). Since every message generated by Exim is necessarily in response
1559 to another message, thes have all been changed to:
1561 Auto-Submitted: auto-replied
1563 in accordance with these statements in the RFC:
1565 The auto-replied keyword:
1567 - SHOULD be used on messages sent in direct response to another
1568 message by an automatic process,
1570 - MUST NOT be used on manually-generated messages,
1572 - MAY be used on Delivery Status Notifications (DSNs) and Message
1573 Disposition Notifications (MDNs),
1575 - MUST NOT be used on messages generated by automatic or periodic
1576 processes, except for messages which are automatic responses to
1579 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1580 to the default Received: header definition.
1582 PH/12 Added log selector acl_warn_skipped (default on).
1584 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1585 variables because (a) they are in the wrong storage pool and (b) even if
1586 they were copied, it wouldn't work properly because of the caching.
1588 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1589 checking when decoding. Apparently there are clients that generate
1590 overlong encoded strings. Why am I not surprised?
1592 PH/15 If the first argument of "${if match_address" was not empty, but did not
1593 contain an "@" character, Exim crashed. Now it writes a panic log message
1594 and treats the condition as false.
1596 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1598 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1599 extension "envelope-auth". The code is finished and in agreement with
1600 other implementations, but there is no documentation so far and in fact,
1601 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1602 not changing the active code.
1604 Print executed "if" and "elsif" statements when debugging is used. This
1605 helps a great deal to understand what a filter does.
1607 Document more things not specified clearly in RFC3028. I had all this
1608 sorted out, when out of a sudden new issues came to my mind. Oops."
1610 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1613 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1614 canonical form (as documented). However, after a host name lookup from
1615 the IP address, check_host() was doing a simple string comparison with
1616 addresses acquired from the DNS when checking that the found name did
1617 have the original IP as one of its addresses. Since any found IPv6
1618 addresses are likely to be in abbreviated form, the comparison could
1619 fail. Luckily, there already exists a function for doing the comparison
1620 by converting both addresses to binary, so now that is used instead of
1621 the text comparison.
1623 PH/20 There was another similar case to PH/19, when a complete host name was
1624 given in a host list; looking up its IP address could give an abbreviated
1625 form, whereas the current host's name might or might not be abbreviated.
1626 The same fix has been applied.
1632 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1633 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1636 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1637 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1639 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1641 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1642 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1643 Also, for "Auto-Submitted", treat anything other than "no" as
1644 non-personal, in accordance with RFC 3834. (Previously it treated
1645 anything starting "auto-" as non-personal.)
1647 TF/01 The control=submission/name=... option had a problem with syntax
1648 errors if the name included a slash character. The /name= option
1649 now slurps the rest of the string, so it can include any characters
1650 but it must come last in the list of options (after /sender_retain
1653 PH/05 Some modifications to the interface to the fake nameserver for the new
1661 TK/01 Added the "success_on_redirect" address verification option. See
1662 NewStuff for rationale and an example.
1664 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1666 PH/02 Patch to exigrep to allow it to work on syslog lines.
1668 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1669 fread() to skip over the body file's header line, because in Cygwin the
1670 header line is locked and is inaccessible.
1672 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1673 co-exist for some time) to make it clear that it is the Exim ID that is
1674 referenced, not the Message-ID: header line.
1676 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1677 string_format() function, because snprintf() does not exist on all
1680 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1681 ${stat: expansion item.
1683 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1684 protocol synchronization error", to keep the pedants happy.
1686 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1687 well as for IRIX systems, when gcc is being used. See the host.c source
1690 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1692 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1695 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1696 transport and to the smtp transport in LMTP mode.
1698 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1700 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1701 run a filter in a subprocess. This could lead to confusion in subsequent
1702 lookups in the parent process. There should also be a search_tidyup() at
1703 the end of the subprocess.
1705 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1706 only if the host matched helo_try_verify_hosts, which caused the
1707 verification to occur when the EHLO/HELO command was issued. The ACL just
1708 tested the remembered result. Now, if a previous verification attempt has
1709 not happened, "verify = helo" does it there and then.
1711 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1713 TK/03 Fix log output including CR from clamd.
1715 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1716 exist provoked a memory error which could cause a segfault.
1718 PH/15 Installed PCRE 6.2
1720 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1722 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1723 of the problem. Specifically, suggested +O2 rather than +O1 for the
1726 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1728 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1729 record, the port was not correctly logged when the outgoing_port log
1730 selector was set (it logged the transort's default port).
1732 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1733 fallback_hosts, and "hosts" in the smtp transport.
1735 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1736 host errors such as "Connection refused".
1738 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1739 authentication with radiusclient 0.4.9:
1741 - Error returned from rc_read_config was caught wrongly
1742 - Username/password not passed on to radius server due to wrong length.
1744 The presumption is that some radiusclient API changes for 4.51/PH/17
1745 were not taken care of correctly. The code is still untested by me (my
1746 Linux distribution still has 0.3.2 of radiusclient), but it was
1747 contributed by a Radius user.
1749 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1750 expanding the "port" option of the smtp transport.
1752 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1753 while reading a MIME header. Thanks to Tom Hughes for a patch.
1755 PH/24 Include config.h inside local_scan.h so that configuration settings are
1758 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1759 that in an ACL for RCPT (for example), you can examine exactly what was
1762 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1763 commands, but it was not correctly comparing the address with the actual
1764 client host address. Thus, it would show the EHLO address in Received:
1765 header lines when this was not necessary.
1767 PH/27 Added the % operator to ${eval:}.
1769 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1770 it should be ignoring failures (because with -C, for example, it has lost
1771 privilege). It wasn't ignoring creation failures other than "already
1774 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1777 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1778 because an input line was too long, either on its own, or by virtue of
1779 too many continuations, the temporary file was not being removed, and the
1780 return code was incorrect.
1782 PH/31 Missing "BOOL" in function definition in filtertest.c.
1784 PH/32 Applied Sieve patches from the maintainer.
1786 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1788 PH/33 Added "verify = not_blind".
1790 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1791 Local/Makefile (with some defaults set). These are used in built scripts
1792 such as exicyclog, but they have never been used in the exim_install
1793 script (though there are many overriding facilities there). I have
1794 arranged that the exim_install script now takes note of these two
1797 PH/35 Installed configuration files for Dragonfly.
1799 PH/36 When a locally submitted message by a trusted user did not contain a
1800 From: header, and the sender address was obtained from -f or from an SMTP
1801 MAIL command, and the trusted user did not use -F to supply a sender
1802 name, $originator_name was incorrectly used when constructing a From:
1803 header. Furthermore, $originator_name was used for submission mode
1804 messages from external hosts without From: headers in a similar way,
1805 which is clearly wrong.
1807 PH/37 Added control=suppress_local_fixups.
1809 PH/38 When log_selector = +received_sender was set, and the addition of the
1810 sender made the log line's construction buffer exactly full, or one byte
1811 less than full, an overflow happened when the terminating "\n" was
1814 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1815 when the result of a list match is failure because a DNS lookup failed.
1817 PH/40 RM_COMMAND is now used in the building process.
1819 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1820 the "build-* directories that it finds.
1822 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1823 address in a domain literal was a prefix of an interface address.
1825 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1826 when verifying a sender address, unless rewrite_headers is false.
1828 PH/44 Wrote a long comment about why errors_to addresses are verified as
1829 recipients, not senders.
1831 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1832 the ratelimit ACL was added.
1834 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1836 PH/46 Added extra information about PostgreSQL errors to the error string.
1838 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1839 avoiding the need to install special zones in a real server. This is
1840 backwards compatible; if it can't find the fake resolver, it drops back.
1841 Thus, both old and new test suites can be run.
1843 TF/02 Added util/ratelimit.pl
1845 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1846 clock is set back in time.
1848 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1849 Candler <B.Candler@pobox.com>.
1851 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1852 OK for addresses that are the result of redirections.
1854 PH/48 A number of further additions for the benefit of the new test suite,
1855 including a fake gethostbyname() that interfaces to the fake DNS resolver
1858 TF/06 The fix for widen_domains has also been applied to qualify_single and
1859 search_parents which are the other dnslookup options that can cause
1862 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1865 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1867 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1868 service" if it encountered an MX record with an empty target hostname.
1869 The message is now "an MX or SRV record indicated no SMTP service".
1871 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1872 if the DNS of the sending site is misconfigured. This is quite a
1873 common situation. This change restores the behaviour of treating a
1874 helo verification defer as a failure.
1876 PH/51 If self=fail was set on a router, the bounce message did not include the
1877 actual error message.
1883 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1885 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1886 command itself ended in error, the underlying message about the transport
1887 filter timeout was being overwritten with the pipe command error. Now the
1888 underlying error message should be appended to the second error message.
1890 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1892 PH/02 Reduce the amount of output that "make" produces by default. Full output
1893 can still be requested.
1895 PH/03 The warning log line about a condition test deferring for a "warn" verb
1896 was being output only once per connection, rather than after each
1897 occurrence (because it was using the same function as for successful
1898 "warn" verbs). This seems wrong, so I have changed it.
1900 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1901 it should not have, which might have caused a crash in the right
1902 circumstances, but probably never did.
1904 PH/04 Installed a modified version of Tony Finch's patch to make submission
1905 mode fix the return path as well as the Sender: header line, and to
1906 add a /name= option so that you can make the user's friendly name appear
1909 TF/03 Added the control = fakedefer ACL modifier.
1911 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1912 Mark Lowes for thorough testing.
1914 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1916 TK/03 Merged latest SRS patch from Miles Wilton.
1918 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1919 with the definition in sysexits.h (which is #included earlier).
1920 Fortunately, Exim does not actually use EX_OK. The code used to try to
1921 preserve the sysexits.h value, by assumimg that macro definitions were
1922 scanned for macro replacements. I have been disabused of this notion,
1923 so now the code just undefines EX_OK before #including unistd.h.
1925 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1926 in the smtp transport. When a block could not be written in a single
1927 write() function, the timeout was being re-applied to each part-write.
1928 This seems wrong - if the receiver was accepting one byte at a time it
1929 would take for ever. The timeout is now adjusted when this happens. It
1930 doesn't have to be particularly precise.
1932 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1933 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1935 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1936 without a domain if the check to <postmaster@domain> fails.
1938 SC/01 Eximstats: added -xls and the ability to specify output files
1939 (patch written by Frank Heydlauf).
1941 SC/02 Eximstats: use FileHandles for outputing results.
1943 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1945 SC/04 Eximstats: fixed display of large numbers with -nvr option
1947 SC/05 Eximstats: fixed merging of reports with empty tables.
1949 SC/06 Eximstats: added the -include_original_destination flag
1951 SC/07 Eximstats: removed tabs and trailing whitespace.
1953 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1955 TK/06 MBOX spool code: Add real "From " MBOX separator line
1956 so the .eml file is really in mbox format (even though
1957 most programs do not really care). Patch from Alex Miller.
1959 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1960 The latter is generated from $received_to and is only set if the
1961 message has one envelope recipient. SA can use these headers,
1962 obviously out-of-the-box. Patch from Alex Miller.
1964 PH/08 The ${def test on a variable was returning false if the variable's
1965 value was "0", contrary to what the specification has always said!
1966 The result should be true unless the variable is empty.
1968 PH/09 The syntax error of a character other than { following "${if
1969 def:variable_name" (after optional whitespace) was not being diagnosed.
1970 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1971 accidental colon was present, for example, could give incorrect results.
1973 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1974 result is used (not including appendfile, where other changes are about
1977 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1978 This involved changing a lot of size variables from int to off_t. It
1979 should work with maildirs and everything.
1981 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1982 spamd dying while we are connected to it.
1984 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1987 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1988 function for BDB 4.3.
1990 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1992 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1993 the sources for PCRE have been re-arranged and more files are now
1996 PH/15 The code I had for printing potentially long long variables in PH/11
1997 above was not the best (it lost precision). The length of off_t variables
1998 is now inspected at build time, and an appropriate printing format (%ld
1999 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2000 to be "long long int" or "long int". This is needed for the internal
2001 formatting function string_vformat().
2003 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2004 the configuration file to be ":syslog", then the script "guesses" where
2005 the logs files are, rather than using the compiled in default. In our
2006 case the guess is not the same as the compiled default, so the script
2007 suddenly stopped working when I started to use syslog. The patch checks
2008 to see if log_file_path is "". If so, it attempts to read it from exim
2009 with no configuration file to get the compiled in version, before it
2010 falls back to the previous guessing code."
2012 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2013 implementing BATV in an Exim configuration. See NewStuff for the gory
2016 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2017 Makefile that are specific to HP-UX.
2019 PH/18 If the "use_postmaster" option was set for a recipient callout together
2020 with the "random" option, the postmaster address was used as the MAIL
2021 FROM address for the random test, but not for the subsequent recipient
2022 test. It is now used for both.
2024 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2025 patch removes a few documentation additions to RFC 3028, because the
2026 latest draft now contains them. It adds the new en;ascii-case comparator
2027 and a new error check for 8bit text in MIME parts. Comparator and
2028 require names are now matched exactly. I enabled the subaddress
2029 extension, but it is not well tested yet (read: it works for me)."
2031 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2032 rework some of the code of TK/09 above to avoid the hardwired use of
2033 "%lld" and "long long". Replaced the call to snprintf() with a call to
2036 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2037 records point to non-existent hosts", "retry timeout exceeded", and
2038 "retry time not reached for any host after a long failure period".
2040 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2041 experimental DomainKeys support:
2043 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2044 (2) On an error such as an illegally used "control", the wrong name for
2045 the control was given.
2047 These problems did NOT occur unless DomainKeys support was compiled.
2049 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2051 PH/24 Added ${if match_ip condition.
2053 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2054 databases so that it will be absolutely obvious if a crash occurs in the
2055 DB library. This is a regular occurrence (often caused by mis-matched
2058 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2059 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2060 picked up on a user's system that detects such things. There doesn't seem
2061 to be a gcc warning option for this - only an attribute that has to be
2062 put on the function's prototype. It seems that in Fedora Core 4 they have
2063 set this on a number of new functions. No doubt there will be more in due
2066 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2067 specify a transport. However, if an address that was verified by such a
2068 router was the subject of a callout, Exim crashed because it tried to
2069 read the rcpt_include_affixes from the non-existent transport. Now it
2070 just assumes that the setting of that option is false. This bug was
2071 introduced by 4.51/PH/31.
2073 PH/28 Changed -d+all to exclude +memory, because that information is very
2074 rarely of interest, but it makes the output a lot bigger. People tend to
2075 do -d+all out of habit.
2077 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2078 code in os-type was giving problems when libc.so lives in lib64, like on
2081 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2082 aren't the modern standard, and it seems that some systems' include files
2083 don't always have them. Exim was already checking for some of the newer
2084 ones like T_AAAA, and defining it itself. I've added checks for all the
2085 record types that Exim uses.
2087 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2088 not automatically generating a new one, as it is supposed to. This
2089 prevented TLS from working. If the file did exist, but contained invalid
2090 data, a new version was generated, as expected. It was only the case of a
2091 non-existent file that was broken.
2093 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2094 with a change in libdomainkeys > 0.64.
2096 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2097 from DNS. If the selector record carries the flag, it now has
2098 precedence over the domain-wide flag.
2100 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2102 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2103 the use of a port name that isn't defined in /etc/services) occurred, the
2104 message was deferred as in a normal delivery, and thus remained on the
2105 spool, instead of being failed because of the mua_wrapper setting. This
2106 is now fixed, and I tidied up some of the mua_wrapper messages at the
2109 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2110 the messages in a hash of arrays rather than using individual hashes.
2111 This is a bit cleaner and results in dramatic memory savings, albeit
2112 at a slight CPU cost.
2114 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2115 as requested by Marc Sherman.
2117 SC/10 Eximstats: added histograms for user specified patterns as requested
2120 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2122 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2123 fopen() in the content-scanning modules that did not already have it.
2129 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2130 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2132 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2134 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2136 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2137 to test Sieve filters that use "vacation".
2139 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2140 that changes the way the GnuTLS parameters are stored in the cache file.
2141 The new format can be generated externally. For backward compatibility,
2142 if the data in the cache doesn't make sense, Exim assumes it has read an
2143 old-format file, and it generates new data and writes a new file. This
2144 means that you can't go back to an older release without removing the
2147 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2148 work if there are any delivery delays because "one_time" forces the
2149 parent to be marked "delivered", so its unseen clone is never tried
2150 again. For this reason, Exim now forbids the simultaneous setting of
2153 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2154 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2155 entry if you want to know the details. The fix, however, neglected to
2156 consider the case when local delivery batching is involved. The test for
2157 "previously delivered" was not happening when checking to see if an
2158 address could be batched with a previous (undelivered) one; under
2159 certain circumstances this could lead to multiple deliveries to the same
2162 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2163 in its include files, and this causes problems building Exim.
2165 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2166 header_syntax) but Exim was just ignoring anything given after a slash.
2167 In particular, this caused confusion with an attempt to use "verify =
2168 reverse_host_lookup/defer_ok". An error is now given when options are
2169 supplied for verify items that do not have them. (Maybe reverse_host_
2170 lookup should have a defer_ok option, but that's a different point.)
2172 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2173 defined by RFC 821) to 2048, because there were problems with some AUTH
2174 commands, and RFC 1869 says the size should be increased for extended
2175 SMTP commands that take arguments.
2177 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2180 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2181 "unknown" error; now it says that the functionality isn't in the binary.
2183 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2184 an address' error message when a string expansion fails (syntax or
2185 whatever). Otherwise the password may appear in the log. Following change
2186 PH/42 below, there is no longer a chance of it appearing in a bounce
2189 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2191 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2192 of its addresses was ever tried. (Bugzilla bug #2.)
2194 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2195 the result incorrectly in the debug output. (It correctly added a newline
2196 to what was transported.)
2198 TF/01 Added $received_time.
2200 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2201 commented out examples of how to interface to a virus scanner and to
2202 SpamAssassin. Also added commented examples of av_scanner and
2203 spamd_address settings.
2205 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2206 and controls are allowed in which ACLs. There were a couple of minor
2207 errors. Some of the entries in the conditions table (which is a table of
2208 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2209 negation of where the condition IS allowed.
2211 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2213 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2214 header file does not have a version number, so I've had to invent a new
2215 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2216 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2217 radiusclient), but it was contributed by a Radius user.
2219 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2220 files or directories, for OpenSSL.
2222 PH/19 When an Exim process that is running as root has to create an Exim log
2223 file, it does so in a subprocess that runs as exim:exim so as to get the
2224 ownership right at creation (otherwise, other Exim processes might see
2225 the file with the wrong ownership). There was no test for failure of this
2226 fork() call, which would lead to the process getting stuck as it waited
2227 for a non-existent subprocess. Forks do occasionally fail when resources
2228 run out. I reviewed all the other calls to fork(); they all seem to check
2231 PH/20 When checking for unexpected SMTP input at connect time (before writing
2232 the banner), Exim was not dealing correctly with a non-positive return
2233 from the read() function. If the client had disconnected by this time,
2234 the result was a log entry for a synchronization error with an empty
2235 string after "input=" when read() returned zero. If read() returned -1
2236 (an event I could not check), uninitialized data bytes were printed.
2237 There were reports of junk text (parts of files, etc) appearing after
2240 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2242 PH/22 Added support for macro redefinition, and (re)definition in between
2243 driver and ACL definitions.
2245 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2246 forgetting to use the resulting value; it was using the unexpanded value.
2248 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2249 hadn't been configured. The fix is from Juergen Kreileder, who
2250 understands it better than I do:
2252 "Here's what I see happening with three configured cyrus_sasl
2253 authenticators configured (plain, login, cram-md5):
2255 On startup auth_cyrus_sasl_init() gets called for each of these.
2256 This means three calls to sasl_listmech() without a specified mech_list.
2257 => SASL tests which mechs of all available mechs actually work
2258 => three warnings about OTP not working
2259 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2261 With the patch, sasl_listmech() also gets called three times. But now
2262 SASL's mech_list option is set to the server_mech specified in the the
2263 authenticator. Or in other words, the answer from sasl_listmech()
2264 gets limited to just the mech you're testing for (which is different
2266 => the return list contains just 'plain' or 'login', 'cram-md5' or
2267 nothing depending on the value of ob->server_mech.
2269 I've just tested the patch: Authentication still works fine,
2270 unavailable mechs specified in the exim configuration are still
2271 caught, and the auth.log warnings about OTP are gone."
2273 PH/25 When debugging is enabled, the contents of the command line are added
2274 to the debugging output, even when log_selector=+arguments is not
2277 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2278 answer is "GNU", and only if the return is "GNU/something" is the answer
2281 PH/27 $acl_verify_message is now set immediately after the failure of a
2282 verification in an ACL, and so is available in subsequent modifiers. In
2283 particular, the message can be preserved by coding like this:
2285 warn !verify = sender
2286 set acl_m0 = $acl_verify_message
2288 Previously, $acl_verify_message was set only while expanding "message"
2289 and "log_message" when a very denied access.
2291 PH/28 Modified OS/os.c-Linux with
2293 -#ifndef OS_LOAD_AVERAGE
2294 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2296 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2297 nomenclature these days.)
2299 PH/29 Installed patch from the Sieve maintainer that adds the options
2300 sieve_useraddress and sieve_subaddress to the redirect router.
2302 PH/30 In these circumstances:
2303 . Two addresses routed to the same list of hosts;
2304 . First host does not offer TLS;
2305 . First host accepts first address;
2306 . First host gives temporary error to second address;
2307 . Second host offers TLS and a TLS session is established;
2308 . Second host accepts second address.
2309 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2310 and peerdn, if requested) that were in fact used only for the second
2313 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2314 attention to any local part prefix or suffix that was matched by the
2315 router that accepted the address. It now behaves in the same way as it
2316 does for delivery: the affixes are removed from the local part unless
2317 rcpt_include_affixes is set on the transport.
2319 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2320 timeout during the DATA phase of an incoming message.
2322 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2323 applied a patch sanctioned by the Sieve maintainer.
2325 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2326 the uid or gid is negative. A case of a negative gid caused this to be
2327 noticed. The fix allows for either to be negative.
2329 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2330 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2333 PH/36 The change PH/12 above was broken. Fixed it.
2335 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2336 the grounds that routing the same address twice would always produce the
2337 same answer. This might have been true once, but it is certainly no
2338 longer true now. Routing a child address may depend on the previous
2339 routing that produced that child. Some complicated redirection strategies
2340 went wrong when messages had multiple recipients, and made Exim's
2341 behaviour dependent on the order in which the addresses were given.
2343 I have moved the duplicate checking until after the routing is complete.
2344 Exim scans the addresses that are assigned to local and remote
2345 transports, and removes any duplicates. This means that more work will be
2346 done, as duplicates will always all be routed, but duplicates are
2347 presumably rare, so I don't expect this is of any significance.
2349 For deliveries to pipes, files, and autoreplies, the duplicate checking
2350 still happens during the routing process, since they are not going to be
2353 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2354 It corrects a timeout issue with spamd. This is Ian's comment: "The
2355 background is that sometimes spamd either never reads data from a
2356 connection it has accepted, or it never writes response data. The exiscan
2357 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2358 blindly assumes that writes won't block so it may never time out."
2360 PH/39 Allow G after quota size as well as K and M.
2362 PH/40 The value set for $authenticated_id in an authenticator may not contain
2363 binary zeroes or newlines because the value is written to log lines and
2364 to spool files. There was no check on this. Now the value is run through
2365 the string_printing() function so that such characters are converted to
2366 printable escape sequences.
2368 PH/41 $message_linecount is a new variable that contains the total number of
2369 lines in the message. Compare $body_linecount, which is the count for the
2372 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2373 bounce and delay warning messages, except in certain special cases, which
2376 (a) An SMTP error message from a remote host;
2377 (b) A message specified in a :fail: redirection;
2378 (c) A message specified in a "fail" command in a system filter;
2379 (d) A message specified in a FAIL return from the queryprogram router;
2380 (e) A message specified by the cannot_route_message router option.
2382 In these cases only, Exim does include the error details in bounce and
2383 warning messages. There are also a few cases where bland messages such
2384 as "unrouteable address" or "local delivery error" are given.
2386 PH/43 $value is now also set for the "else" part of a ${run expansion.
2388 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2389 being worked on, but at least Exim now implements the latest version to
2392 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2393 process to complete was treated as a delivery failure, a timeout while
2394 writing the message to the pipe was logged, but erroneously treated as a
2395 successful delivery. Such timeouts include transport filter timeouts. For
2396 consistency with the overall process timeout, these timeouts are now
2397 treated as errors, giving rise to delivery failures by default. However,
2398 there is now a new Boolean option for the pipe transport called
2399 timeout_defer, which, if set TRUE, converts the failures into defers for
2400 both kinds of timeout. A transport filter timeout is now identified in
2403 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2404 systems where "make" and "gmake" are different, calling "gmake" at top
2405 level broke things. I've arranged for the value of $(MAKE) to be passed
2406 from the Makefile to this script so that it can call the same version of
2410 A note about Exim versions 4.44 and 4.50
2411 ----------------------------------------
2413 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2414 changes of various kinds. As a consequence, a big documentation update was
2415 needed. This delayed the release for rather longer than seemed good, especially
2416 in the light of a couple of (minor) security issues. Therefore, the changes
2417 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2418 release. So 4.44 and 4.50 are in effect two different branches that both start
2421 I have left the 4.50 change log unchanged; it contains all the changes since
2422 4.43. The change log for 4.44 is below; many of its items are identical to
2423 those for 4.50. This seems to be the most sensible way to preserve the
2424 historical information.
2430 1. Minor wording change to the doc/README.SIEVE file.
2432 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2433 computation of the current number of files was incorrect.
2435 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2436 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2439 4. Give more explanation in the error message when the command for a transport
2440 filter fails to execute.
2442 5. There are several places where Exim runs a non-Exim command in a
2443 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2444 was being done only for the command run by the queryprogram router. It is
2445 now done for all such subprocesses. The other cases are: ${run, transport
2446 filters, and the commands run by the lmtp and pipe transports.
2448 6. Added CONFIGURE_GROUP build-time option.
2450 7. Some older OS have a limit of 256 on the maximum number of file
2451 descriptors. Exim was using setrlimit() to set 1000 as a large value
2452 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2453 systems. I've change it so that if it can't get 1000, it tries for 256.
2455 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2456 was an oversight, and furthermore, ever since the addition of extra
2457 controls (e.g. 4.43/32), the checks on when to allow different forms of
2458 "control" were broken. There should now be diagnostics for all cases when a
2459 control that does not make sense is encountered.
2461 9. Added the /retain_sender option to "control=submission".
2463 10. $recipients is now available in the predata ACL (oversight).
2465 11. Tidy the search cache before the fork to do a delivery from a message
2466 received from the command line. Otherwise the child will trigger a lookup
2467 failure and thereby defer the delivery if it tries to use (for example) a
2468 cached ldap connection that the parent has called unbind on.
2470 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2471 of $address_data from the recipient verification was clobbered by the
2472 sender verification.
2474 13. The value of address_data from a sender verification is now available in
2475 $sender_address_data in subsequent conditions in the ACL statement.
2477 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2479 15. Added a new option "connect=<time>" to callout options, to set a different
2482 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2483 was its contents. (It was OK if the option was not defined at all.)
2485 17. A "Completed" log line is now written for messages that are removed from
2486 the spool by the -Mrm option.
2488 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2489 information about exactly what failed.
2491 19. Added -dd to debug only the daemon process.
2493 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2494 handles timeouts, both on the server side and network timeouts. Renamed the
2495 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2497 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2498 It is now set to "smtps".
2500 22. $host_address is now set to the target address during the checking of
2501 ignore_target_hosts.
2503 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2504 being passed; this would have caused $sender_host_name to have been used if
2505 matching the list had actually called for a host name (not very likely,
2506 since this list is usually IP addresses). A host name is now passed as
2509 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2510 code that specifies a non-restarting handler (typically sigaction() in
2511 modern systems) in an attempt to fix a rare and obscure crash bug.
2513 25. Narrowed the window for a race in the daemon that could cause it to ignore
2514 SIGCHLD signals. This is not a major problem, because they are used only to
2515 wake it up if nothing else does.
2517 26. A malformed maildirsize file could cause Exim to calculate negative values
2518 for the mailbox size or file count. Odd effects could occur as a result.
2519 The maildirsize information is now recalculated if the size or filecount
2522 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2523 support for a long time. Removed HAVE_SYS_VFS_H.
2525 28. Installed the latest version of exipick from John Jetmore.
2527 29. In an address list, if the pattern was not a regular expression, an empty
2528 subject address (from a bounce message) matched only if the pattern was an
2529 empty string. Non-empty patterns were not even tested. This was the wrong
2530 because it is perfectly reasonable to use an empty address as part of a
2531 database query. An empty address is now tested by patterns that are
2532 lookups. However, all the other forms of pattern expect the subject to
2533 contain a local part and a domain, and therefore, for them, an empty
2534 address still always fails if the pattern is not itself empty.
2536 30. Exim went into a mad DNS loop when attempting to do a callout where the
2537 host was specified on an smtp transport, and looking it up yielded more
2538 than one IP address.
2540 31. Re-factored the code for checking spool and log partition space into a
2541 function that finds that data and another that does the check. The former
2542 is then used to implement four new variables: $spool_space, $log_space,
2543 $spool_inodes, and $log_inodes.
2545 32. The RFC2047 encoding function was originally intended for short strings
2546 such as real names; it was not keeping to the 75-character limit for
2547 encoded words that the RFC imposes. It now respects the limit, and
2548 generates multiple encoded words if necessary. To be on the safe side, I
2549 have increased the buffer size for the ${rfc2047: expansion operator from
2552 33. It is now permitted to omit both strings after an "if" condition; if the
2553 condition is true, the result is "true". As before, when the second string
2554 is omitted, a false condition yields an empty string. This makes it less
2555 cumbersome to write custom ACL and router conditions.
2557 34. Failure to deliver a bounce message always caused it to be frozen, even if
2558 there was an errors_to setting on the router. The errors_to setting is now
2561 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2562 canonical form (fully expanded) before being placed in
2563 $sender_host_address.
2565 36. The table in the code that translates DNS record types into text (T_A to
2566 "A" for instance) was missing entries for NS and CNAME. It is just possible
2567 that this could have caused confusion if both these types were looked up
2568 for the same domain, because the text type is used as part of Exim's
2569 per-process caching. But the chance of anyone hitting this buglet seems
2572 37. The dnsdb lookup has been extended in a number of ways.
2574 (1) There is a new type, "zns", which walks up the domain tree until it
2575 finds some nameserver records. It should be used with care.
2577 (2) There is a new type, "mxh", which is like "mx" except that it returns
2578 just the host names, not the priorities.
2580 (3) It is now possible to give a list of domains (or IP addresses) to be
2581 looked up. The behaviour when one of the lookups defers can be
2582 controlled by a keyword.
2584 (4) It is now possible to specify the separator character for use when
2585 multiple records are returned.
2587 38. The dnslists ACL condition has been extended: it is now possible to supply
2588 a list of IP addresses and/or domains to be looked up in a particular DNS
2591 39. Added log_selector=+queue_time_overall.
2593 40. When running the queue in the test harness, wait just a tad after forking a
2594 delivery process, to get repeatability of debugging output.
2596 41. Include certificate and key file names in error message when GnuTLS fails
2597 to set them up, because the GnuTLS error message doesn't include the name
2598 of the failing file when there is a problem reading it.
2600 42. Allow both -bf and -bF in the same test run.
2602 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2604 44. The "Exiscan patch" is now merged into the mainline Exim source.
2606 45. Sometimes the final signoff response after QUIT could fail to get
2607 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2608 < 0 before doing a fflush(). This bug looks as though it goes back to the
2609 introduction of TLS in release 3.20, but "sometimes" must have been rare
2610 because the tests only now provoked it.
2612 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2613 (this can affect the format of dates).
2615 47. exim_tidydb, when checking for the continued existence of a message for
2616 which it has found a message-specific retry record, was not finding
2617 messages that were in split spool directories. Consequently, it was
2618 deleting retry records that should have stayed in existence.
2620 48. Steve fixed some bugs in eximstats.
2622 49. The SPA authentication driver was not abandoning authentication and moving
2623 on to the next authenticator when an expansion was forced to fail,
2624 contradicting the general specification for all authenticators. Instead it
2625 was generating a temporary error. It now behaves as specified.
2627 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2628 (the order specifies the preference for clients). The order is now AES256,
2629 AES128, 3DES, ARCFOUR128.
2631 51. Small patch to Sieve code - explicitly set From: when generating an
2634 52. Exim crashed if a remote delivery caused a very long error message to be
2635 recorded - for instance if somebody sent an entire SpamAssassin report back
2636 as a large number of 550 error lines. This bug was coincidentally fixed by
2637 increasing the size of one of Exim's internal buffers (big_buffer) that
2638 happened as part of the Exiscan merge. However, to be on the safe side, I
2639 have made the code more robust (and fixed the comments that describe what
2642 53. Now that there can be additional text after "Completed" in log lines (if
2643 the queue_time_overall log selector is set), a one-byte patch to exigrep
2644 was needed to allow it to recognize "Completed" as not the last thing in
2647 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2648 patch that reportedly fixes this has been added. I am not expert enough to
2649 create a test for it. This is what the patch creator wrote:
2651 "I found a little strange behaviour of ldap code when working with
2652 Windows 2003 AD Domain, where users was placed in more than one
2653 Organization Units. When I tried to give exim partial DN, the exit code
2654 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2655 But simultaneously result of request was absolutely normal ldap result,
2656 so I produce this patch..."
2658 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2659 REFERENCE, so I have modified the code to exclude the patch when that macro
2662 55. Some experimental protocols are using DNS PTR records for new purposes. The
2663 keys for these records are domain names, not reversed IP addresses. The
2664 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2665 leaves it alone. Component reversal etc. now happens only for IP addresses.
2668 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2670 57. Double the size of the debug message buffer (to 2048) so that more of very
2671 long debug lines gets shown.
2673 58. The exicyclog utility now does better if the number of log files to keep
2674 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2676 59. Two changes related to the smtp_active_hostname option:
2678 (1) $smtp_active_hostname is now available as a variable.
2679 (2) The default for smtp_banner uses $smtp_active_hostname instead
2680 of $primary_hostname.
2682 60. The host_aton() function is supposed to be passed a string that is known
2683 to be a valid IP address. However, in the case of IPv6 addresses, it was
2684 not checking this. This is a hostage to fortune. Exim now panics and dies
2685 if the condition is not met. A case was found where this could be provoked
2686 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2687 components; fortuitously, this particular loophole had already been fixed
2688 by change 4.50/55 above.
2690 If there are any other similar loopholes, the new check in host_aton()
2691 itself should stop them being exploited. The report I received stated that
2692 data on the command line could provoke the exploit when Exim was running as
2693 exim, but did not say which command line option was involved. All I could
2694 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2695 running as the user.
2698 61. There was a buffer overflow vulnerability in the SPA authentication code
2699 (which came originally from the Samba project). I have added a test to the
2700 spa_base64_to_bits() function which I hope fixes it.
2703 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2704 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2706 63. The daemon start-up calls getloadavg() while still root for those OS that
2707 need the first call to be done as root, but it missed one case: when
2708 deliver_queue_load_max is set with deliver_drop_privilege. This is
2709 necessary for the benefit of the queue runner, because there is no re-exec
2710 when deliver_drop_privilege is set.
2712 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2713 This has been fixed.
2715 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2716 was in use, was not putting the data itself into the right store pool;
2717 consequently, it could be overwritten for a subsequent message in the same
2718 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2721 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2723 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2724 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2725 the function were treating the return as a boolean value, which happened to
2726 work because 0=false and not-0=true, but is not correct code.
2728 68. The host_aton() function was not handling scoped IPv6 addresses (those
2729 with, for example, "%eth0" on the end) correctly.
2731 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2732 negated items (that is, ~something) in unsigned ints. Some compilers
2733 apparently mutter when there is no cast.
2735 70. If an address verification called from an ACL failed, and did not produce a
2736 user-specific message (i.e. there was only a "system" message), nothing was
2737 put in $acl_verify_message. In this situation, it now puts the system
2740 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2741 session; change 4.31/43 added the unwanted input to the log line - except
2742 that it did not do this in the start of session case. It now does.
2744 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2745 This is wrong and can cause the other end to generate a synchronization
2746 error if it is another Exim or anything else that does the synchronization
2747 check. A QUIT command is no longer sent after a timeout.
2749 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2750 during host lookups.
2752 74. The defer_ok option of callout verification was not working if it was used
2753 when verifying addresses in header lines, that is, for this case:
2755 verify = header_sender/callout=defer_ok
2757 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2758 those file descriptors could be used for SMTP connections. If anything
2759 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2760 could be sent to the SMTP client, causing chaos. The daemon now opens
2761 stdin, stdout, and stderr to /dev/null when it puts itself into the
2764 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2765 log by default. The user can override this with suitable Perl magic.
2767 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2768 the log message when discard triggers, was not working for the DATA ACL or
2769 for the non-SMTP ACL.
2771 78. Error message wording change in sieve.c.
2773 79. If smtp_accept_max_per_host was set, the number of connections could be
2774 restricted to fewer than expected, because the daemon was trying to set up
2775 a new connection before checking whether the processes handling previous
2776 connections had finished. The check for completed processes is now done
2777 earlier. On busy systems, this bug wouldn't be noticed because something
2778 else would have woken the daemon, and it would have reaped the completed
2781 80. If a message was submitted locally by a user whose login name contained one
2782 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2783 It caused a spool format error. I have fixed the spool reading code. A
2784 related problem was that the "from" clause in the Received: line became
2785 illegal because of the space(s). It is now covered by ${quote_local_part.
2787 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2790 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2792 83. Patch from Sieve maintainer to fix unterminated string problem in
2793 "vacation" handling.
2795 84. Some minor changes to the Linux configuration files to help with other
2796 OS variants using glibc.
2798 85. One more patch for Sieve to update vacation handling to latest spec.
2801 ----------------------------------------------------
2802 See the note above about the 4.44 and 4.50 releases.
2803 ----------------------------------------------------
2809 1. Change 4.43/35 introduced a bug that caused file counts to be
2810 incorrectly computed when quota_filecount was set in an appendfile
2813 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2814 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2817 3. Give more explanation in the error message when the command for a transport
2818 filter fails to execute.
2820 4. There are several places where Exim runs a non-Exim command in a
2821 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2822 was being done only for the command run by the queryprogram router. It is
2823 now done for all such subprocesses. The other cases are: ${run, transport
2824 filters, and the commands run by the lmtp and pipe transports.
2826 5. Some older OS have a limit of 256 on the maximum number of file
2827 descriptors. Exim was using setrlimit() to set 1000 as a large value
2828 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2829 systems. I've change it so that if it can't get 1000, it tries for 256.
2831 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2832 was an oversight, and furthermore, ever since the addition of extra
2833 controls (e.g. 4.43/32), the checks on when to allow different forms of
2834 "control" were broken. There should now be diagnostics for all cases when a
2835 control that does not make sense is encountered.
2837 7. $recipients is now available in the predata ACL (oversight).
2839 8. Tidy the search cache before the fork to do a delivery from a message
2840 received from the command line. Otherwise the child will trigger a lookup
2841 failure and thereby defer the delivery if it tries to use (for example) a
2842 cached ldap connection that the parent has called unbind on.
2844 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2845 of $address_data from the recipient verification was clobbered by the
2846 sender verification.
2848 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2849 was its contents. (It was OK if the option was not defined at all.)
2851 11. A "Completed" log line is now written for messages that are removed from
2852 the spool by the -Mrm option.
2854 12. $host_address is now set to the target address during the checking of
2855 ignore_target_hosts.
2857 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2858 being passed; this would have caused $sender_host_name to have been used if
2859 matching the list had actually called for a host name (not very likely,
2860 since this list is usually IP addresses). A host name is now passed as
2863 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2864 code that specifies a non-restarting handler (typically sigaction() in
2865 modern systems) in an attempt to fix a rare and obscure crash bug.
2867 15. Narrowed the window for a race in the daemon that could cause it to ignore
2868 SIGCHLD signals. This is not a major problem, because they are used only to
2869 wake it up if nothing else does.
2871 16. A malformed maildirsize file could cause Exim to calculate negative values
2872 for the mailbox size or file count. Odd effects could occur as a result.
2873 The maildirsize information is now recalculated if the size or filecount
2876 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2877 support for a long time. Removed HAVE_SYS_VFS_H.
2879 18. Updated exipick to current release from John Jetmore.
2881 19. Allow an empty sender to be matched against a lookup in an address list.
2882 Previously the only cases considered were a regular expression, or an
2885 20. Exim went into a mad DNS lookup loop when doing a callout where the
2886 host was specified on the transport, if the DNS lookup yielded more than
2889 21. The RFC2047 encoding function was originally intended for short strings
2890 such as real names; it was not keeping to the 75-character limit for
2891 encoded words that the RFC imposes. It now respects the limit, and
2892 generates multiple encoded words if necessary. To be on the safe side, I
2893 have increased the buffer size for the ${rfc2047: expansion operator from
2896 22. Failure to deliver a bounce message always caused it to be frozen, even if
2897 there was an errors_to setting on the router. The errors_to setting is now
2900 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2901 canonical form (fully expanded) before being placed in
2902 $sender_host_address.
2904 24. Updated eximstats to version 1.33
2906 25. Include certificate and key file names in error message when GnuTLS fails
2907 to set them up, because the GnuTLS error message doesn't include the name
2908 of the failing file when there is a problem reading it.
2910 26. Expand error message when OpenSSL has problems setting up cert/key files.
2913 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2914 (this can affect the format of dates).
2916 28. exim_tidydb, when checking for the continued existence of a message for
2917 which it has found a message-specific retry record, was not finding
2918 messages that were in split spool directories. Consequently, it was
2919 deleting retry records that should have stayed in existence.
2921 29. eximstats updated to version 1.35
2922 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2923 1.35 - bugfix such that pie charts by volume are generated correctly
2925 30. The SPA authentication driver was not abandoning authentication and moving
2926 on to the next authenticator when an expansion was forced to fail,
2927 contradicting the general specification for all authenticators. Instead it
2928 was generating a temporary error. It now behaves as specified.
2930 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2931 (the order specifies the preference for clients). The order is now AES256,
2932 AES128, 3DES, ARCFOUR128.
2934 31. Small patch to Sieve code - explicitly set From: when generating an
2937 32. Exim crashed if a remote delivery caused a very long error message to be
2938 recorded - for instance if somebody sent an entire SpamAssassin report back
2939 as a large number of 550 error lines. This bug was coincidentally fixed by
2940 increasing the size of one of Exim's internal buffers (big_buffer) that
2941 happened as part of the Exiscan merge. However, to be on the safe side, I
2942 have made the code more robust (and fixed the comments that describe what
2945 33. Some experimental protocols are using DNS PTR records for new purposes. The
2946 keys for these records are domain names, not reversed IP addresses. The
2947 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2948 leaves it alone. Component reversal etc. now happens only for IP addresses.
2951 34. The host_aton() function is supposed to be passed a string that is known
2952 to be a valid IP address. However, in the case of IPv6 addresses, it was
2953 not checking this. This is a hostage to fortune. Exim now panics and dies
2954 if the condition is not met. A case was found where this could be provoked
2955 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2956 components; fortuitously, this particular loophole had already been fixed
2957 by change 4.50/55 or 4.44/33 above.
2959 If there are any other similar loopholes, the new check in host_aton()
2960 itself should stop them being exploited. The report I received stated that
2961 data on the command line could provoke the exploit when Exim was running as
2962 exim, but did not say which command line option was involved. All I could
2963 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2964 running as the user.
2967 35. There was a buffer overflow vulnerability in the SPA authentication code
2968 (which came originally from the Samba project). I have added a test to the
2969 spa_base64_to_bits() function which I hope fixes it.
2972 36. The daemon start-up calls getloadavg() while still root for those OS that
2973 need the first call to be done as root, but it missed one case: when
2974 deliver_queue_load_max is set with deliver_drop_privilege. This is
2975 necessary for the benefit of the queue runner, because there is no re-exec
2976 when deliver_drop_privilege is set.
2978 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2979 was in use, was not putting the data itself into the right store pool;
2980 consequently, it could be overwritten for a subsequent message in the same
2981 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2984 38. Sometimes the final signoff response after QUIT could fail to get
2985 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2986 < 0 before doing a fflush(). This bug looks as though it goes back to the
2987 introduction of TLS in release 3.20, but "sometimes" must have been rare
2988 because the tests only now provoked it.
2994 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2995 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2996 Now it returns an integer. A number of calls were still expecting a T/F
2997 return. Fortuitously, in all cases, the tests worked in OK situations,
2998 which is the norm. However, things would have gone wrong on any write
2999 failures on the smtp file descriptor. This function is used when sending
3000 messages over SMTP and also when doing verify callouts.
3002 2. When Exim is called to do synchronous delivery of a locally submitted
3003 message (the -odf or -odi options), it no longer closes stderr before doing
3006 3. Implemented the mua_wrapper option.
3008 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3010 5. Implemented the functions header_remove(), header_testname(),
3011 header_add_at_position(), and receive_remove_recipient(), and exported them
3014 6. If an ACL "warn" statement specified the addition of headers, Exim already
3015 inserted X-ACL-Warn: at the start if there was no header name. However, it
3016 was not making this test for the second and subsequent header lines if
3017 there were newlines in the string. This meant that an invalid header could
3018 be inserted if Exim was badly configured.
3020 7. Allow an ACL "warn" statement to add header lines at the start or after all
3021 the Received: headers, as well as at the end.
3023 8. Added the rcpt_4xx retry error code.
3025 9. Added postmaster_mailfrom=xxx to callout verification option.
3027 10. Added mailfrom=xxxx to the callout verification option, for verify=
3030 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3031 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3033 12. Inserted some casts to stop certain compilers warning when using pointer
3034 differences as field lengths or precisions in printf-type calls (mostly
3035 affecting debugging statements).
3037 13. Added optional readline() support for -be (dynamically loaded).
3039 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3040 same clock tick as a message's arrival, so that its received time was the
3041 same as the "first fail" time on the retry record, and that message
3042 remained on the queue past the ultimate address timeout, every queue runner
3043 would try a delivery (because it was past the ultimate address timeout) but
3044 after another failure, the ultimate address timeout, which should have then
3045 bounced the address, did not kick in. This was a "< instead of <=" error;
3046 in most cases the first failure would have been in the next clock tick
3047 after the received time, and all would be well.
3049 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3050 being recognized when the domain list was tested by the match_domain
3051 condition in an expansion string.
3053 16. Added the ${str2b64: operator.
3055 17. Exim was always calling setrlimit() to set a large limit for the number of
3056 processes, without checking whether the existing limit was already
3057 adequate. (It did check for the limit on file descriptors.) Furthermore,
3058 errors from getrlimit() and setrlimit() were being ignored. Now they are
3059 logged to the main and panic logs, but Exim does carry on, to try to do its
3060 job under whatever limits there are.
3062 18. Imported PCRE 5.0.
3064 19. Trivial typo in log message " temporarily refused connection" (the leading
3067 20. If the log selector return_path_on_delivery was set and an address was
3068 redirected to /dev/null, the delivery process crashed because it assumed
3069 that a return path would always be set for a "successful" delivery. In this
3070 case, the whole delivery is bypassed as an optimization, and therefore no
3073 21. Internal re-arrangement: the function for sending a challenge and reading
3074 a response while authentication was assuming a zero-terminated challenge
3075 string. It's now changed to take a pointer and a length, to allow for
3076 binary data in such strings.
3078 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3080 23. Exim was not respecting finduser_retries when seeking the login of the
3081 uid under which it was called; it was always trying 10 times. (The default
3082 setting of finduser_retries is zero.) Also, it was sleeping after the final
3083 failure, which is pointless.
3085 24. Implemented tls_on_connect_ports.
3087 25. Implemented acl_smtp_predata.
3089 26. If the domain in control=submission is set empty, Exim assumes that the
3090 authenticated id is a complete email address when it generates From: or
3091 Sender: header lines.
3093 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3094 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3095 chown and chgrp in /bin and hostname in /usr/bin.
3097 28. Exim was keeping the "process log" file open after each use, just as it
3098 does for the main log. This opens the possibility of it remaining open for
3099 long periods when the USR1 signal hits a daemon. Occasional processlog
3100 errors were reported, that could have been caused by this. Anyway, it seems
3101 much more sensible not to leave this file open at all, so that is what now
3104 29. The long-running daemon process does not normally write to the log once it
3105 has entered its main loop, and it closes the log before doing so. This is
3106 so that log files can straightforwardly be renamed and moved. However,
3107 there are a couple of unusual error situations where the daemon does write
3108 log entries, and I had neglected to close the log afterwards.
3110 30. The text of an SMTP error response that was received during a remote
3111 delivery was being truncated at 512 bytes. This is too short for some of
3112 the long messages that one sometimes sees. I've increased the limit to
3115 31. It is now possible to make retry rules that apply only when a message has a
3116 specific sender, in particular, an empty sender.
3118 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3119 it possible to be selective about when SMTP synchronization is enforced.
3121 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3123 32. Implemented hosts_connection_nolog.
3125 33. Added an ACL for QUIT.
3127 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3130 35. Added mailbox_size and mailbox_filecount to appendfile.
3132 36. Added control = no_multiline_responses to ACLs.
3134 37. There was a bug in the logic of the code that waits for the clock to tick
3135 in the case where the clock went backwards by a substantial amount such
3136 that the microsecond fraction of "now" was more than the microsecond
3137 fraction of "then" (but the whole seconds number was less).
3139 38. Added support for the libradius Radius client library this is found on
3140 FreeBSD (previously only the radiusclient library was supported).
3146 1. When certain lookups returned multiple values in the form name=value, the
3147 quoting of the values was not always being done properly. Specifically:
3148 (a) If the value started with a double quote, but contained no whitespace,
3150 (b) If the value contained whitespace other than a space character (i.e.
3151 tabs or newlines or carriage returns) it was not quoted.
3152 This fix has been applied to the mysql and pgsql lookups by writing a
3153 separate quoting function and calling it from the lookup code. The fix
3154 should probably also be applied to nisplus, ibase and oracle lookups, but
3155 since I cannot test any of those, I have not disturbed their existing code.
3157 2. A hit in the callout cache for a specific address caused a log line with no
3158 reason for rejecting RCPT. Now it says "Previous (cached) callout
3159 verification failure".
3161 3. There was an off-by-one bug in the queryprogram router. An over-long
3162 return line was truncated at 256 instead of 255 characters, thereby
3163 overflowing its buffer with the terminating zero. As well as fixing this, I
3164 have increased the buffer size to 1024 (and made a note to document this).
3166 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3167 when Exim is waiting for an SMTP response from a remote server, Exim
3168 restarts its select() call on the socket, thereby resetting its timeout.
3169 This is not a problem when such interrupts are rare. Somebody set up a cron
3170 job to run exiwhat every 2 minutes, which is less than the normal select()
3171 timeout (5 or 10 minutes). This meant that the select() timeout never
3172 kicked in because it was always reset. I have fixed this by comparing the
3173 time when an interrupt arrives with the time at the start of the first call
3174 to select(). If more time than the timeout has elapsed, the interrupt is
3175 treated as a timeout.
3177 5. Some internal re-factoring in preparation for the addition of Sieve
3178 extensions (by MH). In particular, the "personal" test is moved to a
3179 separate function, and given an option for scanning Cc: and Bcc: (which is
3180 not set for Exim filters).
3182 6. When Exim created an email address using the login of the caller as the
3183 local part (e.g. when creating a From: or Sender: header line), it was not
3184 quoting the local part when it contained special characters such as @.
3186 7. Installed new OpenBSD configuration files.
3188 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3189 try to make them clearer.
3191 9. Callout options, other than the timeout value, were being ignored when
3192 verifying sender addresses in header lines. For example, when using
3194 verify = header_sender/callout=no_cache
3196 the cache was (incorrectly) being used.
3198 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3199 only the Cygwin environment.
3201 11. When return_path_on_delivery was set as a log selector, if different remote
3202 addresses in the same message used different return paths and parallel
3203 remote delivery occurred, the wrong values would sometimes be logged.
3204 (Whenever a remote delivery process finished, the return path value from
3205 the most recently started remote delivery process was logged.)
3207 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3208 lines when AUTH and/or TLS are in use. This is the "received protocol"
3209 field. Exim used to use "asmtp" for authenticated SMTP, without any
3210 indication (in the protocol name) for TLS use. Now it follows the RFC and
3211 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3212 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3213 names appear in log lines as well as in Received: header lines.
3215 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3216 extensions, and comparison tests, and to fix some bugs.
3218 14. Changes to the "personal" filter test:
3220 (1) The test was buggy in that it was just doing the equivalent of
3221 "contains" tests on header lines. For example, if a user's address was
3222 anne@some.where, the "personal" test would incorrectly be true for
3224 To: susanne@some.where
3226 This test is now done by extracting each address from the header in turn,
3227 and checking the entire address. Other tests that are part of "personal"
3228 are now done using regular expressions (for example, to check local parts
3229 of addresses in From: header lines).
3231 (2) The list of non-personal local parts in From: addresses has been
3232 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3233 taken from the Sieve specification recommendations.
3235 (3) If the message contains any header line starting with "List-" it is
3236 treated as non-personal.
3238 (4) The test for "circular" in the Subject: header line has been removed
3239 because it now seems ill-conceived.
3241 15. Minor typos in src/EDITME comments corrected.
3243 16. Installed latest exipick from John Jetmore.
3245 17. If headers_add on a router specified a text string that was too long for
3246 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3247 of string_sprintf() is now avoided.
3249 18. $message_body_size was not set (it was always zero) when running the DATA
3250 ACL and the local_scan() function.
3252 19. For the "mail" command in an Exim filter, no default was being set for
3253 the once_repeat time, causing a random time value to be used if "once" was
3254 specified. (If the value happened to be <= 0, no repeat happened.) The
3255 default is now 0s, meaning "never repeat". The "vacation" command was OK
3256 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3257 (I found it when inspecting the code).
3259 20. There is now an overall timeout for performing a callout verification. It
3260 defaults to 4 times the callout timeout, which applies to individual SMTP
3261 commands during the callout. The overall timeout applies when there is more
3262 than one host that can be tried. The timeout is checked before trying the
3263 next host. This prevents very long delays if there are a large number of
3264 hosts and all are timing out (e.g. when the network connections are timing
3265 out). The value of the overall timeout can be changed by specifying an
3266 additional sub-option for "callout", called "maxwait". For example:
3268 verify = sender/callout=5s,maxwait=20s
3270 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3271 to the end before writing, but this should make it even safer).
3273 22. Exim was forgetting that it had advertised PIPELINING for the second and
3274 subsequent messages on an SMTP connection. It was also not resetting its
3275 memory on STARTTLS and an internal HELO.
3277 23. When Exim logs an SMTP synchronization error within a session, it now
3278 records whether PIPELINING has been advertised or not.
3280 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3281 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3282 rather than long int.
3284 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3286 26. Added the never_mail option to autoreply.
3292 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3293 crash if the getsockname() call failed; this can happen if a connection is
3294 closed very soon after it is established. The problem was simply in the
3295 order in which certain operations were done, causing Exim to try to write
3296 to the SMTP stream before it had set up the file descriptor. The bug has
3297 been fixed by making things happen in the correct order.
3303 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3304 before the connection was closed, thus losing the rejection response.
3306 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3307 some early Solaris releases, but causes trouble in current releases where
3308 socklen_t is defined.
3310 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3313 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3316 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3317 that releasing the top of it at the end releases what was used for sub-
3318 expansions (unless the block got too big). However, discard this block if
3319 the first thing is a variable or header, so that we can use its block when
3320 it is dynamic (useful for very large $message_headers, for example).
3322 6. Lookups now cache *every* query, not just the most recent. A new, separate
3323 store pool is used for this. It can be recovered when all lookup caches are
3324 flushed. Lookups now release memory at the end of their result strings.
3325 This has involved some general refactoring of the lookup sources.
3327 7. Some code has been added to the store_xxx() functions to reduce the amount
3328 of flapping under certain conditions.
3330 8. log_incoming_interface used to affect only the <= reception log lines. Now
3331 it causes the local interface and port to be added to several more SMTP log
3332 lines, for example "SMTP connection from", and rejection lines.
3334 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3336 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3338 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3339 could be overwritten at the end of the current message (or the start of a
3340 new message if it was set in a HELO ACL). The value is now preserved for
3341 the duration of the SMTP connection.
3343 12. If a transport had a headers_rewrite setting, and a matching header line
3344 contained an unqualified address, that address was qualified, even if it
3345 did not match any rewriting rules. The underlying bug was that the values
3346 of the flags that permit the existence of unqualified sender and recipient
3347 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3348 non-local messages, and by -bnq for local messages) were not being
3349 preserved with the message after it was received.
3351 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3352 "next input=" as part of the text comprising the host identity instead of
3353 the correct text. The code was using the same buffer for two different
3354 strings. However, depending on which order the printing function evaluated
3355 its arguments, the bug did not always show up. Under Linux, for example, my
3356 test suite worked just fine.
3358 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3359 doesn't work with some older versions of Perl. It has been changed to "my",
3360 which in any case is probably the better facility to use.
3362 15. A really picky compiler found some instances of statements for creating
3363 error messages that either had too many or two few arguments for the format
3366 16. The size of the buffer for calls to the DNS resolver has been increased
3367 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3368 for addresses that have a lot of PTR records. This alleviates a problem; it
3369 does not fully solve it.
3371 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3372 buffer now truncates the list and logs the incident, which is the same
3373 action as happens when Exim is looking up a host name and its aliases.
3374 Previously in this situation something unpredictable would happen;
3375 sometimes it was "internal error: store_reset failed".
3377 18. If a server dropped the connection unexpectedly when an Exim client was
3378 using GnuTLS and trying to read a response, the client delivery process
3379 crashed while trying to generate an error log message.
3381 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3382 string, for example:
3384 warn message = H1: something\nH2: something
3386 the text was added as a single header line from Exim's point of view
3387 though it ended up OK in the delivered message. However, searching for the
3388 second and subsequent header lines using $h_h2: did not work. This has been
3389 fixed. Similarly, if a system filter added multiple headers in this way,
3390 the routers could not see them.
3392 20. Expanded the error message when iplsearch is called with an invalid key to
3393 suggest using net-iplsearch in a host list.
3395 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3396 ACLs are no longer actually imposed (and a message to that effect is
3399 22. If a "gecos" field in a passwd entry contained escaped characters, in
3400 particular, if it contained a \" sequence, Exim got it wrong when building
3401 a From: or a Sender: header from that name. A second bug also caused
3402 incorrect handling when an unquoted " was present following a character
3403 that needed quoting.
3405 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3406 was not being matched caselessly.
3408 24. Arranged for all hyphens in the exim.8 source to be escaped with
3411 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3412 properly. Recipient callout cache records were still being keyed to include
3413 the sender, even when use_sender was set false. This led to far more
3414 callouts that were necessary. The sender is no longer included in the key
3415 when use_sender is false.
3417 26. Added "control = submission" modifier to ACLs.
3419 27. Added the ${base62d: operator to decode base 62 numbers.
3421 28. dnsdb lookups can now access SRV records.
3423 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3424 the configuration file.
3426 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3427 (-v) mode. This makes the output for a verbose queue run more intelligible.
3429 31. Added a use_postmaster feature to recipient callouts.
3431 32. Added the $body_zerocount variable, containing the number of binary zero
3432 bytes in the message body.
3434 33. The time of last modification of the "new" subdirectory is now used as the
3435 "mailbox time last read" when there is a quota error for a maildir
3438 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3440 35. Added +ignore_unknown as a special item in host lists.
3442 36. Code for decoding IPv6 addresses in host lists is now included, even if
3443 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3444 address was recognized as an IP address, but was then not correctly decoded
3445 into binary, causing unexpected and incorrect effects when compared with
3452 1. Very minor rewording of debugging text in manualroute to say "list of
3453 hosts" instead of "hostlist".
3455 2. If verify=header_syntax was set, and a header line with an unqualified
3456 address (no domain) and a large number of spaces between the end of the
3457 name and the colon was received, the reception process suffered a buffer
3458 overflow, and (when I tested it) crashed. This was caused by some obsolete
3459 code that should have been removed. The fix is to remove it!
3461 3. When running in the test harness, delay a bit after writing a bounce
3462 message to get a bit more predictability in the log output.
3464 4. Added a call to search_tidyup() just before forking a reception process. In
3465 theory, someone could use a lookup in the expansion of smtp_accept_max_
3466 per_host which, without the tidyup, could leave open a database connection.
3468 5. Added the variables $recipient_data and $sender_data which get set from a
3469 lookup success in an ACL "recipients" or "senders" condition, or a router
3470 "senders" option, similar to $domain_data and $local_part_data.
3472 6. Moved the writing of debug_print from before to after the "senders" test
3475 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3476 problems for message scanning, either using a data ACL, or using
3477 local_scan() because the Received: header was not generated till after they
3478 were called (in order to set the time as the time of reception completion).
3479 I have revised the way this works. The header is now generated after the
3480 body is received, but before the ACL or local_scan() are called. After they
3481 are run, the timestamp in the header is updated.
3487 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3488 before starting a queue runner without re-exec. This happened only when
3489 deliver_drop_privilege was set or when the Exim user was set to root. The
3490 effect of the bug was that timeouts during subsequent deliveries caused
3491 crashes instead of being properly handled. The handler is now left at its
3492 default (and expected) setting.
3494 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3495 message, again when deliver_drop_privilege is set or Exim is run as root.
3496 The bug described in (1) was not present in this case, but the tidying up
3497 of the other signals was missing. I have made the two cases consistent.
3499 3. The ignore_target_hosts setting on a manualroute router was being ignored
3500 for hosts that were looked up using the /MX notation.
3502 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3505 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3506 operated on the sender address. After changing the $sender_address to <>
3507 for the sender address verify, Exim was re-instated it as the original
3508 (before rewriting) address, but remembering that it had rewritten it, so it
3509 wasn't rewriting it again. This bug also had the effect of breaking the
3510 sender address verification caching when the sender address was rewritten.
3512 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3513 This has been changed so that if the ip literal address matches
3514 ignore_target_hosts, the router declines.
3516 7. Added expansion conditions match_domain, match_address, and match_local_
3517 part (NOT match_host).
3519 8. The placeholder for the Received: header didn't have a length field set.
3521 9. Added code to Exim itself and to exim_lock to test for a specific race
3522 condition that could lead to file corruption when using MBX delivery. The
3523 issue is with the lockfile that is created in /tmp. If this file is removed
3524 after a process has opened it but before that process has acquired a lock,
3525 there is the potential for a second process to recreate the file and also
3526 acquire a lock. This could lead to two Exim processes writing to the file
3527 at the same time. The added code performs the same test as UW imapd; it
3528 checks after acquiring the lock that its file descriptor still refers to
3529 the same named file.
3531 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3532 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3535 11. Added the smtp_active_hostname option. If used, this will typically be made
3536 to depend on the incoming interface address. Because $interface_address is
3537 not set up until the daemon has forked a reception process, error responses
3538 that can happen earlier (such as "too many connections") no longer contain
3541 12. If an expansion in a condition on a "warn" statement fails because a lookup
3542 defers, the "warn" statement is abandoned, and the next ACL statement is
3543 processed. Previously this caused the whole ACL to be aborted.
3545 13. Added the iplsearch lookup type.
3547 14. Added ident_timeout as a log selector.
3549 15. Added tls_certificate_verified as a log selector.
3551 16. Added a global option tls_require_ciphers (compare the smtp transport
3552 option of the same name). This controls incoming TLS connections.
3554 17. I finally figured out how to make tls_require_ciphers do a similar thing
3555 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3556 before starting the TLS session.
3558 18. Tabs are now shown as \t in -bP output.
3560 19. If the log selector return_path_on_delivery was set, Exim crashed when
3561 bouncing a message because it had too many Received: header lines.
3563 20. If two routers both had headers_remove settings, and the first one included
3564 a superfluous trailing colon, the final name in the first list and the
3565 first name in the second list were incorrectly joined into one item (with a
3566 colon in the middle).
3572 1. Added -C and -D options to the exinext utility, mainly to make it easier
3573 to include in the automated testing, but these could be helpful when
3574 multiple configurations are in use.
3576 2. The exinext utility was not formatting the output nicely when there was
3577 an alternate port involved in the retry record key, nor when there was a
3578 message id as well (for retries that were specific to a specific message
3579 and a specific host). It was also confused by IPv6 addresses, because of
3580 the additional colons they contain. I have fixed the IPv4 problem, and
3581 patched it up to do a reasonable job for IPv6.
3583 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3584 delivery, the log line now contains "pipelined" if PIPELINING was used.
3586 4. An SMTP transport process used to panic and die if the bind() call to set
3587 an explicit outgoing interface failed. This has been changed; it is now
3588 treated in the same way as a connect() failure.
3590 5. A reference to $sender_host_name in the part of a conditional expansion
3591 that was being skipped was still causing a DNS lookup. This no longer
3594 6. The def: expansion condition was not recognizing references to header lines
3595 that used bh_ and bheader_.
3597 7. Added the _cache feature to named lists.
3599 8. The code for checking quota_filecount in the appendfile transport was
3600 allowing one more file than it should have been.
3602 9. For compatibility with Sendmail, the command line option
3610 and sets the incoming protocol and host name (for trusted callers). The
3611 host name and its colon can be omitted when only the protocol is to be set.
3612 Note the Exim already has two private options, -pd and -ps, that refer to
3613 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3614 "s", but I don't think that's a major issue.
3616 10. A number of refactoring changes to the code, none of which should affect
3619 (a) The number of logging options was getting close to filling up the
3620 32-bit word that was used as a bit map. I have split them into two classes:
3621 those that are passed in the argument to log_write(), and those that are
3622 only ever tested independently outside of that function. These are now in
3623 separate 32-bit words, so there is plenty of room for expansion again.
3624 There is no change in the user interface or the logging behaviour.
3626 (b) When building, for example, log lines, the code previously used a
3627 macro that called string_cat() twice, in order to add two strings. This is
3628 not really sufficiently general. Furthermore, there was one instance where
3629 it was actually wrong because one of the argument was used twice, and in
3630 one call a function was used. (As it happened, calling the function twice
3631 did not affect the overall behaviour.) The macro has been replaced by a
3632 function that can join an arbitrary number of extra strings onto a growing
3635 (c) The code for expansion conditions now uses a table and a binary chop
3636 instead of a serial search (which was left over from when there were very
3637 few conditions). Also, it now recognizes conditions like "pam" even when
3638 the relevant support is not compiled in: a suitably worded error message is
3639 given if an attempt is made to use such a condition.
3641 11. Added ${time_interval:xxxxx}.
3643 12. A bug was causing one of the ddress fields not to be passed back correctly
3644 from remote delivery subprocesses. The field in question was not being
3645 subsequently used, so this caused to problems in practice.
3647 13. Added new log selectors queue_time and deliver_time.
3649 14. Might have fixed a bug in maildirsizefile handling that threw up
3650 "unexpected character" debug warnings, and recalculated the data
3651 unnecessarily. In any case, I expanded the warning message to give more
3654 15. Added the message "Restricted characters in address" to the statements in
3655 the default ACL that block characters like @ and % in local parts.
3657 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3658 Three changes have been made:
3660 (a) There was a serious bug; a negative response to MAIL caused the whole
3661 recipient domain to be cached as invalid, thereby blocking all messages
3662 to all local parts at the same domain, from all senders. This bug has
3663 been fixed. The domain is no longer cached after a negative response to
3664 MAIL if the sender used is not empty.
3666 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3669 (c) A new callout option, "use_sender" has been added for people who want
3670 the modified behaviour.
3676 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3679 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3680 indeed breaks things for older releases.
3682 3. Added additional logging to the case where there is a problem reading data
3683 from a filter that is running in a subprocess using a pipe, in order to
3684 try to track down a specific problem.
3686 4. Testing facility fudge: when running in the test harness and attempting
3687 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3688 getting "No route to host". Convert this to a timeout.
3690 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3693 6. Some OS don't have socklen_t but use size_t instead. This affects the
3694 fifth argument of getsockopt() amongst other things. This is now
3695 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3696 can be set for individual OS. I have set it for SunOS5, OSF1, and
3697 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3698 some earlier ones do not.
3700 7. Change 4.30/15 was not doing the test caselessly.
3702 8. The standard form for an IPv6 address literal was being rejected by address
3703 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3704 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3705 this, as well as the form without the "IPv6" on the front (but only when
3706 address literals are enabled, of course).
3708 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3710 10. Exim crashed if a message with an empty sender address specified by -f
3711 encountered a router with an errors_to setting. This could be provoked only
3712 by a command such as
3716 where an empty string was supplied; "<>" did not hit this bug.
3718 11. Installed PCRE release 4.5.
3720 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3721 remained set. It is now erased.
3723 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3724 times from message ids (which are base 36 rather than the normal 62).
3726 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3727 were being counted as actual protocol errors, and logged if the log
3728 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3729 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3730 and DATA following a set of rejected RCPTs do not count as protocol errors.
3731 In other words, Exim assumes they were pipelined, though this may not
3732 actually be the case. Of course, in all cases the client gets an
3733 appropriate error code.
3735 15. If a lookup fails in an ACL condition, a message about the failure may
3736 be available; it is used if testing the ACL cannot continue, because most
3737 such messages specify what the cause of the deferral is. However, some
3738 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3739 that caused an old message to be retained and used if a later statement
3740 caused a defer, replacing the real cause of the deferral.
3742 16. If an IP address had so many PTR records that the DNS lookup buffer
3743 was not large enough to hold them, Exim could crash while trying to process
3744 the truncated data. It now detects and logs this case.
3746 17. Further to 4.21/58, another change has been made: if (and only if) the
3747 first line of a message (the first header line) ends with CRLF, a bare LF
3748 in a subsequent header line has a space inserted after it, so as not to
3749 terminate the header.
3751 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3752 unnecessarily, used atoi() instead of strtol(), and didn't check the
3753 termination when getting file sizes from file names by regex.
3755 19. Completely re-implemented the support for maildirsize files, in the light
3756 of a number of problems with the previous contributed implementation
3757 (4.30/29). In particular:
3759 . If the quota is zero, the maildirsize file is maintained, but no quota is
3762 . If the maildir directory does not exist, it is created before any attempt
3763 to write a maildirsize file.
3765 . The quota value in the file is just a cache; if the quota is changed in
3766 the transport, the new value overrides.
3768 . A regular expression is available for excluding directories from the
3771 20. The autoreply transport checks the characters in options that define the
3772 message's headers; it allows continued headers, but it was checking with
3773 isspace() after an embedded newline instead of explicitly looking for a
3776 21. If all the "regular" hosts to which an address was routed had passed their
3777 expiry times, and had not reached their retry times, the address was
3778 bounced, even if fallback hosts were defined. Now Exim should go on to try
3781 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3782 equivalent code in the SMTP transport. Some hosts send humungous responses
3783 to HELO/EHLO, more than 1024 it seems.
3785 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3786 gives compiler warnings in some environments. I've now done it "properly",
3789 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3790 (because of problems with the built-in one) was declared to return uschar *
3791 instead of char *, causing compiler failure.
3793 25. Fixed a file descriptor leak when processing alias/forward files.
3795 26. Fixed a minor format string issue in dbfn.c.
3797 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3799 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3800 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3803 29. When a sender address is verified, it is cached, to save repeating the test
3804 when there is more than one recipient in a message. However, when the
3805 verification involves a callout, it is possible for different callout
3806 options to be set for different recipients. It is too complicated to keep
3807 track of this in the cache, so now Exim always runs a verification when a
3808 callout is required, relying on the callout cache for the optimization.
3809 The overhead is duplication of the address routing, but this should not be
3812 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3813 to be verified with callout=postmaster, and the main callout worked but the
3814 postmaster check failed, the verification correctly failed. However, if a
3815 subsequent RCPT command asked for sender verification *without* the
3816 postmaster check, incorrect caching caused this verification also to fail,
3819 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3820 it was not caching the DNS options (qualify_single, search_parents) that
3821 were used when the lookup failed. A subsequent lookup with different
3822 options therefore always gave the same answer, though there were cases
3823 where it should not have. (Example: a "domains = !$mx_any" option on a
3824 dnslookup router: the "domains" option is always processed without any
3825 widening, but the router might have qualify_single set.) Now Exim uses the
3826 cached value only when the same options are set.
3828 32. Added John Jetmore's "exipick" utility to the distribution.
3830 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3831 than a timeout (e.g. a certificate is required, and is not provided), an
3832 Exim server now closes the connection immediately. Previously it waited for
3833 the client to close - but if the client is SSL, it seems that they each
3834 wait for each other, leading to a delay before one of them times out.
3836 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3837 maintained 0.8.x compatibility because I don't think many are using it, and
3838 it is clearly obsolete.
3840 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3843 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3844 client certificate was expired. A simple patch fixes this, though I don't
3845 understand the full logic of why the verify callback is called multiple
3848 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3849 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3850 which causes problems with some clients (such as the Certicom SSL Plus
3851 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3852 disables the coutermeasure allowing Eudora to connect."
3854 38. Exim was not checking that a write() to a log file succeeded. This could
3855 lead to Bad Things if a log got too big, in particular if it hit a file
3856 size limit. Exim now panics and dies if it cannot write to a log file, just
3857 as it does if it cannot open a log file.
3859 39. Modified OS/Makefile-Linux so that it now contains
3861 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3863 The two -D definitions ensure that Exim is compiled with large file
3864 support, which makes it possible to handle log files that are bigger than
3867 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3868 instance) a domain was checked against a named list that involved a lookup,
3869 causing $domain_data to be set, then another domain was checked against the
3870 same list, then the first domain was re-checked, the value of $domain_data
3871 after the final check could be wrong. In particular, if the second check
3872 failed, it could be set empty. This bug probably also applied to
3875 41. The strip_trailing_dot option was not being applied to the address given
3876 with the -f command-line option.
3878 42. The code for reading a message's header from the spool was incrementing
3879 $received_count, but never initializing it. This meant that the value was
3880 incorrect (doubled) while delivering a message in the same process in which
3881 it was received. In the most common configuration of Exim, this never
3882 happens - a fresh exec is done - but it can happen when
3883 deliver_drop_privilege is set.
3885 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3886 it now includes up to 150 characters of the unexpected data in the log
3889 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3890 and building data strings. The size of both of these buffers was 10 000
3891 bytes - far larger than anybody would *ever* want, thought I. Needless to
3892 say, somebody hit the limit. I have increased the maximum line length to
3893 20 000 and the maximum data length of concatenated lines to 100 000. I have
3894 also fixed two bugs, because there was no checking on these buffers. Tsk,
3895 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3896 buffer is too small.
3898 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3899 lsearch lookups. Now it does.
3901 46. When parsing a route_list item in a manualroute router, a fixed-length
3902 buffer was used for the list of hosts. I made this 1024 bytes long,
3903 thinking that nobody would ever have a list of hosts that long. Wrong.
3904 Somebody had a whole pile of complicated expansion conditions, and the
3905 string was silently truncated, leading to an expansion error. It turns out
3906 that it is easier to change to an unlimited length (owing to other changes
3907 that have happened since this code was originally written) than to build
3908 structure for giving a limitation error. The length of the item that
3909 expands into the list of hosts is now unlimited.
3911 47. The lsearch lookup could not handle data where the length of text line was
3912 more than 4095 characters. Such lines were truncated, leading to shortened
3913 data being returned. It should now handle lines of any length.
3915 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3916 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3919 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3920 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3921 no longer happens in comment lines. A list of replacements is now placed
3922 at the head of all of the source files, except those whose only change is
3923 to replace PERL_COMMAND in the very first #! line.
3925 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3926 messages on the queue, with a bottom-up merge sort, using code contributed
3927 by Michael Haardt. This should make operations like -bp somewhat faster on
3928 large queues. It won't affect queue runners, except when queue_run_in_order
3931 51. Installed eximstats 1.31 in the distribution.
3933 52. Added support for SRV lookups to the dnslookup router.
3935 53. If an ACL referred to $message_body or $message_body_end, the value was not
3936 reset for any messages that followed in the same SMTP session.
3938 54. The store-handling optimization for building very long strings was not
3939 differentiating between the different store pools. I don't think this
3940 actually made any difference in practice, but I've tidied it.
3942 55. While running the routers to verify a sender address, $sender_address
3943 was still set to the sender address. This is wrong, because when routing to
3944 send a bounce to the sender, it would be empty. Therefore, I have changed
3945 it so that, while verifying a sender address, $sender_address is set to <>.
3946 (There is no change to what happens when verifying a recipient address.)
3948 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3949 target A or AAAA records (if not already returned) without resetting the
3950 qualify_single or search_parents options of the DNS resolver. These are
3951 inappropriate in this case because the targets of MX and SRV records must
3952 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3953 target that, when qualified, matched something in the local domain. These
3954 two options are now turned off when doing these lookups.
3956 57. It seems that at least some releases of Reiserfs (which does not have the
3957 concept of a fixed number of inodes) returns zero and not -1 for the
3958 number of available inodes. This interacted badly with check_spool_inodes,
3959 which assumed that -1 was the "no such thing" setting. What I have done is
3960 to check that the total number of inodes is greater than zero before doing
3961 the test of how many are available.
3963 58. When a "warn" ACL statement has a log_message modifier, the message is
3964 remembered, and not repeated. This is to avoid a lot of repetition when a
3965 message has many recipients that cause the same warning to be written.
3966 Howewer, Exim was preserving the list of already written lines for an
3967 entire SMTP session, which doesn't seem right. The memory is now reset if a
3968 new message is started.
3970 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3971 debugging output unless log_rewrite was also set.
3973 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3974 of (int)(handle) when we know that handle contains (void *)(-1).
3976 61. The Exim daemon panic-logs an error return when it closes the incoming
3977 connection. However "connection reset by peer" seems to be common, and
3978 isn't really an error worthy of noting specially, so that particular error
3981 62. When Exim is trying to find all the local interfaces, it used to panic and
3982 die if the ioctl to get the interface flags failed. However, it seems that
3983 on at least one OS (Solaris 9) it is possible to have an interface that is
3984 included in the list of interfaces, but for which you get a failure error
3985 for this call. This happens when the interface is not "plumbed" into a
3986 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3987 failure of the "get flags" call assumes that the interface is down.
3989 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3990 makes life easier for people who are doing arithmetic on fields extracted
3991 from dates, where you often get leading zeros that should not be
3992 interpreted as octal.
3994 64. Added qualify_domain to the redirect router, to override the global
3997 65. If a pathologically long header line contained very many addresses (the
3998 report of this problem mentioned 10 000) and each of them was rewritten,
3999 Exim could use up a very large amount of memory. (It kept on making new
4000 copies of the header line as it rewrote, and never released the old ones.)
4001 At the expense of a bit more processing, the header rewriting function has
4002 been changed so that it no longer eats memory in this way.
4004 66. The generation of the Received: header has been moved from the time that a
4005 message starts to be received, to the time that it finishes. The timestamp
4006 in the Received: header should now be very close to that of the <= log
4007 line. There are two side-effects of this change:
4009 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4010 logged header lines no longer include the local Received: line, because
4011 it has not yet been created. The same applies to a copy of the message
4012 that is returned to a non-SMTP sender when a message is rejected.
4014 (b) When a filter file is tested using -bf, no additional Received: header
4015 is added to the test message. After some thought, I decided that this
4018 This change does not affect the value of $received_for. It is still set
4019 after address rewriting, but before local_scan() is called.
4021 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4023 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4024 gave an unhelpful panic error message, and a defer error. I have managed to
4025 change this behaviour so that it now rejects any supplied certificate,
4026 which seems right, as the list of acceptable certificates is empty.
4028 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4029 gave an unhelpful defer error. I have not managed to make this reject any
4030 supplied certificates, but the error message it gives is "no certificate
4031 supplied", which is not helpful.
4033 70. exigrep's output now also includes lines that are not associated with any
4034 message, but which match the given pattern. Implemented by a patch from
4035 Martin Sluka, which also tidied up the Perl a bit.
4037 71. Recipient callout verification, like sender verification, was using <> in
4038 the MAIL FROM command. This isn't really the right thing, since the actual
4039 sender may affect whether the remote host accepts the recipient or not. I
4040 have changed it to use the actual sender in the callout; this means that
4041 the cache record is now keyed on a recipient/sender pair, not just the
4042 recipient address. There doesn't seem to be a real danger of callout loops,
4043 since a callout by the remote host to check the sender would use <>.
4044 [SEE ABOVE: changed after hitting problems.]
4046 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4047 temporary errors. However, in the case of such a code being given after
4048 the end of a data transmission (i.e. after ".") Exim was failing to write
4049 a retry record for the message. (Yes, there was some broken host that was
4050 actually sending 8xx at this point.)
4052 73. An unknown lookup type in a host list could cause Exim to panic-die when
4053 the list was checked. (An example that provoked this was putting <; in the
4054 middle of a list instead of at the start.) If this happened during a DATA
4055 ACL check, a -D file could be left lying around. This kind of configuration
4056 error no longer causes Exim to die; instead it causes a defer errror. The
4057 incident is still logged to the main and panic logs.
4059 74. Buglet left over from Exim 3 conversion. The message "too many messages
4060 in one connection" was written to the rejectlog but not the mainlog, except
4061 when address rewriting (yes!) was being logged.
4063 75. Added write_rejectlog option.
4065 76. When a system filter was run not as root (that is, when system_filter_user
4066 was set), the values of the $n variables were not being returned to the
4067 main process; thus, they were not subsequently available in the $sn
4070 77. Added +return_path_on_delivery log selector.
4072 78. A connection timeout was being treated differently from recipients deferred
4073 when testing hosts_max_try with a message that was older than the host's
4074 retry timeout. (The host should not be counted, thus allowing all hosts to
4075 be tried at least once before bouncing.) This may have been the cause of an
4076 occasionally reported bug whereby a message would remain on the queue
4077 longer than the retry timeout, but would be bounced if a delivery was
4078 forced. I say "may" because I never totally pinned down the problem;
4079 setting up timeout/retry tests is difficult. See also the next item.
4081 79. The ultimate address timeout was not being applied to errors that involved
4082 a combination of host plus message (for example, a timeout on a MAIL
4083 command). When an address resolved to a number of possible hosts, and they
4084 were not all tried for each delivery (e.g. because of hosts_max_try), a
4085 message could remain on the queue longer than the retry timeout.
4087 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4090 81. Fixed an obscure SMTP outgoing bug which required at least the following
4091 conditions: (a) there was another message waiting for the same server;
4092 (b) the server returned 5xx to all RCPT commands in the first message so
4093 that the message was not completed; (c) the server dropped the connection
4094 or gave a negative response to the RSET that Exim sends to abort the
4095 transaction. The observed case was a dropped connection after DATA that had
4096 been sent in pipelining mode. That is, the server had advertised PIPELINING
4097 but was not implementing it correctly. The effect of the bug was incorrect
4098 behaviour, such as trying another host, and this could lead to a crash.
4104 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4105 and daemon.c were passed as pointers to ints; they should have been
4106 pointers to socklen_t variables (which are typically unsigned ints).
4108 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4111 3. Fixed a really odd bug that affected only the testing scheme; patching a
4112 certain fixed string in the binary changed the value of another string that
4113 happened to be identical to the end of the original first string.
4115 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4116 name", it returns that address as the IP address. On some operating
4117 systems (e.g. Solaris), it also passes back the IP address string as the
4118 "host name". However, on others (e.g. Linux), it passes back an empty
4119 string. Exim wasn't checking for this, and was changing the host name to an
4120 empty string, assuming it had been canonicized.
4122 5. Although rare, it is permitted to have more than one PTR record for a given
4123 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4124 all the names associated with an address, because they do in Solaris.
4125 However, it seems that they do not in Linux for data that comes from the
4126 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4127 I found this out when I moved to a new Linux workstation and tried to run
4128 the Exim test suite.
4130 To get round this problem I have changed the code so that it now does its
4131 own call to the DNS to look up PTR records when searching for a host name.
4132 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4133 addresses that are only in /etc/hosts are still found.
4135 This behaviour is, however, controlled by an option called host_lookup_
4136 order, which defaults to "bydns:byaddr". If people want to use the other
4137 order, or indeed, just use one or the other means of lookup, they can
4138 specify it in this variable.
4140 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4141 some operating systems, this comes back from gethostbyaddr() as an empty
4142 string, and this is what Exim used to test for. However, it seems that in
4143 other systems, "." is yielded. Exim now tests for this case too.
4145 7. The values of check_spool_space and check_log_space are now held internally
4146 as a number of kilobytes instead of an absolute number of bytes. If a
4147 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4148 kilobyte. This means that much larger values can be stored.
4150 8. Exim monitor: an attempt to get the action menu when not actually pointing
4151 at a message produces an empty menu entitled "No message selected". This
4152 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4153 no entries in it ("Shell widget menu has zero width and/or height"). So I
4154 have added a single, blank menu entry in this case.
4156 9. Added ${quote_local_part.
4158 10. MIME decoding is now applied to the contents of Subject: header lines when
4161 11. Now that a reference to $sender_host_address automatically causes a reverse
4162 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4163 host lookup before query-style lookups in lists that might use this
4164 variable. This has therefore been abolished, and the "net-" prefix is no
4165 longer necessary for query-style lookups.
4167 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4168 have been a typo for LFLAGS, so it has been changed.
4170 13. The install script calls Exim with "-C /dev/null" in order to find the
4171 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4172 to be output. Howeve, since Exim outputs its version number before the
4173 error, it didn't break the script. It just looked ugly. I fixed this by
4174 always allowing "-C /dev/null" if the caller is root.
4176 14. Ignore overlarge ACL variable number when reading spool file - insurance
4177 against a later release with more variables having written the file.
4179 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4180 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4181 this, as well as the form without the "IPv6" on the front.
4183 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4184 OS/Makefile-Darwin file.
4186 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4187 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4189 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4190 variable. However, some OS use ints for inodes. I've added cast to long int
4191 to get rid of the compiler warning.
4193 19. I had forgotten to lock out "/../" in configuration file names when
4194 ALT_CONFIG_PREFIX was set.
4196 20. Routers used for verification do not need to specify transports. However,
4197 if such a router generated a host list, and callout was configured, Exim
4198 crashed, because it could not find a port number from the (non-existent)
4199 transport. It now assumes port 25 in this circumstance.
4201 21. Added the -t option to exigrep.
4203 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4204 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4205 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4206 them is set, LOOKUP_LSEARCH is forced.
4208 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4210 24. Added sender and host information to the "rejected by local_scan()" log
4211 line; previously there was no indication of these.
4213 25. Added .include_if_exists.
4215 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4216 Linux. It turns out that not all file systems support this. Apparently some
4217 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4218 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4219 not supported on the file descriptor, is now ignored when Exim is trying to
4220 sync a directory. This applies only to Linux.
4222 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4224 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4225 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4228 29. Added support for maildirsize files from supplied patch (modified a bit).
4230 30. The use of :fail: followed by an empty string could lead Exim to respond to
4231 sender verification failures with (e.g.):
4233 550 Verification failed for <xxx>
4234 550 Sender verify failed
4236 where the first response line was missing the '-' that indicates it is not
4237 the final line of the response.
4239 31. The loop for finding the name of the user that called Exim had a hardwired
4240 limit of 10; it now uses the value of finduser_retries, which is used for
4241 all other user lookups.
4243 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4246 33. Exim was neglecting to zero errno before one call of strtol() when
4247 expanding a string and expecting an integer value. On some systems this
4248 resulted in spurious "integer overflow" errors. Also, it was casting the
4249 result into an int without checking.
4251 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4252 did not work. The code looks as if it has *never* worked, though it appears
4253 to have been documented since at least releast 1.62. I have made it work.
4255 35. The "timeout_DNS" error in retry rules, also documented since at least
4256 1.62, also never worked. As it isn't clear exactly what this means, and
4257 clearly it isn't a major issue, I have abolished the feature by treating it
4258 as "timeout", and writing a warning to the main and panic logs.
4260 36. The display of retry rules for -brt wasn't always showing the error code
4263 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4264 timeout_connect_A, timeout_connect_MX.
4266 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4267 to the empty sender.
4269 39. The daemon was not analysing the content of -oX till after it had closed
4270 stderr and disconnected from the controlling terminal. This meant that any
4271 syntax errors were only noted on the panic log, and the return code from
4272 the command was 0. By re-arranging the code a little, I've made the
4273 decoding happen first, so such errors now appear on stderr, and the return
4274 code is 1. However, the actual setting up of the sockets still happens in
4275 the disconnected process, so errors there are still only recorded on the
4278 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4279 connections (as happens on some IP stacks) was logged at start up time as
4280 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4281 it from "IPv6 and IPv4", which means that two separate sockets are being
4284 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4285 says whether AF_INET or AF_INET6 was passed as an argument.
4287 42. Exiwhat output was messed up when time zones were included in log
4290 43. Exiwhat now gives more information about the daemon's listening ports,
4291 and whether -tls-on-connect was used.
4293 44. The "port" option of the smtp transport is now expanded.
4295 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4296 silently ignored. Now an error message is written to the main and panic
4299 46. There's a new ACL modifier called "logwrite" which writes to a log file
4300 as soon as it is encountered.
4302 47. Added $local_user_uid and $local_user_gid at routing time.
4304 48. Exim crashed when trying to verify a sender address that was being
4307 49. Exim was recognizing only a space character after ".include". It now also
4308 recognizes a tab character.
4310 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4311 extracting the relevant information from the specification. The man page no
4312 longer contains scrambled data for the -d option, and I've added a section
4313 at the front about calling Exim under different names.
4315 51. Added "extra_headers" argument to the "mail" command in filter files.
4317 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4320 53. Installed eximstats 1.29.
4322 54. Added transport_filter_timeout as a generic transport option.
4324 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4325 Cc: header lines. This was required by RFC 822, but it not required by RFC
4328 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4329 incoming messages that did not have them. Now it does so only if the
4330 message originates locally, that is, if there is no associated remote host
4331 address. When Resent- header lines are present, this applies to the Resent-
4332 lines rather than the non-Resent- lines.
4334 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4335 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4337 58. Messages for configuration errors now include the name of the main
4338 configuration file - useful now that there may be more than one file in a
4339 list (.included file names were always shown).
4341 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4342 for those rare installations that do not start the daemon as root or run it
4343 setuid root. I've cut out the call to initgroups() if the daemon is not
4346 60. The Exim user and group can now be bound into the binary as text strings
4347 that are looked up at the start of Exim's processing.
4349 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4351 62. Added $mailstore_basename variable.
4353 63. Installed patch to sieve.c from Michael Haardt.
4355 64. When Exim failed to open the panic log after failing to open the main log,
4356 the original message it was trying to log was written to stderr and debug
4357 output, but if they were not available (the usual case in production), it
4358 was lost. Now it is written to syslog before the two lines that record the
4359 failures to open the logs.
4361 65. Users' Exim filters run in subprocesses under the user's uid. It is
4362 possible for a "deliver" command or an alias in a "personal" command to
4363 provoke an address rewrite. If logging of address rewriting is configured,
4364 this fails because the process is not running as root or exim. There may be
4365 a better way of dealing with this, but for the moment (because 4.30 needs
4366 to be released), I have disabled address rewrite logging when running a
4367 filter in a non-root, non-exim process.
4373 1. The buildconfig auxiliary program wasn't quoting the value set for
4374 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4375 not defined. This bug was masked in 4.22 by the effect that was fixed in
4378 2. Some messages that were rejected after a message id was allocated were
4379 shown as "incomplete" by exigrep. It no longer does this for messages that
4380 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4382 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4383 have allow_domain_literals set, the ID did not get logged in the <= line.
4384 Domain literals are now always recognized in Message-ID: header lines.
4386 4. The first argument for a ${extract expansion item is the key name or field
4387 number. Leading and trailing spaces in this item were not being ignored,
4388 causing some misleading effects.
4390 5. When deliver_drop_privilege was set, single queue runner processes started
4391 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4392 same command in the process it spins off) were not dropping privilege.
4394 6. When the daemon running as "exim" started a queue runner, it always
4395 re-executed Exim in the spun-off process. This is a waste of effort when
4396 deliver_drop_privilege is set. The new process now just calls the
4397 queue-runner function directly.
4403 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4406 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4407 ignored. Though the use of -oP was forcing the writing of a pid file, it
4408 was always written to the default place.
4410 3. If the message "no IP address found for host xxxx" is generated during
4411 incoming verification, it is now followed by identification of the incoming
4412 connection (so you can more easily find what provoked it).
4414 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4416 5. Added some features to "harden" Exim a bit more against certain attacks:
4418 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4419 be put in Local/Makefile. This is like the never_users runtime option,
4420 but it cannot be overridden. The default setting is "root".
4422 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4423 prefix string with which any file named in a -C command line option
4426 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4427 is retained for -C and -D only if the caller of Exim is root. Without
4428 it, the exim user may also use -C and -D and retain privilege.
4430 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4431 command line option is disabled.
4433 6. Macro names set by the -D option must start with an upper case letter, just
4434 like macro names defined in the configuration file.
4436 7. Added "dereference=" facility to LDAP.
4438 8. Two instances of the typo "uknown" in the source files are fixed.
4440 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4441 the Configure-Makefile script screwed up while processing it.
4443 10. Incorporated PCRE 4.4.
4445 11. The SMTP synchronization check was not operating right at the start of an
4446 SMTP session. For example, it could not catch a HELO sent before the client
4447 waited for the greeting. There is now a check for outstanding input at the
4448 point when the greeting is written. Because of the duplex, asynchronous
4449 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4450 way, but not yet received, when the check is performed.
4452 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4453 on TCP/IP sockets, because this apparently causes some broken clients to
4456 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4457 unchanged) from the Cygwin maintainer.
4459 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4460 was supported instead of testing for mbx. Effectively a typo.
4462 15. The spa authenticator server code was not checking that the input it
4463 received was valid base64.
4465 16. The debug output line for the "set" modifier in ACLs was not showing the
4466 name of the variable that was being set.
4468 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4470 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4471 lookup on its own. Something else was needed to trigger the lookup. For
4472 example, a match in host_lookup or the need for a host name in a host list.
4473 Now, if $sender_host_name is referenced and the host name has not yet been
4474 looked up, a lookup is performed. If the lookup fails, the variable remains
4475 empty, and $host_lookup_failed is set to "1".
4477 19. Added "eqi" as a case-independent comparison operator.
4479 20. The saslauthd authentication condition could segfault if neither service
4480 nor realm was specified.
4482 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4483 error message that was logged was misleading, and incoming SMTP
4484 connections were dropped. The message is now more accurate, and temporary
4485 errors are given to SMTP connections.
4487 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4488 (except RSET) with a 421 error, until QUIT is received. However, it was
4489 failing to send a response to QUIT.
4491 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4492 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4493 ACL is now run after the helo_try_verify_hosts code.
4495 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4496 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4497 case-independent, so other case variants are also recognized). Apparently
4498 some systems use these upper case variants.
4500 25. If more than two messages were waiting for the same host, and a transport
4501 filter was specified for the transport, Exim sent two messages over the
4502 same TCP/IP connection, and then failed with "socket operation on non-
4503 socket" when it tried to send the third.
4505 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4507 27. The extern definition of crypt16() in expand.c was not being excluded when
4508 the OS had its own crypt16() function.
4510 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4511 as a preferred synonym for return_size_limit, both as an option and as an
4514 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4516 30. Changed the default configuration ACL to relax the local part checking rule
4517 for addresses that are not in any local domains. For these addresses,
4518 slashes and pipe symbols are allowed within local parts, but the sequence
4519 /../ is explicitly forbidden.
4521 31. SPA server authentication was not clearing the challenge buffer before
4524 32. log_message in a "warn" ACL statement was writing to the reject log as
4525 well as to the main log, which contradicts the documentation and doesn't
4526 seem right (because no rejection is happening). So I have stopped it.
4528 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4529 However, I am unable to do any testing of this.
4531 34. Fixed an infelicity in the appendfile transport. When checking directories
4532 for a mailbox, to see if any needed to be created, it was accidentally
4533 using path names with one or more superfluous leading slashes; tracing
4534 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4536 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4537 contents are added to the log line that is written for every discarded
4538 recipient. (Previously a log_message setting was ignored.)
4540 36. The ${quote: operator now quotes the string if it is empty.
4542 37. The install script runs exim in order to find its version number. If for
4543 some reason other than non-existence or emptiness, which it checks, it
4544 could not run './exim', it was installing it with an empty version number,
4545 i.e. as "exim-". This error state is now caught, and the installation is
4548 38. An argument was missing from the function that creates an error message
4549 when Exim fails to connect to the socket for saslauthd authentication.
4550 This could cause Exim to crash, or give a corrupted message.
4552 39. Added isip, isip4, and isip6 to ${if conditions.
4554 40. The ACL variables $acl_xx are now saved with the message, and can be
4555 accessed later in routers, transports, and filters.
4557 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4558 strings in the file are not string-expanded.
4560 42. If a MAIL command specified a SIZE value that was too large to fit into an
4561 int variable, the check against message_size_limit failed. Such values are
4562 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4563 day this will have to be increased, but I don't think I want to be around
4564 when emails are that large.
4571 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4572 iconv() is not standard in FreeBSD.
4574 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4575 IPv6 enabled. The observed symptom was a segmentation fault on return from
4576 the function os_common_find_running_interfaces() in src/os.c.
4578 3. In the check_special_case() function in daemon.c I had used "errno" as an
4579 argument name, which causes warnings on some systems. This was basically a
4580 typo, since it was named "eno" in the comments!
4582 4. The code that waits for the clock to tick (at a resolution of some fraction
4583 of a second) so as to ensure message-id uniqueness was always waiting for
4584 at least one whole tick, when it could have waited for less. [This is
4585 almost certainly not relevant at current processor speeds, where it is
4586 unlikely to ever wait at all. But we try to future-proof.]
4588 5. The function that sleeps for a time interval that includes fractions of a
4589 second contained a race. It did not block SIGALRM between setting the
4590 timer, and suspending (a couple of lines later). If the interval was short
4591 and the sigsuspend() was delayed until after it had expired, the suspension
4592 never ended. On busy systems this could lead to processes getting stuck for
4595 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4596 process, before it forks any delivery processes. The open lookup caching
4597 mechanism meant that the open file or database connection was passed into
4598 the delivery process. The problem was that delivery processes always tidy
4599 up cached lookup data. This could cause a problem for the next delivery
4600 process started by the queue runner, because the external queue runner
4601 process does not know about the closure. So the next delivery process
4602 still has data in the lookup cache. In the case of a file lookup, there was
4603 no problem because closing a file descriptor in a subprocess doesn't affect
4604 the parent. However, if the lookup was caching a connection to a database,
4605 the connection was closed, and the second delivery process was likely to
4606 see errors such as "PGSQL: query failed: server closed the connection
4607 unexpectedly". The problem has been fixed by closing all cached lookups
4608 in a queue runner before running a delivery process.
4610 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4611 seem to have the "const" qualifier which it has on other OS. I've
4614 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4615 *of the same type* (client or server) with the same public name that an
4616 error should be diagnosed.
4618 9. When Exim looked up a host name for an IP address, but failed to find the
4619 original IP address when looking up the host name (a safety check), it
4620 output the message "<ip address> does not match any IP for NULL", which was
4621 confusing, to say the least. The bug was that the host name should have
4622 appeared instead of "NULL".
4624 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4625 user that is built into the binary, and the -C or -D options is used, root
4626 privilege is dropped before the configuration file is read. In addition,
4627 logging is switched to stderr instead of the normal log files. If the
4628 configuration then re-defines the Exim user, the unprivileged environment
4629 is probably not what is expected, so Exim logs a panic warning message (but
4632 However, if deliver_drop_privilege is set, the unprivileged state may well
4633 be exactly what is intended, so the warning has been cut out in that case,
4634 and Exim is allowed to try to write to its normal log files.
4640 1. smtp_return_error_details was not giving details for temporary sender
4641 or receiver verification errors.
4643 2. Diagnose a configuration error if two authenticators have the same public
4646 3. Exim used not to create the message log file for a message until the first
4647 delivery attempt. This could be confusing when incoming messages were held
4648 for policy or load reasons. The message log file is now created at the time
4649 the message is received, and an initial "Received" line is written to it.
4651 4. The automatically generated man page for command line options had a minor
4652 bug that caused no ill effects; however, a more serious problem was that
4653 the procedure for building the man page automatically didn't always
4654 operate. Consequently, release 4.20 contains an out-of-date version. This
4655 shouldn't happen again.
4657 5. When building Exim with embedded Perl support, the script that builds the
4658 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4659 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4661 6. The freeze_tell option was not being used for messages that were frozen on
4662 arrival, either by an ACL or by local_scan().
4664 7. Added the smtp_incomplete_transaction log selector.
4666 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4667 was accepting AUTH without a new EHLO.
4669 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4670 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4673 10. Logging of TCP/IP connections (when configured) now happens in the main
4674 daemon process instead of the child process, so that the TCP/IP connection
4675 count is more accurate (but it can never be perfect).
4677 11. The use of "drop" in a nested ACL was not being handled correctly in the
4678 outer ACL. Now, if condition failure induced by the nested "drop" causes
4679 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4680 or "require"), the connection is dropped.
4682 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4683 that yield "discard" can now be used with an "accept" or a "discard" verb,
4684 but an error is generated for any others (because I can't see a useful way
4685 to define what should happen).
4687 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4688 are now processed in the same way as lines in the Exim configuration file.
4689 In particular, continuation lines are supported.
4691 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4693 15. Added -ti meaning -t -i.
4695 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4696 domains, and warn by logging if others are found.
4698 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4699 returned from GIFCONF when Exim is trying to find the list of interfaces on
4700 a host. The code in os.c has been modified to copy each ifreq to an aligned
4701 structure in all cases.
4703 Also, in some cases, the returned ifreq's were being copied to a 'struct
4704 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4705 means the last couple of bytes of an IPv6 address could be chopped if the
4706 ifreq contained only a normal sockaddr (14 bytes storage).
4708 18. Named domain lists were not supported in the hosts_treat_as_local option.
4709 An entry such as +xxxx was not recognized, and was treated as a literal
4712 19. Ensure that header lines added by a DATA ACL are included in the reject log
4713 if the ACL subsequently rejects the message.
4715 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4716 MD5 (which is deprecated).
4718 21. When testing a filter file using -bf, Exim was writing a message when it
4719 took the sender from a "From " line in the message, but it was not doing so
4720 when it took $return_path from a Return-Path: header line. It now does.
4722 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4723 with a valid header line field name (a series of printing characters
4724 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4726 23. Changed "disc" in the source to "disk" to conform to the documentation and
4727 the book and for uniformity.
4729 24. Ignore Sendmail's -Ooption=value command line item.
4731 25. When execve() failed while trying to run a command in a pipe transport,
4732 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4733 could be confused with a return value of 69 from the command itself. This
4734 has been changed to 127, the value the shell returns if it is asked to run
4735 a non-existent command. The wording for the related log line suggests a
4736 non-existent command as the problem.
4738 26. If received_header_text expands to an empty string, do not add a Received:
4739 header line to the message. (Well, it adds a token one on the spool, but
4740 marks it "old" so that it doesn't get used or transmitted.)
4742 27. Installed eximstats 1.28 (addition of -nt option).
4744 28. There was no check for failure on the call to getsockname() in the daemon
4745 code. This can fail if there is a shortage of resources on the system, with
4746 ENOMEM, for example. A temporary error is now given on failure.
4748 29. Contrary to the C standard, it seems that in some environments, the
4749 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4750 program. Exim now does this explicitly; it affects the formatting of
4751 timestamps using strftime().
4753 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4754 complaints. I've now initialized all the variables, to avoid this.
4756 32. Header lines added by a system filter were not being "seen" during
4757 transport-time rewrites.
4759 33. The info_callback() function passed to OpenSSL is set up with type void
4760 (*)(SSL *, int, int), as described somewhere. However, when calling the
4761 function (actually a macro) that sets it up, the type void(*)() is
4762 expected. I've put in a cast to prevent warnings from picky compilers.
4764 34. If a DNS black list lookup found a CNAME record, but there were no A
4765 records associated with the domain it pointed at, Exim crashed.
4767 35. If a DNS black list lookup returned more than one A record, Exim ignored
4768 all but the first. It now scans all returned addresses if a particular IP
4769 value is being sought. In this situation, the contents of the
4770 $dnslist_value variable are a list of all the addresses, separated by a
4773 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4774 to accept a host name and all its aliases if the forward lookup for any of
4775 them yielded the IP address of the incoming connection. Now it accepts only
4776 those names whose forward lookup yields the correct IP address. Any other
4777 names are discarded. This closes a loophole whereby a rogue DNS
4778 administrator could create reverse DNS records to break through a
4779 wildcarded host restriction in an ACL.
4781 37. If a user filter or a system filter that ran in a subprocess used any of
4782 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4783 the wrong values were passed to the pipe command ($thisaddress had the
4784 value of $0, $0 had the value of $1, etc). This bug was introduced by
4785 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4787 38. Improved the line breaking for long SMTP error messages from ACLs.
4788 Previously, if there was no break point between 40 and 75 characters, Exim
4789 left the rest of the message alone. Two changes have been made: (a) I've
4790 reduced the minimum length to 35 characters; (b) if it can't find a break
4791 point between 35 and 75 characters, it looks ahead and uses the first one
4792 that it finds. This may give the occasional overlong line, but at least the
4793 remaining text gets split now.
4795 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4796 file descriptors might be low, and that setting 1000 would always raise it.
4797 It turns out that in some environments, the limit is already over 1000 and
4798 that lowering it causes trouble. So now Exim takes care not to decrease it.
4800 40. When delivering a message, the value of $return_path is set to $sender_
4801 address at the start of routing (routers may change the value). By an
4802 oversight, this default was not being set up when an address was tested by
4803 -bt or -bv, which affected the outcome if any router or filter referred to
4806 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4807 log only when "message" or "log_message" are set. However, if one of the
4808 conditions was an address verification, or a call to a nested ACL, the
4809 messages generated by the underlying test were being passed through. This
4810 no longer happens. The underlying message is available in $acl_verify_
4811 message for both "message" and "log_message" expansions, so it can be
4812 passed through if needed.
4814 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4815 new expansion $bh_ to give the encoded byte string without charset
4816 translation. Translation happens only if iconv() is available; HAVE_ICONV
4817 indicates this at build time. HEADERS_CHARSET gives the charset to
4818 translate to; headers_charset can change it in the configuration, and
4819 "headers charset" can change it in an individual filter file.
4821 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4822 that creates RFC 2047 encoded "words" labels them as that charset instead
4823 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4824 expansion operator; (ii) when Exim creates a From: line for a local
4825 message; (iii) when a header line is rewritten to include a "phrase" part.
4827 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4828 buggy, causing it to skip the first lines of messages whose message ID
4829 ended in 'D'. This would not have bitten before Exim release 4.14, because
4830 message IDs were unlikely to end in 'D' before then. The effect was to have
4831 incorrect size information for certain domains.
4833 45. #include "config.h" was missing at the start of the crypt16.c module. This
4834 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4837 46. If there was a timeout during a "random" callout check, Exim treated it as
4838 a failure of the random address, and carried on sending RSET and the real
4839 address. If the delay was just some slowness somewhere, the response to the
4840 original RCPT would be taken as a response to RSET and so on, causing
4841 mayhem of various kinds.
4843 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4844 when I implemented it. It didn't allow for the fact that some option values
4845 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4846 the right test for positive values.
4848 48. Domain names in DNS records are case-independent. Exim always looks them up
4849 in lower case. Some resolvers return domain names in exactly the case they
4850 appear in the zone file, that is, they may contain uppercase letters. Not
4851 all resolvers do this - some return always lower case. Exim was treating a
4852 change of case by a resolver as a change of domain, similar to a widening
4853 of a domain abbreviation. This triggered its re-routing code and so it was
4854 trying to route what was effectively the same domain again. This normally
4855 caused routing to fail (because the router wouldn't handle the domain
4856 twice). Now Exim checks for this case specially, and just changes the
4857 casing of the domain that it ultimately uses when it transmits the message
4860 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4863 50. If a filter generated a file delivery with a non-absolute name (possible if
4864 no home directory exists for the router), the forbid_file option was not
4867 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4868 the existing equality matching.
4870 52. Exim was using ints instead of ino_t variables in some places where it was
4871 dealing with inode numbers.
4873 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4874 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4875 called TMPDIR, and if it finds it is different, it changes its value.
4877 54. The smtp_printf() function is now made available to local_scan() so
4878 additional output lines can be written before returning. There is also an
4879 smtp_fflush() function to enable the detection of a dropped connection.
4880 The variables smtp_input and smtp_batched_input are exported to
4883 55. Changed the default runtime configuration: the message "Unknown user"
4884 has been removed from the ACL, and instead placed on the localuser router,
4885 using the cannot_route_message feature. This means that any verification
4886 failures that generate their own messages won't get overridden. Similarly,
4887 the "Unrouteable address" message that was in the ACL for unverifiable
4888 relay addresses has also been removed.
4890 56. Added hosts_avoid_esmtp to the smtp transport.
4892 57. The exicyclog script was not checking for the esoteric option
4893 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4894 will work only if exicyclog is run under the appropriate euid.
4896 58. Following a discussion on the list, the rules by which Exim recognises line
4897 endings on incoming messages have been changed. The -dropcr and drop_cr
4898 options are now no-ops, retained only for backwards compatibility. The
4899 following line terminators are recognized: LF CRLF CR. However, special
4900 processing applies to CR:
4902 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4903 nor a local message in the state where . is a terminator.
4905 (ii) If a bare CR is encountered in a header line, an extra space is added
4906 after the line terminator so as not to end the header. The reasoning
4907 behind this is that bare CRs in header lines are most likely either
4908 to be mistakes, or people trying to play silly games.
4910 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4911 was being incorrectly given as 18 bytes larger than it should have been.
4912 This is a VOB (very old bug).
4914 60. This may never have affected anything current, but just in case it has:
4915 When the local host is found other than at the start of a list of hosts,
4916 the local host, those with the same MX, and any that follow, are discarded.
4917 When the list in question was part of a longer list of hosts, the following
4918 hosts (not currently being processed) were also being discarded. This no
4919 longer happens. I'm not sure if this situation could ever has previously
4922 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4925 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4926 header. This is something that is recommended in a new Internet Draft, and
4927 is something that is documented as being done by Sendmail. There are two
4928 possible values. For messages generated by the autoreply transport, Exim
4931 Auto-Submitted: auto-replied
4933 whereas for all other generated messages (e.g. bounces) it adds
4935 Auto-Submitted: auto-generated
4937 63. The "personal" condition in filters now includes a test for the
4938 Auto-Submitted: header. If it contains the string "auto-" the message it
4939 not considered personal.
4941 64. Added rcpt_include_affixes as a generic transport option.
4943 65. Added queue_only_override (default true).
4945 66. Added the syslog_duplication option.
4947 67. If what should have been the first header line of a message consisted of
4948 a space followed by a colon, Exim was mis-interpreting it as a header line.
4949 It isn't of course - it is syntactically invalid and should therefore be
4950 treated as the start of the message body. The misbehaviour could have
4951 caused a number of strange effects, including loss of data in subsequent
4952 header lines, and spool format errors.
4954 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4955 client host had authenticated. This control can now be exercised by an ACL
4956 for more flexibility.
4958 69. By default, callouts do not happen when testing with -bh. There is now a
4959 variant, -bhc, which does actually run the callout code, including
4960 consulting and updating the callout cache.
4962 70. Added support for saslauthd authentication, courtesy of Alexander
4965 71. If statvfs() failed on the spool or log directories while checking their
4966 size for availability, Exim confusingly gave the error "space shortage".
4967 Furthermore, in debugging mode it crashed with a floating point exception.
4968 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4969 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4970 serious problem, Exim now writes to the main and panic logs when this
4971 happens, with details of the failure. It then refuses to accept the
4972 incoming message, giving the message "spool directory problem" or "log
4973 directory problem" with a 421 code for SMTP messages.
4975 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4976 0, 1, and 2 exist, because some OS complain for execs without them (see
4977 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4978 descriptors. However, the code omitted to check that the open succeeded,
4979 causing mysterious errors if for some reason the permissions on /dev/null
4980 got screwed. Now Exim writes a message to the main and panic logs, and
4981 bombs out if it can't open /dev/null.
4983 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4984 interact so that it is all more flexible. It is supposed to remain
4985 backwards compatible. Also added extra_local_interfaces.
4987 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4988 to bomb out with an assertion failure - to the client this appears as a
4989 connection drop. This problem occurs in the part of the code that was taken
4990 from the Samba project. Fortunately, the assertion is in a very simple
4991 function, so I have fixed this by reproducing the function inline in the
4992 one place where it is called, and arranging for authentication to fail
4993 instead of killing the process with assert().
4995 75. The SPA client code was not working when the server requested OEM rather
4996 than Unicode encoding.
4998 76. Added code to make require_files with a specific uid setting more usable in
4999 the case where statting the file as root fails - usually a non-root-mounted
5000 NFS file system. When this happens and the failure is EACCES, Exim now
5001 forks a subprocess and does the per-uid checking as the relevant uid.
5003 77. Added process_log_path.
5005 78. If log_file_path was not explicitly set, a setting of check_log_space or
5006 check_log_inodes was ignored.
5008 79. If a space check for the spool or log partitions fails, the incident is now
5009 logged. Of course, in the latter case the data may get lost...
5011 80. Added the %p formatting code to string_format() so that it can be used to
5012 print addresses in debug_print(). Adjusted all the address printing in the
5013 debugging in store.c to use %p rather than %d.
5015 81. There was a concern that a line of code in smtp_in.c could overflow a
5016 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5017 initially expressed, the concern was not well-founded, because trailing
5018 spaces are removed early. However, if the trailing spaces were followed by
5019 a NULL, they did not get removed, so the overflow was possible. Two fixes
5022 (a) I re-wrote the offending code in a cleaner fashion.
5023 (b) If an incoming SMTP command contains a NULL character, it is rejected
5026 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5027 runs initgroups(), so that if the Exim user is in any additional groups,
5028 they will be used during message reception.
5034 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob