1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.598 2009/11/20 21:22:20 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.
17 NM/04 Bugzilla 915: Flag broken perl installation during build.
23 TK/01 Added patch by Johannes Berg that expands the main option
24 "spamd_servers" if it starts with a dollar sign.
26 TK/02 Write list of recipients to X-Envelope-Sender header when building
27 the mbox-format spool file for content scanning (suggested by Jakob
30 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
31 (http://www.dcc-servers.net/) support via dccifd. Activated by
32 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
33 experimental_spec.txt for more documentation.
35 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
36 by Mark Daniel Reidel <mr@df.eu>.
38 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
39 When building exim an external PCRE library is now needed -
40 PCRE is a system library on the majority of modern systems.
41 See entry on PCRE_LIBS in EDITME file.
43 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
44 conversation. Added nologin parameter to request.
45 Patch contributed by Kirill Miazine
47 TF/01 Do not log submission mode rewrites if they do not change the address.
49 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
51 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
52 log files in place. Contributed by Roberto Lima
54 NM/04 Bugzilla 667: close socket used by dovecot authenticator
56 TF/03 Bugzilla 615: When checking the local_parts router precondition
57 after a local_part_suffix or local_part_prefix option, Exim now
58 does not use the address's named list lookup cache, since this
59 contains cached lookups for the whole local part.
61 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
62 Robert Millan. Documentation is in experimental-spec.txt
64 TF/04 Bugzilla 668: Fix parallel build (make -j).
66 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
68 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
69 Patch provided by Jan Srzednicki
71 TF/05 Leading white space used to be stripped from $spam_report which
72 wrecked the formatting. Now it is preserved.
74 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
75 that they are available at delivery time.
77 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
79 TF/08 TLS error reporting now respects the incoming_interface and
80 incoming_port log selectors.
82 TF/09 Produce a more useful error message if an SMTP transport's hosts
83 setting expands to an empty string.
85 NM/06 Bugzilla 744: EXPN did not work under TLS.
86 Patch provided by Phil Pennock
88 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
89 Patch provided by Richard Godbee
91 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
92 acl_smtp_notquit, added index entry.
94 NM/09 Bugzilla 787: Potential buffer overflow in string_format
95 Patch provided by Eugene Bujak
97 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
98 Patch provided by Maxim Dounin
100 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
101 Patch provided by Phil Pennock
103 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
105 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
106 Patch provided by Brad "anomie" Jorsch
108 NM/14 Bugzilla 622: Added timeout setting to transport filter.
109 Patch provided by Dean Brooks
111 TK/05 Add native DKIM support (does not depend on external libraries).
113 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
114 Patch provided by Graeme Fowler
116 NM/16 Bugzilla 851: Documentation example syntax fix.
118 NM/17 Changed NOTICE file to remove references to embedded PCRE.
120 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
122 NM/19 Bugzilla 745: TLS version reporting
123 Patch provided by Phil Pennock
125 NM/20 Bugzilla 167: bool: condition support
126 Patch provided by Phil Pennock
128 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
129 Patch provided by Phil Pennock
131 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
132 Patch provided by Brad "anomie" Jorsch
134 NM/23 Bugzilla 687: Fix misparses in eximstats
135 Patch provided by Heiko Schlittermann
137 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
138 Patch provided by Heiko Schlittermann
140 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
141 plus update to original patch.
143 NM/26 Bugzilla 799: Documentation correction for ratelimit
145 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
146 Patch provided by David Brownlee
148 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
150 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
152 NM/30 Bugzilla 888: TLS documentation bugfixes
154 NM/31 Bugzilla 896: Dovecot buffer overrun fix
156 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
157 Unlike the original bugzilla I have changed all shell scripts in src tree
159 NM/33 Bugzilla 898: Transport filter timeout fix
160 Patch by Todd Rinaldo
162 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
163 Patch by Serge Demonchaux
165 NM/35 Bugzilla 39: Base64 decode bug fixes
166 Patch by Jakob Hirsch
168 NM/36 Bugzilla 909: Correct connect() call in dcc code
170 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
172 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
174 NM/39 Bugzilla 911: Fixed MakeLinks build script
180 TK/01 Add preliminary DKIM support. Currently requires a forked version of
181 ALT-N's libdkim that I have put here:
182 http://duncanthrax.net/exim-experimental/
184 Note to Michael Haardt: I had to rename some vars in sieve.c. They
185 were called 'true' and it seems that C99 defines that as a reserved
186 keyword to be used with 'bool' variable types. That means you could
187 not include C99-style headers which use bools without triggering
188 build errors in sieve.c.
190 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
191 as mailq or other aliases. Changed the --help handling significantly
192 to do whats expected. exim_usage() emits usage/help information.
194 SC/01 Added the -bylocaldomain option to eximstats.
196 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
198 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
200 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
206 PH/01 Another patch from the Sieve maintainer.
208 PH/02 When an IPv6 address is converted to a string for single-key lookup
209 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
210 dots are used instead of colons so that keys in lsearch files need not
211 contain colons. This was done some time before quoting was made available
212 in lsearch files. However, iplsearch files do require colons in IPv6 keys
213 (notated using the quote facility) so as to distinguish them from IPv4
214 keys. This meant that lookups for IP addresses in host lists did not work
215 for iplsearch lookups.
217 This has been fixed by arranging for IPv6 addresses to be expressed with
218 colons if the lookup type is iplsearch. This is not incompatible, because
219 previously such lookups could never work.
221 The situation is now rather anomolous, since one *can* have colons in
222 ordinary lsearch keys. However, making the change in all cases is
223 incompatible and would probably break a number of configurations.
225 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
228 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
229 conversion specification without a maximum field width, thereby enabling
230 a rogue spamd server to cause a buffer overflow. While nobody in their
231 right mind would setup Exim to query an untrusted spamd server, an
232 attacker that gains access to a server running spamd could potentially
233 exploit this vulnerability to run arbitrary code as the Exim user.
235 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
236 $primary_hostname instead of what libspf2 thinks the hosts name is.
238 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
239 a directory entry by the name of the lookup key. Previously, if a
240 symlink pointed to a non-existing file or a file in a directory that
241 Exim lacked permissions to read, a lookup for a key matching that
242 symlink would fail. Now it is enough that a matching directory entry
243 exists, symlink or not. (Bugzilla 503.)
245 PH/03 The body_linecount and body_zerocount variables are now exported in the
248 PH/04 Added the $dnslist_matched variable.
250 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
251 This means they are set thereafter only if the connection becomes
254 PH/06 Added the client_condition to authenticators so that some can be skipped
255 by clients under certain conditions.
257 PH/07 The error message for a badly-placed control=no_multiline_responses left
258 "_responses" off the end of the name.
260 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
262 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
263 (without spaces) instead of just copying the configuration text.
265 PH/10 Added the /noupdate option to the ratelimit ACL condition.
267 PH/11 Added $max_received_linelength.
269 PH/12 Added +ignore_defer and +include_defer to host lists.
271 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
272 way in which PCRE > 7.0 is built.
274 PH/14 Implemented queue_only_load_latch.
276 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
277 MAIL command. The effect was to mangle the value on 64-bit systems.
279 PH/16 Another patch from the Sieve maintainer.
281 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
283 PH/18 If a system quota error occurred while trying to create the file for
284 a maildir delivery, the message "Mailbox is full" was not appended to the
285 bounce if the delivery eventually timed out. Change 4.67/27 below applied
286 only to a quota excession during the actual writing of the file.
288 PH/19 It seems that peer DN values may contain newlines (and other non-printing
289 characters?) which causes problems in log lines. The DN values are now
290 passed through string_printing() before being added to log lines.
292 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
293 and InterBase are left for another time.)
295 PH/21 Added message_body_newlines option.
297 PH/22 Guard against possible overflow in moan_check_errorcopy().
299 PH/23 POSIX allows open() to be a macro; guard against that.
301 PH/24 If the recipient of an error message contained an @ in the local part
302 (suitably quoted, of course), incorrect values were put in $domain and
303 $local_part during the evaluation of errors_copy.
309 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
310 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
313 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
314 issue a MAIL command.
316 PH/02 In an ACL statement such as
318 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
320 if a client was not listed at all, or was listed with a value other than
321 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
322 the condition was not true (as it should be), so access was not denied.
323 The bug was that the ! inversion was incorrectly passed on to the second
324 item. This has been fixed.
326 PH/03 Added additional dnslists conditions == and =& which are different from
327 = and & when the dns lookup returns more than one IP address.
329 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
330 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
332 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
333 FSYNC, which compiles an option called disable_fsync that allows for
334 bypassing fsync(). The documentation is heavily laced with warnings.
336 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
338 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
339 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
340 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
341 including adding "make clean"; (3) Added -fPIC when compiling the test
342 dynamically loaded module, to get rid of a warning.
344 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
345 message fails, move_frozen_messages = true and ignore_bounce_errors_after
346 = 0s. The bug is otherwise harmless.
348 PH/07 There was a bug in the dovecot authenticator such that the value of
349 $auth1 could be overwritten, and so not correctly preserved, after a
350 successful authentication. This usually meant that the value preserved by
351 the server_setid option was incorrect.
353 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
355 PH/09 Installed PCRE release 7.0.
357 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
358 run for batched SMTP input. It is now run at the start of every message
359 in the batch. While fixing this I discovered that the process information
360 (output by running exiwhat) was not always getting set for -bs and -bS
361 input. This is fixed, and it now also says "batched" for BSMTP.
363 PH/11 Added control=no_pipelining.
365 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
366 patch, slightly modified), and move the expansion of helo_data till after
367 the connection is made in the smtp transport (so it can use these
370 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
372 PH/14 Added log_selector = +pid.
374 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
376 PH/16 Add ${if forany and ${if forall.
378 PH/17 Added dsn_from option to vary the From: line in DSNs.
380 PH/18 Flush SMTP output before performing a callout, unless control =
381 no_callout_flush is set.
383 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
384 was true (the default) a successful delivery failed to delete the retry
385 item, thus causing premature timeout of the address. The bug is now
388 PH/20 Added hosts_avoid_pipelining to the smtp transport.
390 PH/21 Long custom messages for fakedefer and fakereject are now split up
391 into multiline reponses in the same way that messages for "deny" and
392 other ACL rejections are.
394 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
395 with slight modification.
397 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
398 draft, changing the syntax and factoring some duplicate code".
400 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
401 for deliveries of the second and subsequent messages over the same SMTP
404 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
405 ${reduce, with only minor "tidies".
407 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
409 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
410 expansion side effects.
412 PH/27 When a message times out after an over-quota error from an Exim-imposed
413 quota, the bounce message says "mailbox is full". This message was not
414 being given when it was a system quota that was exceeded. It now should
417 MH/03 Made $recipients available in local_scan(). local_scan() already has
418 better access to the recipient list through recipients_list[], but
419 $recipients can be useful in postmaster-provided expansion strings.
421 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
422 in the case of a MAIL command with additional options following the
423 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
424 were accidentally chopped off.
426 PH/29 SMTP synchronization checks are implemented when a command is read -
427 there is a check that no more input is waiting when there shouldn't be
428 any. However, for some commands, a delay in an ACL can mean that it is
429 some time before the response is written. In this time, more input might
430 arrive, invalidly. So now there are extra checks after an ACL has run for
431 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
432 pipelining has not been advertised.
434 PH/30 MH's patch to allow iscntrl() characters to be list separators.
436 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
437 returned in the SMTP response when smtp_return_error_details was false.
440 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
441 instead of the C I/O that was originally supplied, because problems were
444 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
445 Exim which did not show up earlier: it was assuming that a call to
446 SSL_CTX_set_info_callback() might give an error value. In fact, there is
447 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
448 was a macro that became an assignment, so it seemed to work. This has
449 changed to a proper function call with a void return, hence the compile
450 error. Exim's code has been fixed.
452 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
455 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
457 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
459 PH/37 If a message is not accepted after it has had an id assigned (e.g.
460 because it turns out to be too big or there is a timeout) there is no
461 "Completed" line in the log. When some messages of this type were
462 selected by exigrep, they were listed as "not completed". Others were
463 picked up by some special patterns. I have improved the selection
464 criteria to be more general.
466 PH/38 The host_find_failed option in the manualroute router can now be set
467 to "ignore", to completely ignore a host whose IP address cannot be
468 found. If all hosts are ignored, the behaviour is controlled by the new
469 host_all_ignored option.
471 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
472 homing or because of multiple MX records with /mx) generated more than
473 one IP address, and the following item turned out to be the local host,
474 all the secondary addresses of the first item were incorrectly removed
475 from the list, along with the local host and any following hosts (which
476 is what is supposed to happen).
478 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
479 whoever called Exim into the -H file. In the case of the daemon it was
480 behaving confusingly. When first started, it used values for whoever
481 started the daemon, but after a SIGHUP it used the Exim user (because it
482 calls itself on a restart). I have changed the code so that it now always
485 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
486 message are rejected with the same error (e.g. no authentication or bad
487 sender address), and a DATA command is nevertheless sent (as can happen
488 with PIPELINING or a stupid MUA), the error message that was given to the
489 RCPT commands is included in the rejection of the DATA command. This is
490 intended to be helpful for MUAs that show only the final error to their
493 PH/42 Another patch from the Sieve maintainer.
495 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
496 Eximstats - Fixed some broken HTML links and added missing column headers
498 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
499 columns for Rejects, Temp Rejects, Ham, and Spam rows.
501 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
503 PH/43 Yet another patch from the Sieve maintainer.
505 PH/44 I found a way to check for a TCP/IP connection going away before sending
506 the response to the final '.' that terminates a message, but only in the
507 case where the client has not sent further data following the '.'
508 (unfortunately, this is allowed). However, in many cases there won't be
509 any further data because there won't be any more messages to send. A call
510 to select() can be used: if it shows that the input is "ready", there is
511 either input waiting, or the socket has been closed. An attempt to read
512 the next input character can distinguish the two cases. Previously, Exim
513 would have sent an OK response which the client would never have see.
514 This could lead to message repetition. This fix should cure that, at
515 least in a lot of common cases.
517 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
518 advertised in response to EHLO.
524 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
525 fixed by 4.65/MH/01 (is this a record?) are fixed:
527 (i) An empty string was always treated as zero by the numeric comparison
528 operators. This behaviour has been restored.
530 (ii) It is documented that the numeric comparison operators always treat
531 their arguments as decimal numbers. This was broken in that numbers
532 starting with 0 were being interpreted as octal.
534 While fixing these problems I realized that there was another issue that
535 hadn't been noticed. Values of message_size_limit (both the global option
536 and the transport option) were treated as octal if they started with 0.
537 The documentation was vague. These values are now always treated as
538 decimal, and I will make that clear in the documentation.
544 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
545 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
548 MH/01 Don't check that the operands of numeric comparison operators are
549 integers when their expansion is in "skipping" mode (fixes bug
550 introduced by 4.64-PH/07).
552 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
553 child addresses, Exim now panics and dies. Previously, because the count
554 is held in a short int, deliveries were likely to be lost. As such a
555 large number of recipients for a single message is ridiculous
556 (performance will be very, very poor), I have chosen to impose a limit
557 rather than extend the field.
563 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
564 leftover -K file (the existence of which was triggered by #402).
565 While we were at it, introduced process PID as part of the -K
566 filename. This should rule out race conditions when creating
569 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
570 processing considerably. Previous code took too long for large mails,
571 triggering a timeout which in turn triggers #401.
573 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
574 in the DK code in transports.c. sendfile() is not really portable,
575 hence the _LINUX specificness.
577 TF/01 In the add_headers option to the mail command in an Exim filter,
578 there was a bug that Exim would claim a syntax error in any
579 header after the first one which had an odd number of characters
582 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
583 callout verification, Exim cached a "reject" for the entire domain. This
584 is correct for most verifications, but it is not correct for a recipient
585 verification with use_sender or use_postmaster set, because in that case
586 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
587 case of MAIL FROM:<> rejection from other early rejections (e.g.
588 rejection of HELO). When verifying a recipient using a non-null MAIL
589 address, the cache is ignored if it shows MAIL FROM:<> rejection.
590 Whatever the result of the callout, the value of the domain cache is
591 left unchanged (for any other kind of callout, getting as far as trying
592 RCPT means that the domain itself is ok).
594 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
597 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
598 manifest itself as EPIPE rather than ECONNECT. When tidying away a
599 session, the daemon ignores ECONNECT errors and logs others; it now
600 ignores EPIPE as well.
602 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
603 (quoted-printable decoding).
605 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
606 later the small subsequent patch to fix an introduced bug.
608 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
610 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
612 PH/08 An error is now given if message_size_limit is specified negative.
614 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
615 to be given (somewhat) arbitrary names.
617 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
620 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
621 miscellaneous code fixes
623 PH/10 Added the log_reject_target ACL modifier to specify where to log
626 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
627 hostname. This is wrong, because it relates to the incoming message (and
628 probably the interface on which it is arriving) and not to the outgoing
629 callout (which could be using a different interface). This has been
630 changed to use the value of the helo_data option from the smtp transport
631 instead - this is what is used when a message is actually being sent. If
632 there is no remote transport (possible with a router that sets up host
633 addresses), $smtp_active_hostname is used.
635 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
636 tweaks were necessary in order to get it to work (see also 21 below):
637 (a) The code assumed that strncpy() returns a negative number on buffer
638 overflow, which isn't the case. Replaced with Exim's string_format()
640 (b) There were several signed/unsigned issues. I just did the minimum
641 hacking in of casts. There is scope for a larger refactoring.
642 (c) The code used strcasecmp() which is not a standard C function.
643 Replaced with Exim's strcmpic() function.
644 (d) The code set only $1; it now sets $auth1 as well.
645 (e) A simple test gave the error "authentication client didn't specify
646 service in request". It would seem that Dovecot has changed its
647 interface. Fortunately there's a specification; I followed it and
648 changed what the client sends and it appears to be working now.
650 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
653 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
654 address A is aliased to B and C, where B exists and C does not. Without
655 -v the output is "A verified" because verification stops after a
656 successful redirection if more than one address is generated. However,
657 with -v the child addresses are also verified. Exim was outputting "A
658 failed to verify" and then showing the successful verification for C,
659 with its parentage. It now outputs "B failed to verify", showing B's
660 parentage before showing the successful verification of C.
662 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
663 look up a TXT record in a specific list after matching in a combined
666 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
667 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
668 they consult the DNS. I had assumed they would set it the way they
669 wanted; and indeed my experiments on Linux seem to show that in some
670 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
671 To be on the safe side, however, I have now made the interface to
672 host_find_byname() similar to host_find_bydns(), with an argument
673 containing the DNS resolver options. The host_find_byname() function now
674 sets these options at its start, just as host_find_bydns() does. The smtp
675 transport options dns_qualify_single and dns_search_parents are passed to
676 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
677 of host_find_byname() use the default settings of RES_DEFNAMES
678 (qualify_single) but not RES_DNSRCH (search_parents).
680 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
681 spool_read_header() do less string testing, by means of a preliminary
682 switch on the second character of optional "-foo" lines. (This is
683 overdue, caused by the large number of possibilities that now exist.
684 Originally there were few.) While I was there, I also converted the
685 str(n)cmp tests so they don't re-test the leading "-" and the first
686 character, in the hope this might squeeze out yet more improvement.
688 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
689 flag allowing group syntax was set by the header_syntax check but not
690 turned off, possible causing trouble later; (2) The flag was not being
691 set at all for the header_verify test, causing "group"-style headers to
692 be rejected. I have now set it in this case, and also caused header_
693 verify to ignore an empty address taken from a group. While doing this, I
694 came across some other cases where the code for allowing group syntax
695 while scanning a header line wasn't quite right (mostly, not resetting
696 the flag correctly in the right place). These bugs could have caused
697 trouble for malformed header lines. I hope it is now all correct.
699 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
700 with the "reply" argument non-NULL. The code, however (which originally
701 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
702 but it didn't always do it. This confused somebody who was copying the
703 code for some other use. I have removed all the tests.
705 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
706 feature that was used to support insecure browsers during the U.S. crypto
707 embargo. It requires special client support, and Exim is probably the
708 only MTA that supported it -- and would never use it because real RSA is
709 always available. This code has been removed, because it had the bad
710 effect of slowing Exim down by computing (never used) parameters for the
711 RSA_EXPORT functionality.
713 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
714 authenticator to fail if there's a tab character in the incoming data
715 (there should never be unless someone is messing about, as it's supposed
716 to be base64-encoded). Also added, on Timo's advice, the "secured" option
717 if the connection is using TLS or if the remote IP is the same as the
718 local IP, and the "valid-client-cert option" if a client certificate has
721 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
722 authenticators. This can be used for authorization after authentication
723 succeeds. (In the case of plaintext, it servers for both authentication
726 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
727 if any retry times were supplied.
729 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
730 connection, where there is no client IP address to check. In this
731 situation, the verify now always succeeds.
733 PH/25 Applied John Jetmore's -Mset patch.
735 PH/26 Added -bem to be like -Mset, but loading a message from a file.
737 PH/27 In a string expansion for a processed (not raw) header when multiple
738 headers of the same name were present, leading whitespace was being
739 removed from all of them, but trailing whitespace was being removed only
740 from the last one. Now trailing whitespace is removed from each header
741 before concatenation. Completely empty headers in a concatenation (as
744 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
745 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
747 PH/29 [Removed. This was a change that I later backed out, and forgot to
748 correct the ChangeLog entry (that I had efficiently created) before
749 committing the later change.]
751 PH/30 Exim was sometimes attempting to deliver messages that had suffered
752 address errors (4xx response to RCPT) over the same connection as other
753 messages routed to the same hosts. Such deliveries are always "forced",
754 so retry times are not inspected. This resulted in far too many retries
755 for the affected addresses. The effect occurred only when there were more
756 hosts than the hosts_max_try setting in the smtp transport when it had
757 the 4xx errors. Those hosts that it had tried were not added to the list
758 of hosts for which the message was waiting, so if all were tried, there
759 was no problem. Two fixes have been applied:
761 (i) If there are any address or message errors in an SMTP delivery, none
762 of the hosts (tried or untried) are now added to the list of hosts
763 for which the message is waiting, so the message should not be a
764 candidate for sending over the same connection that was used for a
765 successful delivery of some other message. This seems entirely
766 reasonable: after all the message is NOT "waiting for some host".
767 This is so "obvious" that I'm not sure why it wasn't done
768 previously. Hope I haven't missed anything, but it can't do any
769 harm, as the worst effect is to miss an optimization.
771 (ii) If, despite (i), such a delivery is accidentally attempted, the
772 routing retry time is respected, so at least it doesn't keep
773 hammering the server.
775 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
776 in ${readsocket because some servers need this prod.
778 PH/32 Added some extra debug output when updating a wait-xxx database.
780 PH/33 The hint "could be header name not terminated by colon", which has been
781 given for certain expansion errors for a long time, was not being given
782 for the ${if def:h_colon_omitted{... case.
784 PH/34 The spec says: "With one important exception, whenever a domain list is
785 being scanned, $domain contains the subject domain." There was at least
786 one case where this was not true.
788 PH/35 The error "getsockname() failed: connection reset by peer" was being
789 written to the panic log as well as the main log, but it isn't really
790 panic-worthy as it just means the connection died rather early on. I have
791 removed the panic log writing for the ECONNRESET error when getsockname()
794 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
795 runs only) independently of the message's sender address. This meant
796 that, if the 4xx error was in fact related to the sender, a different
797 message to the same recipient with a different sender could confuse
798 things. In particualar, this can happen when sending to a greylisting
799 server, but other circumstances could also provoke similar problems.
800 I have changed the default so that the retry time for these errors is now
801 based a combination of the sender and recipient addresses. This change
802 can be overridden by setting address_retry_include_sender=false in the
805 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
806 remote server are returned as part of bounce messages. This was not
807 happening for LMTP over a pipe (the lmtp transport), but now it is the
808 same for both kinds of LMTP.
810 PH/38 Despite being documented as not happening, Exim was rewriting addresses
811 in header lines that were in fact CNAMEs. This is no longer the case.
813 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
814 and queue runs started by the daemon processed all messages. This has
815 been fixed so that -R and -S can now usefully be given with -q<time>.
817 PH/40 Import PCRE release 6.7 (fixes some bugs).
819 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
821 PH/42 Give an error if -q is specified more than once.
823 PH/43 Renamed the variables $interface_address and $interface_port as
824 $received_ip_address and $received_port, to make it clear that these
825 values apply to message reception, and not to the outgoing interface when
826 a message is delivered. (The old names remain recognized, of course.)
828 PH/44 There was no timeout on the connect() call when using a Unix domain
829 socket in the ${readsocket expansion. There now is.
831 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
832 be meaningful with "accept".
834 SC/01 Eximstats V1.43
835 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
837 SC/02 Eximstats V1.44
838 Use a glob alias rather than an array ref in the generated
839 parser. This improves both readability and performance.
841 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
842 Collect SpamAssassin and rejection statistics.
843 Don't display local sender or destination tables unless
844 there is data to show.
845 Added average volumes into the top table text output.
847 SC/04 Eximstats V1.46
848 Collect data on the number of addresses (recipients)
849 as well as the number of messages.
851 SC/05 Eximstats V1.47
852 Added 'Message too big' to the list of mail rejection
853 reasons (thanks to Marco Gaiarin).
855 SC/06 Eximstats V1.48
856 Mainlog lines which have GMT offsets and are too short to
857 have a flag are now skipped.
859 SC/07 Eximstats V1.49 (Alain Williams)
860 Added the -emptyok flag.
862 SC/08 Eximstats V1.50
863 Fixes for obtaining the IP address from reject messages.
865 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
866 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
867 whitesspace changes from 4.64-PH/27
869 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
872 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
873 are found, allow negative numbers in numeric criteria)
875 JJ/06 exipick.20061117.2, added new $message_body_missing variable
877 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
878 to match changes made in 4.64-PH/43
880 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
882 PH/47 Put in an explicit test for a DNS lookup of an address record where the
883 "domain" is actually an IP address, and force a failure. This locks out
884 those revolvers/nameservers that support "A-for-A" lookups, in
885 contravention of the specifications.
887 PH/48 When a host name was looked up from an IP address, and the subsequent
888 forward lookup of the name timed out, the host name was left in
889 $sender_host_name, contrary to the specification.
891 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
892 restricted to single-key lookups, Exim was not diagnosing an error if
893 * or *@ was used with a query-style lookup.
895 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
897 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
898 long ago, but noone interested enough thought of it. Let's just say that
899 the "1.1" means that there are some new functions that weren't there at
900 some point in the past.
902 PH/51 Error processing for expansion failure of helo_data from an smtp
903 transport during callout processing was broken.
905 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
906 tested/used via the -bh/-bhc/-bs options.
908 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
909 bug, fixed in subsequent PCRE releases).
911 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
912 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
914 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
920 SC/01 Use a glob alias rather than an array ref in eximstats generated
921 parser. This improves both readability and performance.
923 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
924 Don't display local sender or destination tables in eximstats unless
925 there is data to show.
926 Added average volumes into the eximstats top table text output.
928 SC/03 Collect data on the number of addresses (recipients) as well
929 as the number of messages in eximstats.
931 TF/01 Correct an error in the documentation for the redirect router. Exim
932 does (usually) call initgroups() when daemonizing.
934 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
935 with consistent privilege compared to when running as a daemon.
937 TF/03 Note in the spec that $authenticated_id is not set for local
938 submissions from trusted users.
940 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
941 Thanks to Dean Brooks <dean@iglou.com> for the patch.
943 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
944 by adding some example configuration directives to the default
945 configuration file. A little bit of work is required to uncomment the
946 directives and define how usernames and passwords are checked, but
947 there is now a framework to start from.
949 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
950 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
951 without this. I don't know how relevant this is to other LDAP libraries.
953 PH/02 Add the verb name to the "unknown ACL verb" error.
955 PH/03 Magnus Holmgren's patch for filter_prepend_home.
957 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
959 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
960 directory not expanded when it should be if an expanded home directory
961 was set for the address (which is overridden by the transport).
963 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
966 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
967 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
968 because it is too late at that time, and has no effect.
970 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
971 security issue with \' (bugzilla #107). I could not use the
972 PQescapeStringConn() function, because it needs a PGconn value as one of
975 PH/08 When testing addresses using -bt, indicate those final addresses that
976 are duplicates that would not cause an additional delivery. At least one
977 person was confused, thinking that -bt output corresponded to deliveries.
978 (Suppressing duplicates isn't a good idea as you lose the information
979 about possibly different redirections that led to the duplicates.)
981 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
982 systems where poll() doesn't work, in particular OS X.
984 PH/10 Added more information to debugging output for retry time not reached.
986 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
987 operations in malware.c.
989 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
992 PH/13 If write_rejectlog was set false when logging was sent to syslog with
993 syslog_duplication set false, log lines that would normally be written
994 both the the main log and to the reject log were not written to syslog at
997 PH/14 In the default configuration, change the use of "message" in ACL warn
998 statements to "add_header".
1000 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1001 not followed by a command (e.g. "seen endif").
1003 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1004 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1007 PH/17 Added extra conditions to the default value of delay_warning_condition
1011 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1012 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1013 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1016 The Auto-Submitted: and various List- headers are standardised, whereas I
1017 don't think Precedence: ever was.
1019 PH/18 Refactored debugging code in route_finduser() to show more information,
1020 in particular, the error code if getpwnam() issues one.
1022 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1023 This is apparently needed in addition to the PH/07 change above to avoid
1024 any possible encoding problems.
1026 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1027 but not after initializing Perl.
1029 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1030 output them only if debugging. By default they are written stderr,
1031 apparently, which is not desirable.
1033 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1036 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1039 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1041 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1042 authenticated or an ident call has been made. Suppress the default
1043 values for $authenticated_id and $authenticated_sender (but permit -oMai
1044 and -oMas) when testing with -bh.
1046 PH/24 Re-jigged the order of the tests in the default configuration so that the
1047 tests for valid domains and recipients precede the DNS black list and CSA
1048 tests, on the grounds that those ones are more expensive.
1050 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1051 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1052 This bug exists in every version of Exim that I still have, right back to
1055 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1056 However, an attempt to turn on case-sensitivity in a regex key by
1057 including (?-i) didn't work because the subject string was already
1058 lowercased, and the effects were non-intuitive. It turns out that a
1059 one-line patch can be used to allow (?-i) to work as expected.
1065 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1066 other effects) broke the use of negated acl sub-conditions.
1068 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1071 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1072 "Deny" causes Exim to reject the incoming connection with a 554 error.
1073 Unfortunately, if there is a major crisis, such as a disk failure,
1074 tcp-wrappers gives "deny", whereas what one would like would be some
1075 kind of temporary error. A kludge has been added to help with this.
1076 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1077 554 error is used if errno is still zero or contains ENOENT (which occurs
1078 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1081 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1083 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1084 errors. Otherwise a message that provokes a temporary error (when other
1085 messages do not) can cause a whole host to time out.
1087 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1088 addresses were routed directly to files or pipes from a redirect router.
1089 File deliveries just didn't batch; pipe deliveries might have suffered
1092 PH/06 A failure to get a lock for a hints database would erroneously always say
1093 "Failed to get write lock", even when it was really a read lock.
1095 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1096 of 0600. This has been changed to use the value of the lockfile_mode
1097 option (which defaults to 0600).
1099 PH/08 Applied small patch from the Sieve maintainer.
1101 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1102 folder from quota calculations, a direct delivery into this folder messed
1103 up the contents of the maildirsize file. This was because the regex was
1104 used only to exclude .Trash (or whatever) when the size of the mailbox
1105 was calculated. There was no check that a delivery was happening into an
1106 excluded directory. This bug has been fixed by ignoring all quota
1107 processing for deliveries into excluded directories.
1109 PH/10 Added the maildirfolder_create_regex option to appendfile.
1115 PH/01 The code for finding all the local interface addresses on a FreeBSD
1116 system running IPv6 was broken. This may well have applied to all BSD
1117 systems, as well as to others that have similar system calls. The broken
1118 code found IPv4 interfaces correctly, but gave incorrect values for the
1119 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1120 that it would not match correctly against @[] and not recognize the IPv6
1123 PH/02 The ipliteral router was not recognizing addresses of the form user@
1124 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1126 PH/03 Added disable_ipv6.
1128 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1129 decoded form, because it is most often used to construct To: headers
1130 lines in autoreplies, and the decoded form may well be syntactically
1131 invalid. However, $reply_address has leading white space removed, and all
1132 newlines turned into spaces so that the autoreply transport does not
1135 PH/05 If group was specified without a user on a router, and no group or user
1136 was specified on a transport, the group from the router was ignored.
1138 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1139 for visible compile-time settings that can be used to change these
1140 numbers, for those that want even more. Backwards compatibility with old
1141 spool files has been maintained. However, going back to a previous Exim
1142 release will lost any variables that are in spool files.
1144 PH/07 Two small changes when running in the test harness: increase delay when
1145 passing a TCP/IP connection to a new process, in case the original
1146 process has to generate a bounce, and remove special handling of
1147 127.0.0.2 (sic), which is no longer necessary.
1149 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1150 be the same on different OS.
1152 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1155 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1156 whether --show-vars was specified or not
1158 JJ/02 exipick: Added support for new ACL variable spool format introduced
1161 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1162 syntactically invalid From: or Reply-to: line, and a filter used this to
1163 generate an autoreply, and therefore failed to obtain an address for the
1164 autoreply, Exim could try to deliver to a non-existent relative file
1165 name, causing unrelated and misleading errors. What now happens is that
1166 it logs this as a hard delivery error, but does not attempt to create a
1169 PH/11 The exinext utility has a -C option for testing purposes, but although
1170 the given file was scanned by exinext itself; it wasn't being passed on
1171 when Exim was called.
1173 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1174 an end-of-file indication when reading a command response.
1176 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1177 compiled. In many other places in Exim, IPv6 addresses are always
1178 recognized, so I have changed this. It also means that IPv4 domain
1179 literals of the form [IPV4:n.n.n.n] are now always recognized.
1181 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1182 used if the router is not running as root, for example, when verifying at
1183 ACL time, or when using -bh. The debugging output from this situation was
1184 non-existent - all you got was a failure to exec. I have made two
1187 (a) Failures to set uid/gid, the current directory, or a process leader
1188 in a subprocess such as that created by queryprogram now generate
1189 suitable debugging ouput when -d is set.
1191 (b) The queryprogram router detects when it is not running as root,
1192 outputs suitable debugging information if -d is set, and then runs
1193 the subprocess without attempting to change uid/gid.
1195 PH/15 Minor change to Makefile for building test_host (undocumented testing
1198 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1199 additional section of a DNS packet that returns MX or SRV records.
1200 Instead, it always explicitly searches for A/AAAA records. This avoids
1201 major problems that occur when a DNS server includes only records of one
1202 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1203 fixed another bug: if SRV records were looked up and the corresponding
1204 address records were *not* found in the additional section, the port
1205 values from the SRV records were lost.
1207 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1208 using the correct key (the original address) when searching the retry
1209 rules in order to find which one to use for generating the retry hint.
1211 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1212 adding the default one. Similarly, if it contains a Reply-To: header, the
1213 errors_reply_to option, if set, is not used.
1215 PH/19 When calculating a retry time, Exim used to measure the "time since
1216 failure" by looking at the "first failed" field in the retry record. Now
1217 it does not use this if it is later than than the arrival time of the
1218 message. Instead it uses the arrival time. This makes for better
1219 behaviour in cases where some deliveries succeed, thus re-setting the
1220 "first failed" field. An example is a quota failure for a huge message
1221 when small messages continue to be delivered. Without this change, the
1222 "time since failure" will always be short, possible causing more frequent
1223 delivery attempts for the huge message than are intended.
1224 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1226 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1227 $1, $2, $3) because the numerical variables can be reset during some
1228 expansion items (e.g. "match"), thereby losing the authentication data.
1230 PH/21 Make -bV show the size of off_t variables so that the test suite can
1231 decide whether to run tests for quotas > 2G.
1233 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1234 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1235 filecount value is greater than 2G or if a quota value is greater than 2G
1236 on a system where the size of off_t is not greater than 4, a panic error
1239 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1240 never match. The debug and -bh output now contains an explicit error
1241 message indicating a malformed IPv4 address or mask.
1243 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1244 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1245 PH/23 above applies.
1247 PH/25 Do not write to syslog when running in the test harness. The only
1248 occasion when this arises is a failure to open the main or panic logs
1249 (for which there is an explicit test).
1251 PH/26 Added the /no_tell option to "control=freeze".
1253 PH/27 If a host name lookup failed very early in a connection, for example, if
1254 the IP address matched host_lookup and the reverse lookup yielded a name
1255 that did not have a forward lookup, an error message of the form "no IP
1256 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1257 could be logged. Now it outputs the IP address instead of "NULL".
1259 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1260 allows the sender and the authenticated sender to be set when
1261 submitting a message from within Exim. Since child_open_exim() is
1262 documented for local_scan(), the new function should be too.
1264 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1265 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1266 results in an empty string is now treated as unset.
1268 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1270 PH/31 Added sender_verify_fail logging option.
1272 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1273 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1274 tidied the source and removed it altogether.
1276 PH/33 When a queue run was abandoned because the load average was too high, a
1277 log line was always written; now it is written only if the queue_run log
1278 selector is set. In addition, the log line for abandonment now contains
1279 information about the queue run such as the pid. This is always present
1280 in "start" and "stop" lines but was omitted from the "abandon" line.
1282 PH/34 Omit spaces between a header name and the colon in the error message that
1283 is given when verify = headers_syntax fails (if there are lots of them,
1284 the message gets confusing).
1286 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1287 names, as there are now some PTR records that contain slashes. This check
1288 is only to protect against broken name servers that fall over on strange
1289 characters, so the fact that it applies to all lookups doesn't matter.
1291 PH/36 Now that the new test suite is complete, we can remove some of the
1292 special code in Exim that was needed for the old test suite. For example,
1293 sorting DNS records because real resolvers return them in an arbitrary
1294 order. The new test suite's fake resolver always returns records in the
1297 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1298 bounces) except when queue_only is set, to avoid logging races between
1299 the different processes.
1301 PH/38 Panic-die if .include specifies a non-absolute path.
1303 PH/39 A tweak to the "H" retry rule from its user.
1305 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1306 a label. They prevented compilation on older perls.
1308 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1309 a warning to be raised on newish perls.
1311 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1312 on queue. Changes to match documented behaviour of showing count of
1313 messages matching specified criteria.
1315 PH/40 Changed the default ident timeout from 30s to 5s.
1317 PH/41 Added support for the use of login_cap features, on those BSD systems
1318 that have them, for controlling the resources used by pipe deliveries.
1320 PH/42 The content-scanning code uses fopen() to create files in which to put
1321 message data. Previously it was not paying any attention to the mode of
1322 the files. Exim runs with umask(0) because the rest of the code creates
1323 files with open(), and sets the required mode explicitly. Thus, these
1324 files were ending up world-writeable. This was not a big issue, because,
1325 being within the spool directory, they were not world-accessible. I have
1326 created a function called modefopen, which takes an additional mode
1327 argument. It sets umask(777), creates the file, chmods it to the required
1328 mode, then resets the umask. All the relevant calls to fopen() in the
1329 content scanning code have been changed to use this function.
1331 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1332 to 24 hours. This avoids potential overflow problems when processing G
1333 and H retry rules. I suspect nobody ever tinkers with this value.
1335 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1337 PH/45 When the plaintext authenticator is running as a client, the server's
1338 challenges are checked to ensure they are valid base64 strings. By
1339 default, the authentication attempt is cancelled if an invalid string is
1340 received. Setting client_ignore_invalid_base64 true ignores these errors.
1341 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1342 they are received. Thus, the responses can be made to depend on the
1343 challenges. If an invalid string is ignored, an empty string is placed in
1346 PH/46 Messages that are created by the autoreply transport now contains a
1347 References: header, in accordance with RFCs 2822 and 3834.
1349 PH/47 Added authenticated_sender_force to the smtp transport.
1351 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1353 PH/49 Installed latest patch from the Sieve maintainer.
1355 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1356 was also set, the appendfile transport was unnecessarily scanning a
1357 directory of message files (e.g. for maildir delivery) to find the count
1358 of files (along with the size), even though it did not need this
1359 information. It now does the scan only if it needs to find either the
1360 size of the count of files.
1362 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1364 PH/52 Two bugs concerned with error handling when the smtp transport is
1367 (i) Exim was not creating retry information for temporary errors given
1368 for individual recipients after the DATA command when the smtp transport
1369 was used in LMTP mode. This meant that they could be retried too
1370 frequently, and not timed out correctly.
1372 (ii) Exim was setting the flag that allows error details to be returned
1373 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1374 recipients that were returned after the DATA command.
1376 PH/53 This is related to PH/52, but is more general: for any failing address,
1377 when detailed error information was permitted to be returned to the
1378 sender, but the error was temporary, then after the final timeout, only
1379 "retry timeout exceeded" was returned. Now it returns the full error as
1380 well as "retry timeout exceeded".
1382 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1383 do this, and (what is worse) MTAs that accept it.
1385 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1386 will now be deprecated.
1388 PH/56 New os.c-cygwin from the Cygwin maintainer.
1390 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1391 formats (previously only available in exim formats via -bpr, -bpru,
1392 and -bpra. Now also available in native and exiqgrep formats)
1394 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1395 with very large, slow to parse queues
1397 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1399 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1401 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1402 responses to authentication challenges, though it was showing the
1403 challenges; (ii) I've removed the CR characters from the debug output for
1406 PH/58 Allow for the insertion of a newline as well as a space when a string
1407 is turned into more than one encoded-word during RFC 2047 encoding. The
1408 Sieve code now uses this.
1410 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1411 data_4xx, lost_connection, tls_required.
1413 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1414 message was being sent as an SMTP response.
1416 PH/61 Add -l and -k options to exicyclog.
1418 PH/62 When verifying, if an address was redirected to one new address, so that
1419 verification continued, and the new address failed or deferred after
1420 having set something in $address_data, the value of $address_data was not
1421 passed back to the ACL. This was different to the case when no
1422 redirection occurred. The value is now passed back in both cases.
1424 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1425 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1426 use login_cap.h, so on its own it isn't the distinguishing feature. The
1427 new name refers directly to the setclassresources() function.
1429 PH/65 Added configuration files for NetBSD3.
1431 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1433 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1434 is preferred over IPv4.
1436 PH/68 The bounce_return_message and bounce_return_body options were not being
1437 honoured for bounces generated during the reception of non-SMTP messages.
1438 In particular, this applied to messages rejected by the ACL. This bug has
1439 been fixed. However, if bounce_return_message is true and bounce_return_
1440 body is false, the headers that are returned for a non-SMTP message
1441 include only those that have been read before the error was detected.
1442 (In the case of an ACL rejection, they have all been read.)
1444 PH/69 The HTML version of the specification is now built in a directory called
1445 spec_html instead of spec.html, because the latter looks like a path with
1446 a MIME-type, and this confuses some software.
1448 PH/70 Catch two compiler warnings in sieve.c.
1450 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1451 function verify_get_ident() calls ip_connect() to connect a socket, but
1452 if the "connect()" function timed out, ip_connect() used to close the
1453 socket. However, verify_get_ident() also closes the socket later, and in
1454 between Exim writes to the log, which may get opened at this point. When
1455 the socket was closed in ip_connect(), the log could get the same file
1456 descriptor number as the socket. This naturally causes chaos. The fix is
1457 not to close the socket in ip_connect(); the socket should be closed by
1458 the function that creates it. There was only one place in the code where
1459 this was missing, in the iplookup router, which I don't think anybody now
1460 uses, but I've fixed it anyway.
1462 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1463 well as to direct DNS lookups. Otherwise the handling of names in host
1464 lists is inconsistent and therefore confusing.
1470 PH/01 Two changes to the default runtime configuration:
1472 (1) Move the checks for relay_from_hosts and authenticated clients from
1473 after to before the (commented out) DNS black list checks.
1475 (2) Add control=submission to the relay_from_hosts and authenticated
1476 clients checks, on the grounds that messages accepted by these
1477 statements are most likely to be submissions.
1479 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1481 (1) Generate an error if the third argument for the ${prvs expansion is
1484 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1487 (3) Reset the variables that are obtained from the first argument of
1488 ${prvscheck and used in the second argument before leaving the code,
1489 because their memory is reclaimed, so using them afterwards may do
1492 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1493 one (it's much easier than Tom thought :-).
1495 (5) Because of (4), we can now allow for the use of $prvscheck_result
1496 inside the third argument.
1498 PH/03 For some reason, the default setting of PATH when running a command from
1499 a pipe transport was just "/usr/bin". I have changed it to
1502 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1503 anything to be listed in the output from -bV.
1505 PH/05 When a filter generated an autoreply, the entire To: header line was
1506 quoted in the delivery log line, like this:
1508 => >A.N.Other <ano@some.domain> <original@ddress> ...
1510 This has been changed so that it extracts the operative address. There
1511 may be more than one such address. If so, they are comma-separated, like
1514 => >ano@some.domain,ona@other.domain <original@ddress> ...
1516 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1517 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1518 not being looked up in the rDNS to get a host name, Exim was showing the
1519 IP address twice in Received: lines, even though the IP addresses were
1520 identical. For example:
1522 Received: from [1.2.3.4] (helo=[1.2.3.4])
1524 However, if the real host name was known, it was omitting the HELO data
1525 if it matched the actual IP address. This has been tidied up so that it
1526 doesn't show the same IP address twice.
1528 PH/07 When both +timestamp and +memory debugging was on, the value given by
1529 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1530 called by the debug printing, thereby overwriting the timestamp buffer.
1531 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1533 PH/08 When the original message was included in an autoreply transport, it
1534 always said "this is a copy of the message, including all the headers",
1535 even if body_only or headers_only was set. It now gives an appropriate
1538 PH/09 Applied a patch from the Sieve maintainer which:
1540 o fixes some comments
1541 o adds the (disabled) notify extension core
1542 o adds some debug output for the result of if/elsif tests
1543 o points to the current vacation draft in the documentation
1544 and documents the missing references header update
1548 o fixes a bug in processing the envelope test (when testing
1549 multiple envelope elements, the last element determinted the
1552 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1553 Electronic Mail") by including:
1555 Auto-submitted: auto-generated
1557 in the messages that it generates (bounce messages and others, such as
1558 warnings). In the case of bounce messages for non-SMTP mesages, there was
1559 also a typo: it was using "Auto_submitted" (underscore instead of
1560 hyphen). Since every message generated by Exim is necessarily in response
1561 to another message, thes have all been changed to:
1563 Auto-Submitted: auto-replied
1565 in accordance with these statements in the RFC:
1567 The auto-replied keyword:
1569 - SHOULD be used on messages sent in direct response to another
1570 message by an automatic process,
1572 - MUST NOT be used on manually-generated messages,
1574 - MAY be used on Delivery Status Notifications (DSNs) and Message
1575 Disposition Notifications (MDNs),
1577 - MUST NOT be used on messages generated by automatic or periodic
1578 processes, except for messages which are automatic responses to
1581 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1582 to the default Received: header definition.
1584 PH/12 Added log selector acl_warn_skipped (default on).
1586 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1587 variables because (a) they are in the wrong storage pool and (b) even if
1588 they were copied, it wouldn't work properly because of the caching.
1590 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1591 checking when decoding. Apparently there are clients that generate
1592 overlong encoded strings. Why am I not surprised?
1594 PH/15 If the first argument of "${if match_address" was not empty, but did not
1595 contain an "@" character, Exim crashed. Now it writes a panic log message
1596 and treats the condition as false.
1598 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1600 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1601 extension "envelope-auth". The code is finished and in agreement with
1602 other implementations, but there is no documentation so far and in fact,
1603 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1604 not changing the active code.
1606 Print executed "if" and "elsif" statements when debugging is used. This
1607 helps a great deal to understand what a filter does.
1609 Document more things not specified clearly in RFC3028. I had all this
1610 sorted out, when out of a sudden new issues came to my mind. Oops."
1612 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1615 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1616 canonical form (as documented). However, after a host name lookup from
1617 the IP address, check_host() was doing a simple string comparison with
1618 addresses acquired from the DNS when checking that the found name did
1619 have the original IP as one of its addresses. Since any found IPv6
1620 addresses are likely to be in abbreviated form, the comparison could
1621 fail. Luckily, there already exists a function for doing the comparison
1622 by converting both addresses to binary, so now that is used instead of
1623 the text comparison.
1625 PH/20 There was another similar case to PH/19, when a complete host name was
1626 given in a host list; looking up its IP address could give an abbreviated
1627 form, whereas the current host's name might or might not be abbreviated.
1628 The same fix has been applied.
1634 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1635 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1638 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1639 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1641 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1643 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1644 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1645 Also, for "Auto-Submitted", treat anything other than "no" as
1646 non-personal, in accordance with RFC 3834. (Previously it treated
1647 anything starting "auto-" as non-personal.)
1649 TF/01 The control=submission/name=... option had a problem with syntax
1650 errors if the name included a slash character. The /name= option
1651 now slurps the rest of the string, so it can include any characters
1652 but it must come last in the list of options (after /sender_retain
1655 PH/05 Some modifications to the interface to the fake nameserver for the new
1663 TK/01 Added the "success_on_redirect" address verification option. See
1664 NewStuff for rationale and an example.
1666 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1668 PH/02 Patch to exigrep to allow it to work on syslog lines.
1670 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1671 fread() to skip over the body file's header line, because in Cygwin the
1672 header line is locked and is inaccessible.
1674 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1675 co-exist for some time) to make it clear that it is the Exim ID that is
1676 referenced, not the Message-ID: header line.
1678 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1679 string_format() function, because snprintf() does not exist on all
1682 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1683 ${stat: expansion item.
1685 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1686 protocol synchronization error", to keep the pedants happy.
1688 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1689 well as for IRIX systems, when gcc is being used. See the host.c source
1692 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1694 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1697 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1698 transport and to the smtp transport in LMTP mode.
1700 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1702 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1703 run a filter in a subprocess. This could lead to confusion in subsequent
1704 lookups in the parent process. There should also be a search_tidyup() at
1705 the end of the subprocess.
1707 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1708 only if the host matched helo_try_verify_hosts, which caused the
1709 verification to occur when the EHLO/HELO command was issued. The ACL just
1710 tested the remembered result. Now, if a previous verification attempt has
1711 not happened, "verify = helo" does it there and then.
1713 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1715 TK/03 Fix log output including CR from clamd.
1717 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1718 exist provoked a memory error which could cause a segfault.
1720 PH/15 Installed PCRE 6.2
1722 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1724 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1725 of the problem. Specifically, suggested +O2 rather than +O1 for the
1728 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1730 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1731 record, the port was not correctly logged when the outgoing_port log
1732 selector was set (it logged the transort's default port).
1734 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1735 fallback_hosts, and "hosts" in the smtp transport.
1737 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1738 host errors such as "Connection refused".
1740 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1741 authentication with radiusclient 0.4.9:
1743 - Error returned from rc_read_config was caught wrongly
1744 - Username/password not passed on to radius server due to wrong length.
1746 The presumption is that some radiusclient API changes for 4.51/PH/17
1747 were not taken care of correctly. The code is still untested by me (my
1748 Linux distribution still has 0.3.2 of radiusclient), but it was
1749 contributed by a Radius user.
1751 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1752 expanding the "port" option of the smtp transport.
1754 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1755 while reading a MIME header. Thanks to Tom Hughes for a patch.
1757 PH/24 Include config.h inside local_scan.h so that configuration settings are
1760 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1761 that in an ACL for RCPT (for example), you can examine exactly what was
1764 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1765 commands, but it was not correctly comparing the address with the actual
1766 client host address. Thus, it would show the EHLO address in Received:
1767 header lines when this was not necessary.
1769 PH/27 Added the % operator to ${eval:}.
1771 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1772 it should be ignoring failures (because with -C, for example, it has lost
1773 privilege). It wasn't ignoring creation failures other than "already
1776 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1779 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1780 because an input line was too long, either on its own, or by virtue of
1781 too many continuations, the temporary file was not being removed, and the
1782 return code was incorrect.
1784 PH/31 Missing "BOOL" in function definition in filtertest.c.
1786 PH/32 Applied Sieve patches from the maintainer.
1788 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1790 PH/33 Added "verify = not_blind".
1792 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1793 Local/Makefile (with some defaults set). These are used in built scripts
1794 such as exicyclog, but they have never been used in the exim_install
1795 script (though there are many overriding facilities there). I have
1796 arranged that the exim_install script now takes note of these two
1799 PH/35 Installed configuration files for Dragonfly.
1801 PH/36 When a locally submitted message by a trusted user did not contain a
1802 From: header, and the sender address was obtained from -f or from an SMTP
1803 MAIL command, and the trusted user did not use -F to supply a sender
1804 name, $originator_name was incorrectly used when constructing a From:
1805 header. Furthermore, $originator_name was used for submission mode
1806 messages from external hosts without From: headers in a similar way,
1807 which is clearly wrong.
1809 PH/37 Added control=suppress_local_fixups.
1811 PH/38 When log_selector = +received_sender was set, and the addition of the
1812 sender made the log line's construction buffer exactly full, or one byte
1813 less than full, an overflow happened when the terminating "\n" was
1816 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1817 when the result of a list match is failure because a DNS lookup failed.
1819 PH/40 RM_COMMAND is now used in the building process.
1821 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1822 the "build-* directories that it finds.
1824 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1825 address in a domain literal was a prefix of an interface address.
1827 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1828 when verifying a sender address, unless rewrite_headers is false.
1830 PH/44 Wrote a long comment about why errors_to addresses are verified as
1831 recipients, not senders.
1833 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1834 the ratelimit ACL was added.
1836 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1838 PH/46 Added extra information about PostgreSQL errors to the error string.
1840 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1841 avoiding the need to install special zones in a real server. This is
1842 backwards compatible; if it can't find the fake resolver, it drops back.
1843 Thus, both old and new test suites can be run.
1845 TF/02 Added util/ratelimit.pl
1847 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1848 clock is set back in time.
1850 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1851 Candler <B.Candler@pobox.com>.
1853 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1854 OK for addresses that are the result of redirections.
1856 PH/48 A number of further additions for the benefit of the new test suite,
1857 including a fake gethostbyname() that interfaces to the fake DNS resolver
1860 TF/06 The fix for widen_domains has also been applied to qualify_single and
1861 search_parents which are the other dnslookup options that can cause
1864 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1867 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1869 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1870 service" if it encountered an MX record with an empty target hostname.
1871 The message is now "an MX or SRV record indicated no SMTP service".
1873 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1874 if the DNS of the sending site is misconfigured. This is quite a
1875 common situation. This change restores the behaviour of treating a
1876 helo verification defer as a failure.
1878 PH/51 If self=fail was set on a router, the bounce message did not include the
1879 actual error message.
1885 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1887 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1888 command itself ended in error, the underlying message about the transport
1889 filter timeout was being overwritten with the pipe command error. Now the
1890 underlying error message should be appended to the second error message.
1892 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1894 PH/02 Reduce the amount of output that "make" produces by default. Full output
1895 can still be requested.
1897 PH/03 The warning log line about a condition test deferring for a "warn" verb
1898 was being output only once per connection, rather than after each
1899 occurrence (because it was using the same function as for successful
1900 "warn" verbs). This seems wrong, so I have changed it.
1902 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1903 it should not have, which might have caused a crash in the right
1904 circumstances, but probably never did.
1906 PH/04 Installed a modified version of Tony Finch's patch to make submission
1907 mode fix the return path as well as the Sender: header line, and to
1908 add a /name= option so that you can make the user's friendly name appear
1911 TF/03 Added the control = fakedefer ACL modifier.
1913 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1914 Mark Lowes for thorough testing.
1916 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1918 TK/03 Merged latest SRS patch from Miles Wilton.
1920 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1921 with the definition in sysexits.h (which is #included earlier).
1922 Fortunately, Exim does not actually use EX_OK. The code used to try to
1923 preserve the sysexits.h value, by assumimg that macro definitions were
1924 scanned for macro replacements. I have been disabused of this notion,
1925 so now the code just undefines EX_OK before #including unistd.h.
1927 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1928 in the smtp transport. When a block could not be written in a single
1929 write() function, the timeout was being re-applied to each part-write.
1930 This seems wrong - if the receiver was accepting one byte at a time it
1931 would take for ever. The timeout is now adjusted when this happens. It
1932 doesn't have to be particularly precise.
1934 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1935 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1937 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1938 without a domain if the check to <postmaster@domain> fails.
1940 SC/01 Eximstats: added -xls and the ability to specify output files
1941 (patch written by Frank Heydlauf).
1943 SC/02 Eximstats: use FileHandles for outputing results.
1945 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1947 SC/04 Eximstats: fixed display of large numbers with -nvr option
1949 SC/05 Eximstats: fixed merging of reports with empty tables.
1951 SC/06 Eximstats: added the -include_original_destination flag
1953 SC/07 Eximstats: removed tabs and trailing whitespace.
1955 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1957 TK/06 MBOX spool code: Add real "From " MBOX separator line
1958 so the .eml file is really in mbox format (even though
1959 most programs do not really care). Patch from Alex Miller.
1961 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1962 The latter is generated from $received_to and is only set if the
1963 message has one envelope recipient. SA can use these headers,
1964 obviously out-of-the-box. Patch from Alex Miller.
1966 PH/08 The ${def test on a variable was returning false if the variable's
1967 value was "0", contrary to what the specification has always said!
1968 The result should be true unless the variable is empty.
1970 PH/09 The syntax error of a character other than { following "${if
1971 def:variable_name" (after optional whitespace) was not being diagnosed.
1972 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1973 accidental colon was present, for example, could give incorrect results.
1975 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1976 result is used (not including appendfile, where other changes are about
1979 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1980 This involved changing a lot of size variables from int to off_t. It
1981 should work with maildirs and everything.
1983 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1984 spamd dying while we are connected to it.
1986 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1989 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1990 function for BDB 4.3.
1992 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1994 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1995 the sources for PCRE have been re-arranged and more files are now
1998 PH/15 The code I had for printing potentially long long variables in PH/11
1999 above was not the best (it lost precision). The length of off_t variables
2000 is now inspected at build time, and an appropriate printing format (%ld
2001 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2002 to be "long long int" or "long int". This is needed for the internal
2003 formatting function string_vformat().
2005 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2006 the configuration file to be ":syslog", then the script "guesses" where
2007 the logs files are, rather than using the compiled in default. In our
2008 case the guess is not the same as the compiled default, so the script
2009 suddenly stopped working when I started to use syslog. The patch checks
2010 to see if log_file_path is "". If so, it attempts to read it from exim
2011 with no configuration file to get the compiled in version, before it
2012 falls back to the previous guessing code."
2014 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2015 implementing BATV in an Exim configuration. See NewStuff for the gory
2018 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2019 Makefile that are specific to HP-UX.
2021 PH/18 If the "use_postmaster" option was set for a recipient callout together
2022 with the "random" option, the postmaster address was used as the MAIL
2023 FROM address for the random test, but not for the subsequent recipient
2024 test. It is now used for both.
2026 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2027 patch removes a few documentation additions to RFC 3028, because the
2028 latest draft now contains them. It adds the new en;ascii-case comparator
2029 and a new error check for 8bit text in MIME parts. Comparator and
2030 require names are now matched exactly. I enabled the subaddress
2031 extension, but it is not well tested yet (read: it works for me)."
2033 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2034 rework some of the code of TK/09 above to avoid the hardwired use of
2035 "%lld" and "long long". Replaced the call to snprintf() with a call to
2038 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2039 records point to non-existent hosts", "retry timeout exceeded", and
2040 "retry time not reached for any host after a long failure period".
2042 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2043 experimental DomainKeys support:
2045 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2046 (2) On an error such as an illegally used "control", the wrong name for
2047 the control was given.
2049 These problems did NOT occur unless DomainKeys support was compiled.
2051 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2053 PH/24 Added ${if match_ip condition.
2055 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2056 databases so that it will be absolutely obvious if a crash occurs in the
2057 DB library. This is a regular occurrence (often caused by mis-matched
2060 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2061 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2062 picked up on a user's system that detects such things. There doesn't seem
2063 to be a gcc warning option for this - only an attribute that has to be
2064 put on the function's prototype. It seems that in Fedora Core 4 they have
2065 set this on a number of new functions. No doubt there will be more in due
2068 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2069 specify a transport. However, if an address that was verified by such a
2070 router was the subject of a callout, Exim crashed because it tried to
2071 read the rcpt_include_affixes from the non-existent transport. Now it
2072 just assumes that the setting of that option is false. This bug was
2073 introduced by 4.51/PH/31.
2075 PH/28 Changed -d+all to exclude +memory, because that information is very
2076 rarely of interest, but it makes the output a lot bigger. People tend to
2077 do -d+all out of habit.
2079 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2080 code in os-type was giving problems when libc.so lives in lib64, like on
2083 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2084 aren't the modern standard, and it seems that some systems' include files
2085 don't always have them. Exim was already checking for some of the newer
2086 ones like T_AAAA, and defining it itself. I've added checks for all the
2087 record types that Exim uses.
2089 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2090 not automatically generating a new one, as it is supposed to. This
2091 prevented TLS from working. If the file did exist, but contained invalid
2092 data, a new version was generated, as expected. It was only the case of a
2093 non-existent file that was broken.
2095 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2096 with a change in libdomainkeys > 0.64.
2098 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2099 from DNS. If the selector record carries the flag, it now has
2100 precedence over the domain-wide flag.
2102 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2104 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2105 the use of a port name that isn't defined in /etc/services) occurred, the
2106 message was deferred as in a normal delivery, and thus remained on the
2107 spool, instead of being failed because of the mua_wrapper setting. This
2108 is now fixed, and I tidied up some of the mua_wrapper messages at the
2111 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2112 the messages in a hash of arrays rather than using individual hashes.
2113 This is a bit cleaner and results in dramatic memory savings, albeit
2114 at a slight CPU cost.
2116 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2117 as requested by Marc Sherman.
2119 SC/10 Eximstats: added histograms for user specified patterns as requested
2122 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2124 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2125 fopen() in the content-scanning modules that did not already have it.
2131 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2132 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2134 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2136 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2138 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2139 to test Sieve filters that use "vacation".
2141 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2142 that changes the way the GnuTLS parameters are stored in the cache file.
2143 The new format can be generated externally. For backward compatibility,
2144 if the data in the cache doesn't make sense, Exim assumes it has read an
2145 old-format file, and it generates new data and writes a new file. This
2146 means that you can't go back to an older release without removing the
2149 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2150 work if there are any delivery delays because "one_time" forces the
2151 parent to be marked "delivered", so its unseen clone is never tried
2152 again. For this reason, Exim now forbids the simultaneous setting of
2155 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2156 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2157 entry if you want to know the details. The fix, however, neglected to
2158 consider the case when local delivery batching is involved. The test for
2159 "previously delivered" was not happening when checking to see if an
2160 address could be batched with a previous (undelivered) one; under
2161 certain circumstances this could lead to multiple deliveries to the same
2164 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2165 in its include files, and this causes problems building Exim.
2167 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2168 header_syntax) but Exim was just ignoring anything given after a slash.
2169 In particular, this caused confusion with an attempt to use "verify =
2170 reverse_host_lookup/defer_ok". An error is now given when options are
2171 supplied for verify items that do not have them. (Maybe reverse_host_
2172 lookup should have a defer_ok option, but that's a different point.)
2174 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2175 defined by RFC 821) to 2048, because there were problems with some AUTH
2176 commands, and RFC 1869 says the size should be increased for extended
2177 SMTP commands that take arguments.
2179 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2182 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2183 "unknown" error; now it says that the functionality isn't in the binary.
2185 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2186 an address' error message when a string expansion fails (syntax or
2187 whatever). Otherwise the password may appear in the log. Following change
2188 PH/42 below, there is no longer a chance of it appearing in a bounce
2191 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2193 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2194 of its addresses was ever tried. (Bugzilla bug #2.)
2196 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2197 the result incorrectly in the debug output. (It correctly added a newline
2198 to what was transported.)
2200 TF/01 Added $received_time.
2202 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2203 commented out examples of how to interface to a virus scanner and to
2204 SpamAssassin. Also added commented examples of av_scanner and
2205 spamd_address settings.
2207 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2208 and controls are allowed in which ACLs. There were a couple of minor
2209 errors. Some of the entries in the conditions table (which is a table of
2210 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2211 negation of where the condition IS allowed.
2213 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2215 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2216 header file does not have a version number, so I've had to invent a new
2217 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2218 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2219 radiusclient), but it was contributed by a Radius user.
2221 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2222 files or directories, for OpenSSL.
2224 PH/19 When an Exim process that is running as root has to create an Exim log
2225 file, it does so in a subprocess that runs as exim:exim so as to get the
2226 ownership right at creation (otherwise, other Exim processes might see
2227 the file with the wrong ownership). There was no test for failure of this
2228 fork() call, which would lead to the process getting stuck as it waited
2229 for a non-existent subprocess. Forks do occasionally fail when resources
2230 run out. I reviewed all the other calls to fork(); they all seem to check
2233 PH/20 When checking for unexpected SMTP input at connect time (before writing
2234 the banner), Exim was not dealing correctly with a non-positive return
2235 from the read() function. If the client had disconnected by this time,
2236 the result was a log entry for a synchronization error with an empty
2237 string after "input=" when read() returned zero. If read() returned -1
2238 (an event I could not check), uninitialized data bytes were printed.
2239 There were reports of junk text (parts of files, etc) appearing after
2242 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2244 PH/22 Added support for macro redefinition, and (re)definition in between
2245 driver and ACL definitions.
2247 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2248 forgetting to use the resulting value; it was using the unexpanded value.
2250 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2251 hadn't been configured. The fix is from Juergen Kreileder, who
2252 understands it better than I do:
2254 "Here's what I see happening with three configured cyrus_sasl
2255 authenticators configured (plain, login, cram-md5):
2257 On startup auth_cyrus_sasl_init() gets called for each of these.
2258 This means three calls to sasl_listmech() without a specified mech_list.
2259 => SASL tests which mechs of all available mechs actually work
2260 => three warnings about OTP not working
2261 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2263 With the patch, sasl_listmech() also gets called three times. But now
2264 SASL's mech_list option is set to the server_mech specified in the the
2265 authenticator. Or in other words, the answer from sasl_listmech()
2266 gets limited to just the mech you're testing for (which is different
2268 => the return list contains just 'plain' or 'login', 'cram-md5' or
2269 nothing depending on the value of ob->server_mech.
2271 I've just tested the patch: Authentication still works fine,
2272 unavailable mechs specified in the exim configuration are still
2273 caught, and the auth.log warnings about OTP are gone."
2275 PH/25 When debugging is enabled, the contents of the command line are added
2276 to the debugging output, even when log_selector=+arguments is not
2279 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2280 answer is "GNU", and only if the return is "GNU/something" is the answer
2283 PH/27 $acl_verify_message is now set immediately after the failure of a
2284 verification in an ACL, and so is available in subsequent modifiers. In
2285 particular, the message can be preserved by coding like this:
2287 warn !verify = sender
2288 set acl_m0 = $acl_verify_message
2290 Previously, $acl_verify_message was set only while expanding "message"
2291 and "log_message" when a very denied access.
2293 PH/28 Modified OS/os.c-Linux with
2295 -#ifndef OS_LOAD_AVERAGE
2296 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2298 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2299 nomenclature these days.)
2301 PH/29 Installed patch from the Sieve maintainer that adds the options
2302 sieve_useraddress and sieve_subaddress to the redirect router.
2304 PH/30 In these circumstances:
2305 . Two addresses routed to the same list of hosts;
2306 . First host does not offer TLS;
2307 . First host accepts first address;
2308 . First host gives temporary error to second address;
2309 . Second host offers TLS and a TLS session is established;
2310 . Second host accepts second address.
2311 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2312 and peerdn, if requested) that were in fact used only for the second
2315 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2316 attention to any local part prefix or suffix that was matched by the
2317 router that accepted the address. It now behaves in the same way as it
2318 does for delivery: the affixes are removed from the local part unless
2319 rcpt_include_affixes is set on the transport.
2321 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2322 timeout during the DATA phase of an incoming message.
2324 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2325 applied a patch sanctioned by the Sieve maintainer.
2327 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2328 the uid or gid is negative. A case of a negative gid caused this to be
2329 noticed. The fix allows for either to be negative.
2331 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2332 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2335 PH/36 The change PH/12 above was broken. Fixed it.
2337 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2338 the grounds that routing the same address twice would always produce the
2339 same answer. This might have been true once, but it is certainly no
2340 longer true now. Routing a child address may depend on the previous
2341 routing that produced that child. Some complicated redirection strategies
2342 went wrong when messages had multiple recipients, and made Exim's
2343 behaviour dependent on the order in which the addresses were given.
2345 I have moved the duplicate checking until after the routing is complete.
2346 Exim scans the addresses that are assigned to local and remote
2347 transports, and removes any duplicates. This means that more work will be
2348 done, as duplicates will always all be routed, but duplicates are
2349 presumably rare, so I don't expect this is of any significance.
2351 For deliveries to pipes, files, and autoreplies, the duplicate checking
2352 still happens during the routing process, since they are not going to be
2355 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2356 It corrects a timeout issue with spamd. This is Ian's comment: "The
2357 background is that sometimes spamd either never reads data from a
2358 connection it has accepted, or it never writes response data. The exiscan
2359 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2360 blindly assumes that writes won't block so it may never time out."
2362 PH/39 Allow G after quota size as well as K and M.
2364 PH/40 The value set for $authenticated_id in an authenticator may not contain
2365 binary zeroes or newlines because the value is written to log lines and
2366 to spool files. There was no check on this. Now the value is run through
2367 the string_printing() function so that such characters are converted to
2368 printable escape sequences.
2370 PH/41 $message_linecount is a new variable that contains the total number of
2371 lines in the message. Compare $body_linecount, which is the count for the
2374 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2375 bounce and delay warning messages, except in certain special cases, which
2378 (a) An SMTP error message from a remote host;
2379 (b) A message specified in a :fail: redirection;
2380 (c) A message specified in a "fail" command in a system filter;
2381 (d) A message specified in a FAIL return from the queryprogram router;
2382 (e) A message specified by the cannot_route_message router option.
2384 In these cases only, Exim does include the error details in bounce and
2385 warning messages. There are also a few cases where bland messages such
2386 as "unrouteable address" or "local delivery error" are given.
2388 PH/43 $value is now also set for the "else" part of a ${run expansion.
2390 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2391 being worked on, but at least Exim now implements the latest version to
2394 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2395 process to complete was treated as a delivery failure, a timeout while
2396 writing the message to the pipe was logged, but erroneously treated as a
2397 successful delivery. Such timeouts include transport filter timeouts. For
2398 consistency with the overall process timeout, these timeouts are now
2399 treated as errors, giving rise to delivery failures by default. However,
2400 there is now a new Boolean option for the pipe transport called
2401 timeout_defer, which, if set TRUE, converts the failures into defers for
2402 both kinds of timeout. A transport filter timeout is now identified in
2405 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2406 systems where "make" and "gmake" are different, calling "gmake" at top
2407 level broke things. I've arranged for the value of $(MAKE) to be passed
2408 from the Makefile to this script so that it can call the same version of
2412 A note about Exim versions 4.44 and 4.50
2413 ----------------------------------------
2415 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2416 changes of various kinds. As a consequence, a big documentation update was
2417 needed. This delayed the release for rather longer than seemed good, especially
2418 in the light of a couple of (minor) security issues. Therefore, the changes
2419 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2420 release. So 4.44 and 4.50 are in effect two different branches that both start
2423 I have left the 4.50 change log unchanged; it contains all the changes since
2424 4.43. The change log for 4.44 is below; many of its items are identical to
2425 those for 4.50. This seems to be the most sensible way to preserve the
2426 historical information.
2432 1. Minor wording change to the doc/README.SIEVE file.
2434 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2435 computation of the current number of files was incorrect.
2437 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2438 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2441 4. Give more explanation in the error message when the command for a transport
2442 filter fails to execute.
2444 5. There are several places where Exim runs a non-Exim command in a
2445 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2446 was being done only for the command run by the queryprogram router. It is
2447 now done for all such subprocesses. The other cases are: ${run, transport
2448 filters, and the commands run by the lmtp and pipe transports.
2450 6. Added CONFIGURE_GROUP build-time option.
2452 7. Some older OS have a limit of 256 on the maximum number of file
2453 descriptors. Exim was using setrlimit() to set 1000 as a large value
2454 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2455 systems. I've change it so that if it can't get 1000, it tries for 256.
2457 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2458 was an oversight, and furthermore, ever since the addition of extra
2459 controls (e.g. 4.43/32), the checks on when to allow different forms of
2460 "control" were broken. There should now be diagnostics for all cases when a
2461 control that does not make sense is encountered.
2463 9. Added the /retain_sender option to "control=submission".
2465 10. $recipients is now available in the predata ACL (oversight).
2467 11. Tidy the search cache before the fork to do a delivery from a message
2468 received from the command line. Otherwise the child will trigger a lookup
2469 failure and thereby defer the delivery if it tries to use (for example) a
2470 cached ldap connection that the parent has called unbind on.
2472 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2473 of $address_data from the recipient verification was clobbered by the
2474 sender verification.
2476 13. The value of address_data from a sender verification is now available in
2477 $sender_address_data in subsequent conditions in the ACL statement.
2479 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2481 15. Added a new option "connect=<time>" to callout options, to set a different
2484 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2485 was its contents. (It was OK if the option was not defined at all.)
2487 17. A "Completed" log line is now written for messages that are removed from
2488 the spool by the -Mrm option.
2490 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2491 information about exactly what failed.
2493 19. Added -dd to debug only the daemon process.
2495 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2496 handles timeouts, both on the server side and network timeouts. Renamed the
2497 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2499 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2500 It is now set to "smtps".
2502 22. $host_address is now set to the target address during the checking of
2503 ignore_target_hosts.
2505 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2506 being passed; this would have caused $sender_host_name to have been used if
2507 matching the list had actually called for a host name (not very likely,
2508 since this list is usually IP addresses). A host name is now passed as
2511 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2512 code that specifies a non-restarting handler (typically sigaction() in
2513 modern systems) in an attempt to fix a rare and obscure crash bug.
2515 25. Narrowed the window for a race in the daemon that could cause it to ignore
2516 SIGCHLD signals. This is not a major problem, because they are used only to
2517 wake it up if nothing else does.
2519 26. A malformed maildirsize file could cause Exim to calculate negative values
2520 for the mailbox size or file count. Odd effects could occur as a result.
2521 The maildirsize information is now recalculated if the size or filecount
2524 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2525 support for a long time. Removed HAVE_SYS_VFS_H.
2527 28. Installed the latest version of exipick from John Jetmore.
2529 29. In an address list, if the pattern was not a regular expression, an empty
2530 subject address (from a bounce message) matched only if the pattern was an
2531 empty string. Non-empty patterns were not even tested. This was the wrong
2532 because it is perfectly reasonable to use an empty address as part of a
2533 database query. An empty address is now tested by patterns that are
2534 lookups. However, all the other forms of pattern expect the subject to
2535 contain a local part and a domain, and therefore, for them, an empty
2536 address still always fails if the pattern is not itself empty.
2538 30. Exim went into a mad DNS loop when attempting to do a callout where the
2539 host was specified on an smtp transport, and looking it up yielded more
2540 than one IP address.
2542 31. Re-factored the code for checking spool and log partition space into a
2543 function that finds that data and another that does the check. The former
2544 is then used to implement four new variables: $spool_space, $log_space,
2545 $spool_inodes, and $log_inodes.
2547 32. The RFC2047 encoding function was originally intended for short strings
2548 such as real names; it was not keeping to the 75-character limit for
2549 encoded words that the RFC imposes. It now respects the limit, and
2550 generates multiple encoded words if necessary. To be on the safe side, I
2551 have increased the buffer size for the ${rfc2047: expansion operator from
2554 33. It is now permitted to omit both strings after an "if" condition; if the
2555 condition is true, the result is "true". As before, when the second string
2556 is omitted, a false condition yields an empty string. This makes it less
2557 cumbersome to write custom ACL and router conditions.
2559 34. Failure to deliver a bounce message always caused it to be frozen, even if
2560 there was an errors_to setting on the router. The errors_to setting is now
2563 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2564 canonical form (fully expanded) before being placed in
2565 $sender_host_address.
2567 36. The table in the code that translates DNS record types into text (T_A to
2568 "A" for instance) was missing entries for NS and CNAME. It is just possible
2569 that this could have caused confusion if both these types were looked up
2570 for the same domain, because the text type is used as part of Exim's
2571 per-process caching. But the chance of anyone hitting this buglet seems
2574 37. The dnsdb lookup has been extended in a number of ways.
2576 (1) There is a new type, "zns", which walks up the domain tree until it
2577 finds some nameserver records. It should be used with care.
2579 (2) There is a new type, "mxh", which is like "mx" except that it returns
2580 just the host names, not the priorities.
2582 (3) It is now possible to give a list of domains (or IP addresses) to be
2583 looked up. The behaviour when one of the lookups defers can be
2584 controlled by a keyword.
2586 (4) It is now possible to specify the separator character for use when
2587 multiple records are returned.
2589 38. The dnslists ACL condition has been extended: it is now possible to supply
2590 a list of IP addresses and/or domains to be looked up in a particular DNS
2593 39. Added log_selector=+queue_time_overall.
2595 40. When running the queue in the test harness, wait just a tad after forking a
2596 delivery process, to get repeatability of debugging output.
2598 41. Include certificate and key file names in error message when GnuTLS fails
2599 to set them up, because the GnuTLS error message doesn't include the name
2600 of the failing file when there is a problem reading it.
2602 42. Allow both -bf and -bF in the same test run.
2604 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2606 44. The "Exiscan patch" is now merged into the mainline Exim source.
2608 45. Sometimes the final signoff response after QUIT could fail to get
2609 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2610 < 0 before doing a fflush(). This bug looks as though it goes back to the
2611 introduction of TLS in release 3.20, but "sometimes" must have been rare
2612 because the tests only now provoked it.
2614 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2615 (this can affect the format of dates).
2617 47. exim_tidydb, when checking for the continued existence of a message for
2618 which it has found a message-specific retry record, was not finding
2619 messages that were in split spool directories. Consequently, it was
2620 deleting retry records that should have stayed in existence.
2622 48. Steve fixed some bugs in eximstats.
2624 49. The SPA authentication driver was not abandoning authentication and moving
2625 on to the next authenticator when an expansion was forced to fail,
2626 contradicting the general specification for all authenticators. Instead it
2627 was generating a temporary error. It now behaves as specified.
2629 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2630 (the order specifies the preference for clients). The order is now AES256,
2631 AES128, 3DES, ARCFOUR128.
2633 51. Small patch to Sieve code - explicitly set From: when generating an
2636 52. Exim crashed if a remote delivery caused a very long error message to be
2637 recorded - for instance if somebody sent an entire SpamAssassin report back
2638 as a large number of 550 error lines. This bug was coincidentally fixed by
2639 increasing the size of one of Exim's internal buffers (big_buffer) that
2640 happened as part of the Exiscan merge. However, to be on the safe side, I
2641 have made the code more robust (and fixed the comments that describe what
2644 53. Now that there can be additional text after "Completed" in log lines (if
2645 the queue_time_overall log selector is set), a one-byte patch to exigrep
2646 was needed to allow it to recognize "Completed" as not the last thing in
2649 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2650 patch that reportedly fixes this has been added. I am not expert enough to
2651 create a test for it. This is what the patch creator wrote:
2653 "I found a little strange behaviour of ldap code when working with
2654 Windows 2003 AD Domain, where users was placed in more than one
2655 Organization Units. When I tried to give exim partial DN, the exit code
2656 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2657 But simultaneously result of request was absolutely normal ldap result,
2658 so I produce this patch..."
2660 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2661 REFERENCE, so I have modified the code to exclude the patch when that macro
2664 55. Some experimental protocols are using DNS PTR records for new purposes. The
2665 keys for these records are domain names, not reversed IP addresses. The
2666 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2667 leaves it alone. Component reversal etc. now happens only for IP addresses.
2670 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2672 57. Double the size of the debug message buffer (to 2048) so that more of very
2673 long debug lines gets shown.
2675 58. The exicyclog utility now does better if the number of log files to keep
2676 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2678 59. Two changes related to the smtp_active_hostname option:
2680 (1) $smtp_active_hostname is now available as a variable.
2681 (2) The default for smtp_banner uses $smtp_active_hostname instead
2682 of $primary_hostname.
2684 60. The host_aton() function is supposed to be passed a string that is known
2685 to be a valid IP address. However, in the case of IPv6 addresses, it was
2686 not checking this. This is a hostage to fortune. Exim now panics and dies
2687 if the condition is not met. A case was found where this could be provoked
2688 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2689 components; fortuitously, this particular loophole had already been fixed
2690 by change 4.50/55 above.
2692 If there are any other similar loopholes, the new check in host_aton()
2693 itself should stop them being exploited. The report I received stated that
2694 data on the command line could provoke the exploit when Exim was running as
2695 exim, but did not say which command line option was involved. All I could
2696 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2697 running as the user.
2700 61. There was a buffer overflow vulnerability in the SPA authentication code
2701 (which came originally from the Samba project). I have added a test to the
2702 spa_base64_to_bits() function which I hope fixes it.
2705 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2706 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2708 63. The daemon start-up calls getloadavg() while still root for those OS that
2709 need the first call to be done as root, but it missed one case: when
2710 deliver_queue_load_max is set with deliver_drop_privilege. This is
2711 necessary for the benefit of the queue runner, because there is no re-exec
2712 when deliver_drop_privilege is set.
2714 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2715 This has been fixed.
2717 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2718 was in use, was not putting the data itself into the right store pool;
2719 consequently, it could be overwritten for a subsequent message in the same
2720 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2723 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2725 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2726 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2727 the function were treating the return as a boolean value, which happened to
2728 work because 0=false and not-0=true, but is not correct code.
2730 68. The host_aton() function was not handling scoped IPv6 addresses (those
2731 with, for example, "%eth0" on the end) correctly.
2733 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2734 negated items (that is, ~something) in unsigned ints. Some compilers
2735 apparently mutter when there is no cast.
2737 70. If an address verification called from an ACL failed, and did not produce a
2738 user-specific message (i.e. there was only a "system" message), nothing was
2739 put in $acl_verify_message. In this situation, it now puts the system
2742 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2743 session; change 4.31/43 added the unwanted input to the log line - except
2744 that it did not do this in the start of session case. It now does.
2746 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2747 This is wrong and can cause the other end to generate a synchronization
2748 error if it is another Exim or anything else that does the synchronization
2749 check. A QUIT command is no longer sent after a timeout.
2751 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2752 during host lookups.
2754 74. The defer_ok option of callout verification was not working if it was used
2755 when verifying addresses in header lines, that is, for this case:
2757 verify = header_sender/callout=defer_ok
2759 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2760 those file descriptors could be used for SMTP connections. If anything
2761 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2762 could be sent to the SMTP client, causing chaos. The daemon now opens
2763 stdin, stdout, and stderr to /dev/null when it puts itself into the
2766 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2767 log by default. The user can override this with suitable Perl magic.
2769 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2770 the log message when discard triggers, was not working for the DATA ACL or
2771 for the non-SMTP ACL.
2773 78. Error message wording change in sieve.c.
2775 79. If smtp_accept_max_per_host was set, the number of connections could be
2776 restricted to fewer than expected, because the daemon was trying to set up
2777 a new connection before checking whether the processes handling previous
2778 connections had finished. The check for completed processes is now done
2779 earlier. On busy systems, this bug wouldn't be noticed because something
2780 else would have woken the daemon, and it would have reaped the completed
2783 80. If a message was submitted locally by a user whose login name contained one
2784 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2785 It caused a spool format error. I have fixed the spool reading code. A
2786 related problem was that the "from" clause in the Received: line became
2787 illegal because of the space(s). It is now covered by ${quote_local_part.
2789 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2792 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2794 83. Patch from Sieve maintainer to fix unterminated string problem in
2795 "vacation" handling.
2797 84. Some minor changes to the Linux configuration files to help with other
2798 OS variants using glibc.
2800 85. One more patch for Sieve to update vacation handling to latest spec.
2803 ----------------------------------------------------
2804 See the note above about the 4.44 and 4.50 releases.
2805 ----------------------------------------------------
2811 1. Change 4.43/35 introduced a bug that caused file counts to be
2812 incorrectly computed when quota_filecount was set in an appendfile
2815 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2816 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2819 3. Give more explanation in the error message when the command for a transport
2820 filter fails to execute.
2822 4. There are several places where Exim runs a non-Exim command in a
2823 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2824 was being done only for the command run by the queryprogram router. It is
2825 now done for all such subprocesses. The other cases are: ${run, transport
2826 filters, and the commands run by the lmtp and pipe transports.
2828 5. Some older OS have a limit of 256 on the maximum number of file
2829 descriptors. Exim was using setrlimit() to set 1000 as a large value
2830 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2831 systems. I've change it so that if it can't get 1000, it tries for 256.
2833 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2834 was an oversight, and furthermore, ever since the addition of extra
2835 controls (e.g. 4.43/32), the checks on when to allow different forms of
2836 "control" were broken. There should now be diagnostics for all cases when a
2837 control that does not make sense is encountered.
2839 7. $recipients is now available in the predata ACL (oversight).
2841 8. Tidy the search cache before the fork to do a delivery from a message
2842 received from the command line. Otherwise the child will trigger a lookup
2843 failure and thereby defer the delivery if it tries to use (for example) a
2844 cached ldap connection that the parent has called unbind on.
2846 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2847 of $address_data from the recipient verification was clobbered by the
2848 sender verification.
2850 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2851 was its contents. (It was OK if the option was not defined at all.)
2853 11. A "Completed" log line is now written for messages that are removed from
2854 the spool by the -Mrm option.
2856 12. $host_address is now set to the target address during the checking of
2857 ignore_target_hosts.
2859 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2860 being passed; this would have caused $sender_host_name to have been used if
2861 matching the list had actually called for a host name (not very likely,
2862 since this list is usually IP addresses). A host name is now passed as
2865 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2866 code that specifies a non-restarting handler (typically sigaction() in
2867 modern systems) in an attempt to fix a rare and obscure crash bug.
2869 15. Narrowed the window for a race in the daemon that could cause it to ignore
2870 SIGCHLD signals. This is not a major problem, because they are used only to
2871 wake it up if nothing else does.
2873 16. A malformed maildirsize file could cause Exim to calculate negative values
2874 for the mailbox size or file count. Odd effects could occur as a result.
2875 The maildirsize information is now recalculated if the size or filecount
2878 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2879 support for a long time. Removed HAVE_SYS_VFS_H.
2881 18. Updated exipick to current release from John Jetmore.
2883 19. Allow an empty sender to be matched against a lookup in an address list.
2884 Previously the only cases considered were a regular expression, or an
2887 20. Exim went into a mad DNS lookup loop when doing a callout where the
2888 host was specified on the transport, if the DNS lookup yielded more than
2891 21. The RFC2047 encoding function was originally intended for short strings
2892 such as real names; it was not keeping to the 75-character limit for
2893 encoded words that the RFC imposes. It now respects the limit, and
2894 generates multiple encoded words if necessary. To be on the safe side, I
2895 have increased the buffer size for the ${rfc2047: expansion operator from
2898 22. Failure to deliver a bounce message always caused it to be frozen, even if
2899 there was an errors_to setting on the router. The errors_to setting is now
2902 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2903 canonical form (fully expanded) before being placed in
2904 $sender_host_address.
2906 24. Updated eximstats to version 1.33
2908 25. Include certificate and key file names in error message when GnuTLS fails
2909 to set them up, because the GnuTLS error message doesn't include the name
2910 of the failing file when there is a problem reading it.
2912 26. Expand error message when OpenSSL has problems setting up cert/key files.
2915 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2916 (this can affect the format of dates).
2918 28. exim_tidydb, when checking for the continued existence of a message for
2919 which it has found a message-specific retry record, was not finding
2920 messages that were in split spool directories. Consequently, it was
2921 deleting retry records that should have stayed in existence.
2923 29. eximstats updated to version 1.35
2924 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2925 1.35 - bugfix such that pie charts by volume are generated correctly
2927 30. The SPA authentication driver was not abandoning authentication and moving
2928 on to the next authenticator when an expansion was forced to fail,
2929 contradicting the general specification for all authenticators. Instead it
2930 was generating a temporary error. It now behaves as specified.
2932 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2933 (the order specifies the preference for clients). The order is now AES256,
2934 AES128, 3DES, ARCFOUR128.
2936 31. Small patch to Sieve code - explicitly set From: when generating an
2939 32. Exim crashed if a remote delivery caused a very long error message to be
2940 recorded - for instance if somebody sent an entire SpamAssassin report back
2941 as a large number of 550 error lines. This bug was coincidentally fixed by
2942 increasing the size of one of Exim's internal buffers (big_buffer) that
2943 happened as part of the Exiscan merge. However, to be on the safe side, I
2944 have made the code more robust (and fixed the comments that describe what
2947 33. Some experimental protocols are using DNS PTR records for new purposes. The
2948 keys for these records are domain names, not reversed IP addresses. The
2949 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2950 leaves it alone. Component reversal etc. now happens only for IP addresses.
2953 34. The host_aton() function is supposed to be passed a string that is known
2954 to be a valid IP address. However, in the case of IPv6 addresses, it was
2955 not checking this. This is a hostage to fortune. Exim now panics and dies
2956 if the condition is not met. A case was found where this could be provoked
2957 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2958 components; fortuitously, this particular loophole had already been fixed
2959 by change 4.50/55 or 4.44/33 above.
2961 If there are any other similar loopholes, the new check in host_aton()
2962 itself should stop them being exploited. The report I received stated that
2963 data on the command line could provoke the exploit when Exim was running as
2964 exim, but did not say which command line option was involved. All I could
2965 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2966 running as the user.
2969 35. There was a buffer overflow vulnerability in the SPA authentication code
2970 (which came originally from the Samba project). I have added a test to the
2971 spa_base64_to_bits() function which I hope fixes it.
2974 36. The daemon start-up calls getloadavg() while still root for those OS that
2975 need the first call to be done as root, but it missed one case: when
2976 deliver_queue_load_max is set with deliver_drop_privilege. This is
2977 necessary for the benefit of the queue runner, because there is no re-exec
2978 when deliver_drop_privilege is set.
2980 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2981 was in use, was not putting the data itself into the right store pool;
2982 consequently, it could be overwritten for a subsequent message in the same
2983 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2986 38. Sometimes the final signoff response after QUIT could fail to get
2987 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2988 < 0 before doing a fflush(). This bug looks as though it goes back to the
2989 introduction of TLS in release 3.20, but "sometimes" must have been rare
2990 because the tests only now provoked it.
2996 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2997 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2998 Now it returns an integer. A number of calls were still expecting a T/F
2999 return. Fortuitously, in all cases, the tests worked in OK situations,
3000 which is the norm. However, things would have gone wrong on any write
3001 failures on the smtp file descriptor. This function is used when sending
3002 messages over SMTP and also when doing verify callouts.
3004 2. When Exim is called to do synchronous delivery of a locally submitted
3005 message (the -odf or -odi options), it no longer closes stderr before doing
3008 3. Implemented the mua_wrapper option.
3010 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3012 5. Implemented the functions header_remove(), header_testname(),
3013 header_add_at_position(), and receive_remove_recipient(), and exported them
3016 6. If an ACL "warn" statement specified the addition of headers, Exim already
3017 inserted X-ACL-Warn: at the start if there was no header name. However, it
3018 was not making this test for the second and subsequent header lines if
3019 there were newlines in the string. This meant that an invalid header could
3020 be inserted if Exim was badly configured.
3022 7. Allow an ACL "warn" statement to add header lines at the start or after all
3023 the Received: headers, as well as at the end.
3025 8. Added the rcpt_4xx retry error code.
3027 9. Added postmaster_mailfrom=xxx to callout verification option.
3029 10. Added mailfrom=xxxx to the callout verification option, for verify=
3032 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3033 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3035 12. Inserted some casts to stop certain compilers warning when using pointer
3036 differences as field lengths or precisions in printf-type calls (mostly
3037 affecting debugging statements).
3039 13. Added optional readline() support for -be (dynamically loaded).
3041 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3042 same clock tick as a message's arrival, so that its received time was the
3043 same as the "first fail" time on the retry record, and that message
3044 remained on the queue past the ultimate address timeout, every queue runner
3045 would try a delivery (because it was past the ultimate address timeout) but
3046 after another failure, the ultimate address timeout, which should have then
3047 bounced the address, did not kick in. This was a "< instead of <=" error;
3048 in most cases the first failure would have been in the next clock tick
3049 after the received time, and all would be well.
3051 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3052 being recognized when the domain list was tested by the match_domain
3053 condition in an expansion string.
3055 16. Added the ${str2b64: operator.
3057 17. Exim was always calling setrlimit() to set a large limit for the number of
3058 processes, without checking whether the existing limit was already
3059 adequate. (It did check for the limit on file descriptors.) Furthermore,
3060 errors from getrlimit() and setrlimit() were being ignored. Now they are
3061 logged to the main and panic logs, but Exim does carry on, to try to do its
3062 job under whatever limits there are.
3064 18. Imported PCRE 5.0.
3066 19. Trivial typo in log message " temporarily refused connection" (the leading
3069 20. If the log selector return_path_on_delivery was set and an address was
3070 redirected to /dev/null, the delivery process crashed because it assumed
3071 that a return path would always be set for a "successful" delivery. In this
3072 case, the whole delivery is bypassed as an optimization, and therefore no
3075 21. Internal re-arrangement: the function for sending a challenge and reading
3076 a response while authentication was assuming a zero-terminated challenge
3077 string. It's now changed to take a pointer and a length, to allow for
3078 binary data in such strings.
3080 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3082 23. Exim was not respecting finduser_retries when seeking the login of the
3083 uid under which it was called; it was always trying 10 times. (The default
3084 setting of finduser_retries is zero.) Also, it was sleeping after the final
3085 failure, which is pointless.
3087 24. Implemented tls_on_connect_ports.
3089 25. Implemented acl_smtp_predata.
3091 26. If the domain in control=submission is set empty, Exim assumes that the
3092 authenticated id is a complete email address when it generates From: or
3093 Sender: header lines.
3095 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3096 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3097 chown and chgrp in /bin and hostname in /usr/bin.
3099 28. Exim was keeping the "process log" file open after each use, just as it
3100 does for the main log. This opens the possibility of it remaining open for
3101 long periods when the USR1 signal hits a daemon. Occasional processlog
3102 errors were reported, that could have been caused by this. Anyway, it seems
3103 much more sensible not to leave this file open at all, so that is what now
3106 29. The long-running daemon process does not normally write to the log once it
3107 has entered its main loop, and it closes the log before doing so. This is
3108 so that log files can straightforwardly be renamed and moved. However,
3109 there are a couple of unusual error situations where the daemon does write
3110 log entries, and I had neglected to close the log afterwards.
3112 30. The text of an SMTP error response that was received during a remote
3113 delivery was being truncated at 512 bytes. This is too short for some of
3114 the long messages that one sometimes sees. I've increased the limit to
3117 31. It is now possible to make retry rules that apply only when a message has a
3118 specific sender, in particular, an empty sender.
3120 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3121 it possible to be selective about when SMTP synchronization is enforced.
3123 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3125 32. Implemented hosts_connection_nolog.
3127 33. Added an ACL for QUIT.
3129 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3132 35. Added mailbox_size and mailbox_filecount to appendfile.
3134 36. Added control = no_multiline_responses to ACLs.
3136 37. There was a bug in the logic of the code that waits for the clock to tick
3137 in the case where the clock went backwards by a substantial amount such
3138 that the microsecond fraction of "now" was more than the microsecond
3139 fraction of "then" (but the whole seconds number was less).
3141 38. Added support for the libradius Radius client library this is found on
3142 FreeBSD (previously only the radiusclient library was supported).
3148 1. When certain lookups returned multiple values in the form name=value, the
3149 quoting of the values was not always being done properly. Specifically:
3150 (a) If the value started with a double quote, but contained no whitespace,
3152 (b) If the value contained whitespace other than a space character (i.e.
3153 tabs or newlines or carriage returns) it was not quoted.
3154 This fix has been applied to the mysql and pgsql lookups by writing a
3155 separate quoting function and calling it from the lookup code. The fix
3156 should probably also be applied to nisplus, ibase and oracle lookups, but
3157 since I cannot test any of those, I have not disturbed their existing code.
3159 2. A hit in the callout cache for a specific address caused a log line with no
3160 reason for rejecting RCPT. Now it says "Previous (cached) callout
3161 verification failure".
3163 3. There was an off-by-one bug in the queryprogram router. An over-long
3164 return line was truncated at 256 instead of 255 characters, thereby
3165 overflowing its buffer with the terminating zero. As well as fixing this, I
3166 have increased the buffer size to 1024 (and made a note to document this).
3168 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3169 when Exim is waiting for an SMTP response from a remote server, Exim
3170 restarts its select() call on the socket, thereby resetting its timeout.
3171 This is not a problem when such interrupts are rare. Somebody set up a cron
3172 job to run exiwhat every 2 minutes, which is less than the normal select()
3173 timeout (5 or 10 minutes). This meant that the select() timeout never
3174 kicked in because it was always reset. I have fixed this by comparing the
3175 time when an interrupt arrives with the time at the start of the first call
3176 to select(). If more time than the timeout has elapsed, the interrupt is
3177 treated as a timeout.
3179 5. Some internal re-factoring in preparation for the addition of Sieve
3180 extensions (by MH). In particular, the "personal" test is moved to a
3181 separate function, and given an option for scanning Cc: and Bcc: (which is
3182 not set for Exim filters).
3184 6. When Exim created an email address using the login of the caller as the
3185 local part (e.g. when creating a From: or Sender: header line), it was not
3186 quoting the local part when it contained special characters such as @.
3188 7. Installed new OpenBSD configuration files.
3190 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3191 try to make them clearer.
3193 9. Callout options, other than the timeout value, were being ignored when
3194 verifying sender addresses in header lines. For example, when using
3196 verify = header_sender/callout=no_cache
3198 the cache was (incorrectly) being used.
3200 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3201 only the Cygwin environment.
3203 11. When return_path_on_delivery was set as a log selector, if different remote
3204 addresses in the same message used different return paths and parallel
3205 remote delivery occurred, the wrong values would sometimes be logged.
3206 (Whenever a remote delivery process finished, the return path value from
3207 the most recently started remote delivery process was logged.)
3209 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3210 lines when AUTH and/or TLS are in use. This is the "received protocol"
3211 field. Exim used to use "asmtp" for authenticated SMTP, without any
3212 indication (in the protocol name) for TLS use. Now it follows the RFC and
3213 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3214 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3215 names appear in log lines as well as in Received: header lines.
3217 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3218 extensions, and comparison tests, and to fix some bugs.
3220 14. Changes to the "personal" filter test:
3222 (1) The test was buggy in that it was just doing the equivalent of
3223 "contains" tests on header lines. For example, if a user's address was
3224 anne@some.where, the "personal" test would incorrectly be true for
3226 To: susanne@some.where
3228 This test is now done by extracting each address from the header in turn,
3229 and checking the entire address. Other tests that are part of "personal"
3230 are now done using regular expressions (for example, to check local parts
3231 of addresses in From: header lines).
3233 (2) The list of non-personal local parts in From: addresses has been
3234 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3235 taken from the Sieve specification recommendations.
3237 (3) If the message contains any header line starting with "List-" it is
3238 treated as non-personal.
3240 (4) The test for "circular" in the Subject: header line has been removed
3241 because it now seems ill-conceived.
3243 15. Minor typos in src/EDITME comments corrected.
3245 16. Installed latest exipick from John Jetmore.
3247 17. If headers_add on a router specified a text string that was too long for
3248 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3249 of string_sprintf() is now avoided.
3251 18. $message_body_size was not set (it was always zero) when running the DATA
3252 ACL and the local_scan() function.
3254 19. For the "mail" command in an Exim filter, no default was being set for
3255 the once_repeat time, causing a random time value to be used if "once" was
3256 specified. (If the value happened to be <= 0, no repeat happened.) The
3257 default is now 0s, meaning "never repeat". The "vacation" command was OK
3258 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3259 (I found it when inspecting the code).
3261 20. There is now an overall timeout for performing a callout verification. It
3262 defaults to 4 times the callout timeout, which applies to individual SMTP
3263 commands during the callout. The overall timeout applies when there is more
3264 than one host that can be tried. The timeout is checked before trying the
3265 next host. This prevents very long delays if there are a large number of
3266 hosts and all are timing out (e.g. when the network connections are timing
3267 out). The value of the overall timeout can be changed by specifying an
3268 additional sub-option for "callout", called "maxwait". For example:
3270 verify = sender/callout=5s,maxwait=20s
3272 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3273 to the end before writing, but this should make it even safer).
3275 22. Exim was forgetting that it had advertised PIPELINING for the second and
3276 subsequent messages on an SMTP connection. It was also not resetting its
3277 memory on STARTTLS and an internal HELO.
3279 23. When Exim logs an SMTP synchronization error within a session, it now
3280 records whether PIPELINING has been advertised or not.
3282 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3283 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3284 rather than long int.
3286 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3288 26. Added the never_mail option to autoreply.
3294 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3295 crash if the getsockname() call failed; this can happen if a connection is
3296 closed very soon after it is established. The problem was simply in the
3297 order in which certain operations were done, causing Exim to try to write
3298 to the SMTP stream before it had set up the file descriptor. The bug has
3299 been fixed by making things happen in the correct order.
3305 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3306 before the connection was closed, thus losing the rejection response.
3308 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3309 some early Solaris releases, but causes trouble in current releases where
3310 socklen_t is defined.
3312 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3315 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3318 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3319 that releasing the top of it at the end releases what was used for sub-
3320 expansions (unless the block got too big). However, discard this block if
3321 the first thing is a variable or header, so that we can use its block when
3322 it is dynamic (useful for very large $message_headers, for example).
3324 6. Lookups now cache *every* query, not just the most recent. A new, separate
3325 store pool is used for this. It can be recovered when all lookup caches are
3326 flushed. Lookups now release memory at the end of their result strings.
3327 This has involved some general refactoring of the lookup sources.
3329 7. Some code has been added to the store_xxx() functions to reduce the amount
3330 of flapping under certain conditions.
3332 8. log_incoming_interface used to affect only the <= reception log lines. Now
3333 it causes the local interface and port to be added to several more SMTP log
3334 lines, for example "SMTP connection from", and rejection lines.
3336 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3338 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3340 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3341 could be overwritten at the end of the current message (or the start of a
3342 new message if it was set in a HELO ACL). The value is now preserved for
3343 the duration of the SMTP connection.
3345 12. If a transport had a headers_rewrite setting, and a matching header line
3346 contained an unqualified address, that address was qualified, even if it
3347 did not match any rewriting rules. The underlying bug was that the values
3348 of the flags that permit the existence of unqualified sender and recipient
3349 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3350 non-local messages, and by -bnq for local messages) were not being
3351 preserved with the message after it was received.
3353 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3354 "next input=" as part of the text comprising the host identity instead of
3355 the correct text. The code was using the same buffer for two different
3356 strings. However, depending on which order the printing function evaluated
3357 its arguments, the bug did not always show up. Under Linux, for example, my
3358 test suite worked just fine.
3360 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3361 doesn't work with some older versions of Perl. It has been changed to "my",
3362 which in any case is probably the better facility to use.
3364 15. A really picky compiler found some instances of statements for creating
3365 error messages that either had too many or two few arguments for the format
3368 16. The size of the buffer for calls to the DNS resolver has been increased
3369 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3370 for addresses that have a lot of PTR records. This alleviates a problem; it
3371 does not fully solve it.
3373 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3374 buffer now truncates the list and logs the incident, which is the same
3375 action as happens when Exim is looking up a host name and its aliases.
3376 Previously in this situation something unpredictable would happen;
3377 sometimes it was "internal error: store_reset failed".
3379 18. If a server dropped the connection unexpectedly when an Exim client was
3380 using GnuTLS and trying to read a response, the client delivery process
3381 crashed while trying to generate an error log message.
3383 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3384 string, for example:
3386 warn message = H1: something\nH2: something
3388 the text was added as a single header line from Exim's point of view
3389 though it ended up OK in the delivered message. However, searching for the
3390 second and subsequent header lines using $h_h2: did not work. This has been
3391 fixed. Similarly, if a system filter added multiple headers in this way,
3392 the routers could not see them.
3394 20. Expanded the error message when iplsearch is called with an invalid key to
3395 suggest using net-iplsearch in a host list.
3397 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3398 ACLs are no longer actually imposed (and a message to that effect is
3401 22. If a "gecos" field in a passwd entry contained escaped characters, in
3402 particular, if it contained a \" sequence, Exim got it wrong when building
3403 a From: or a Sender: header from that name. A second bug also caused
3404 incorrect handling when an unquoted " was present following a character
3405 that needed quoting.
3407 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3408 was not being matched caselessly.
3410 24. Arranged for all hyphens in the exim.8 source to be escaped with
3413 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3414 properly. Recipient callout cache records were still being keyed to include
3415 the sender, even when use_sender was set false. This led to far more
3416 callouts that were necessary. The sender is no longer included in the key
3417 when use_sender is false.
3419 26. Added "control = submission" modifier to ACLs.
3421 27. Added the ${base62d: operator to decode base 62 numbers.
3423 28. dnsdb lookups can now access SRV records.
3425 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3426 the configuration file.
3428 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3429 (-v) mode. This makes the output for a verbose queue run more intelligible.
3431 31. Added a use_postmaster feature to recipient callouts.
3433 32. Added the $body_zerocount variable, containing the number of binary zero
3434 bytes in the message body.
3436 33. The time of last modification of the "new" subdirectory is now used as the
3437 "mailbox time last read" when there is a quota error for a maildir
3440 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3442 35. Added +ignore_unknown as a special item in host lists.
3444 36. Code for decoding IPv6 addresses in host lists is now included, even if
3445 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3446 address was recognized as an IP address, but was then not correctly decoded
3447 into binary, causing unexpected and incorrect effects when compared with
3454 1. Very minor rewording of debugging text in manualroute to say "list of
3455 hosts" instead of "hostlist".
3457 2. If verify=header_syntax was set, and a header line with an unqualified
3458 address (no domain) and a large number of spaces between the end of the
3459 name and the colon was received, the reception process suffered a buffer
3460 overflow, and (when I tested it) crashed. This was caused by some obsolete
3461 code that should have been removed. The fix is to remove it!
3463 3. When running in the test harness, delay a bit after writing a bounce
3464 message to get a bit more predictability in the log output.
3466 4. Added a call to search_tidyup() just before forking a reception process. In
3467 theory, someone could use a lookup in the expansion of smtp_accept_max_
3468 per_host which, without the tidyup, could leave open a database connection.
3470 5. Added the variables $recipient_data and $sender_data which get set from a
3471 lookup success in an ACL "recipients" or "senders" condition, or a router
3472 "senders" option, similar to $domain_data and $local_part_data.
3474 6. Moved the writing of debug_print from before to after the "senders" test
3477 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3478 problems for message scanning, either using a data ACL, or using
3479 local_scan() because the Received: header was not generated till after they
3480 were called (in order to set the time as the time of reception completion).
3481 I have revised the way this works. The header is now generated after the
3482 body is received, but before the ACL or local_scan() are called. After they
3483 are run, the timestamp in the header is updated.
3489 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3490 before starting a queue runner without re-exec. This happened only when
3491 deliver_drop_privilege was set or when the Exim user was set to root. The
3492 effect of the bug was that timeouts during subsequent deliveries caused
3493 crashes instead of being properly handled. The handler is now left at its
3494 default (and expected) setting.
3496 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3497 message, again when deliver_drop_privilege is set or Exim is run as root.
3498 The bug described in (1) was not present in this case, but the tidying up
3499 of the other signals was missing. I have made the two cases consistent.
3501 3. The ignore_target_hosts setting on a manualroute router was being ignored
3502 for hosts that were looked up using the /MX notation.
3504 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3507 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3508 operated on the sender address. After changing the $sender_address to <>
3509 for the sender address verify, Exim was re-instated it as the original
3510 (before rewriting) address, but remembering that it had rewritten it, so it
3511 wasn't rewriting it again. This bug also had the effect of breaking the
3512 sender address verification caching when the sender address was rewritten.
3514 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3515 This has been changed so that if the ip literal address matches
3516 ignore_target_hosts, the router declines.
3518 7. Added expansion conditions match_domain, match_address, and match_local_
3519 part (NOT match_host).
3521 8. The placeholder for the Received: header didn't have a length field set.
3523 9. Added code to Exim itself and to exim_lock to test for a specific race
3524 condition that could lead to file corruption when using MBX delivery. The
3525 issue is with the lockfile that is created in /tmp. If this file is removed
3526 after a process has opened it but before that process has acquired a lock,
3527 there is the potential for a second process to recreate the file and also
3528 acquire a lock. This could lead to two Exim processes writing to the file
3529 at the same time. The added code performs the same test as UW imapd; it
3530 checks after acquiring the lock that its file descriptor still refers to
3531 the same named file.
3533 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3534 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3537 11. Added the smtp_active_hostname option. If used, this will typically be made
3538 to depend on the incoming interface address. Because $interface_address is
3539 not set up until the daemon has forked a reception process, error responses
3540 that can happen earlier (such as "too many connections") no longer contain
3543 12. If an expansion in a condition on a "warn" statement fails because a lookup
3544 defers, the "warn" statement is abandoned, and the next ACL statement is
3545 processed. Previously this caused the whole ACL to be aborted.
3547 13. Added the iplsearch lookup type.
3549 14. Added ident_timeout as a log selector.
3551 15. Added tls_certificate_verified as a log selector.
3553 16. Added a global option tls_require_ciphers (compare the smtp transport
3554 option of the same name). This controls incoming TLS connections.
3556 17. I finally figured out how to make tls_require_ciphers do a similar thing
3557 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3558 before starting the TLS session.
3560 18. Tabs are now shown as \t in -bP output.
3562 19. If the log selector return_path_on_delivery was set, Exim crashed when
3563 bouncing a message because it had too many Received: header lines.
3565 20. If two routers both had headers_remove settings, and the first one included
3566 a superfluous trailing colon, the final name in the first list and the
3567 first name in the second list were incorrectly joined into one item (with a
3568 colon in the middle).
3574 1. Added -C and -D options to the exinext utility, mainly to make it easier
3575 to include in the automated testing, but these could be helpful when
3576 multiple configurations are in use.
3578 2. The exinext utility was not formatting the output nicely when there was
3579 an alternate port involved in the retry record key, nor when there was a
3580 message id as well (for retries that were specific to a specific message
3581 and a specific host). It was also confused by IPv6 addresses, because of
3582 the additional colons they contain. I have fixed the IPv4 problem, and
3583 patched it up to do a reasonable job for IPv6.
3585 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3586 delivery, the log line now contains "pipelined" if PIPELINING was used.
3588 4. An SMTP transport process used to panic and die if the bind() call to set
3589 an explicit outgoing interface failed. This has been changed; it is now
3590 treated in the same way as a connect() failure.
3592 5. A reference to $sender_host_name in the part of a conditional expansion
3593 that was being skipped was still causing a DNS lookup. This no longer
3596 6. The def: expansion condition was not recognizing references to header lines
3597 that used bh_ and bheader_.
3599 7. Added the _cache feature to named lists.
3601 8. The code for checking quota_filecount in the appendfile transport was
3602 allowing one more file than it should have been.
3604 9. For compatibility with Sendmail, the command line option
3612 and sets the incoming protocol and host name (for trusted callers). The
3613 host name and its colon can be omitted when only the protocol is to be set.
3614 Note the Exim already has two private options, -pd and -ps, that refer to
3615 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3616 "s", but I don't think that's a major issue.
3618 10. A number of refactoring changes to the code, none of which should affect
3621 (a) The number of logging options was getting close to filling up the
3622 32-bit word that was used as a bit map. I have split them into two classes:
3623 those that are passed in the argument to log_write(), and those that are
3624 only ever tested independently outside of that function. These are now in
3625 separate 32-bit words, so there is plenty of room for expansion again.
3626 There is no change in the user interface or the logging behaviour.
3628 (b) When building, for example, log lines, the code previously used a
3629 macro that called string_cat() twice, in order to add two strings. This is
3630 not really sufficiently general. Furthermore, there was one instance where
3631 it was actually wrong because one of the argument was used twice, and in
3632 one call a function was used. (As it happened, calling the function twice
3633 did not affect the overall behaviour.) The macro has been replaced by a
3634 function that can join an arbitrary number of extra strings onto a growing
3637 (c) The code for expansion conditions now uses a table and a binary chop
3638 instead of a serial search (which was left over from when there were very
3639 few conditions). Also, it now recognizes conditions like "pam" even when
3640 the relevant support is not compiled in: a suitably worded error message is
3641 given if an attempt is made to use such a condition.
3643 11. Added ${time_interval:xxxxx}.
3645 12. A bug was causing one of the ddress fields not to be passed back correctly
3646 from remote delivery subprocesses. The field in question was not being
3647 subsequently used, so this caused to problems in practice.
3649 13. Added new log selectors queue_time and deliver_time.
3651 14. Might have fixed a bug in maildirsizefile handling that threw up
3652 "unexpected character" debug warnings, and recalculated the data
3653 unnecessarily. In any case, I expanded the warning message to give more
3656 15. Added the message "Restricted characters in address" to the statements in
3657 the default ACL that block characters like @ and % in local parts.
3659 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3660 Three changes have been made:
3662 (a) There was a serious bug; a negative response to MAIL caused the whole
3663 recipient domain to be cached as invalid, thereby blocking all messages
3664 to all local parts at the same domain, from all senders. This bug has
3665 been fixed. The domain is no longer cached after a negative response to
3666 MAIL if the sender used is not empty.
3668 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3671 (c) A new callout option, "use_sender" has been added for people who want
3672 the modified behaviour.
3678 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3681 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3682 indeed breaks things for older releases.
3684 3. Added additional logging to the case where there is a problem reading data
3685 from a filter that is running in a subprocess using a pipe, in order to
3686 try to track down a specific problem.
3688 4. Testing facility fudge: when running in the test harness and attempting
3689 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3690 getting "No route to host". Convert this to a timeout.
3692 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3695 6. Some OS don't have socklen_t but use size_t instead. This affects the
3696 fifth argument of getsockopt() amongst other things. This is now
3697 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3698 can be set for individual OS. I have set it for SunOS5, OSF1, and
3699 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3700 some earlier ones do not.
3702 7. Change 4.30/15 was not doing the test caselessly.
3704 8. The standard form for an IPv6 address literal was being rejected by address
3705 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3706 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3707 this, as well as the form without the "IPv6" on the front (but only when
3708 address literals are enabled, of course).
3710 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3712 10. Exim crashed if a message with an empty sender address specified by -f
3713 encountered a router with an errors_to setting. This could be provoked only
3714 by a command such as
3718 where an empty string was supplied; "<>" did not hit this bug.
3720 11. Installed PCRE release 4.5.
3722 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3723 remained set. It is now erased.
3725 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3726 times from message ids (which are base 36 rather than the normal 62).
3728 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3729 were being counted as actual protocol errors, and logged if the log
3730 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3731 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3732 and DATA following a set of rejected RCPTs do not count as protocol errors.
3733 In other words, Exim assumes they were pipelined, though this may not
3734 actually be the case. Of course, in all cases the client gets an
3735 appropriate error code.
3737 15. If a lookup fails in an ACL condition, a message about the failure may
3738 be available; it is used if testing the ACL cannot continue, because most
3739 such messages specify what the cause of the deferral is. However, some
3740 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3741 that caused an old message to be retained and used if a later statement
3742 caused a defer, replacing the real cause of the deferral.
3744 16. If an IP address had so many PTR records that the DNS lookup buffer
3745 was not large enough to hold them, Exim could crash while trying to process
3746 the truncated data. It now detects and logs this case.
3748 17. Further to 4.21/58, another change has been made: if (and only if) the
3749 first line of a message (the first header line) ends with CRLF, a bare LF
3750 in a subsequent header line has a space inserted after it, so as not to
3751 terminate the header.
3753 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3754 unnecessarily, used atoi() instead of strtol(), and didn't check the
3755 termination when getting file sizes from file names by regex.
3757 19. Completely re-implemented the support for maildirsize files, in the light
3758 of a number of problems with the previous contributed implementation
3759 (4.30/29). In particular:
3761 . If the quota is zero, the maildirsize file is maintained, but no quota is
3764 . If the maildir directory does not exist, it is created before any attempt
3765 to write a maildirsize file.
3767 . The quota value in the file is just a cache; if the quota is changed in
3768 the transport, the new value overrides.
3770 . A regular expression is available for excluding directories from the
3773 20. The autoreply transport checks the characters in options that define the
3774 message's headers; it allows continued headers, but it was checking with
3775 isspace() after an embedded newline instead of explicitly looking for a
3778 21. If all the "regular" hosts to which an address was routed had passed their
3779 expiry times, and had not reached their retry times, the address was
3780 bounced, even if fallback hosts were defined. Now Exim should go on to try
3783 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3784 equivalent code in the SMTP transport. Some hosts send humungous responses
3785 to HELO/EHLO, more than 1024 it seems.
3787 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3788 gives compiler warnings in some environments. I've now done it "properly",
3791 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3792 (because of problems with the built-in one) was declared to return uschar *
3793 instead of char *, causing compiler failure.
3795 25. Fixed a file descriptor leak when processing alias/forward files.
3797 26. Fixed a minor format string issue in dbfn.c.
3799 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3801 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3802 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3805 29. When a sender address is verified, it is cached, to save repeating the test
3806 when there is more than one recipient in a message. However, when the
3807 verification involves a callout, it is possible for different callout
3808 options to be set for different recipients. It is too complicated to keep
3809 track of this in the cache, so now Exim always runs a verification when a
3810 callout is required, relying on the callout cache for the optimization.
3811 The overhead is duplication of the address routing, but this should not be
3814 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3815 to be verified with callout=postmaster, and the main callout worked but the
3816 postmaster check failed, the verification correctly failed. However, if a
3817 subsequent RCPT command asked for sender verification *without* the
3818 postmaster check, incorrect caching caused this verification also to fail,
3821 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3822 it was not caching the DNS options (qualify_single, search_parents) that
3823 were used when the lookup failed. A subsequent lookup with different
3824 options therefore always gave the same answer, though there were cases
3825 where it should not have. (Example: a "domains = !$mx_any" option on a
3826 dnslookup router: the "domains" option is always processed without any
3827 widening, but the router might have qualify_single set.) Now Exim uses the
3828 cached value only when the same options are set.
3830 32. Added John Jetmore's "exipick" utility to the distribution.
3832 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3833 than a timeout (e.g. a certificate is required, and is not provided), an
3834 Exim server now closes the connection immediately. Previously it waited for
3835 the client to close - but if the client is SSL, it seems that they each
3836 wait for each other, leading to a delay before one of them times out.
3838 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3839 maintained 0.8.x compatibility because I don't think many are using it, and
3840 it is clearly obsolete.
3842 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3845 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3846 client certificate was expired. A simple patch fixes this, though I don't
3847 understand the full logic of why the verify callback is called multiple
3850 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3851 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3852 which causes problems with some clients (such as the Certicom SSL Plus
3853 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3854 disables the coutermeasure allowing Eudora to connect."
3856 38. Exim was not checking that a write() to a log file succeeded. This could
3857 lead to Bad Things if a log got too big, in particular if it hit a file
3858 size limit. Exim now panics and dies if it cannot write to a log file, just
3859 as it does if it cannot open a log file.
3861 39. Modified OS/Makefile-Linux so that it now contains
3863 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3865 The two -D definitions ensure that Exim is compiled with large file
3866 support, which makes it possible to handle log files that are bigger than
3869 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3870 instance) a domain was checked against a named list that involved a lookup,
3871 causing $domain_data to be set, then another domain was checked against the
3872 same list, then the first domain was re-checked, the value of $domain_data
3873 after the final check could be wrong. In particular, if the second check
3874 failed, it could be set empty. This bug probably also applied to
3877 41. The strip_trailing_dot option was not being applied to the address given
3878 with the -f command-line option.
3880 42. The code for reading a message's header from the spool was incrementing
3881 $received_count, but never initializing it. This meant that the value was
3882 incorrect (doubled) while delivering a message in the same process in which
3883 it was received. In the most common configuration of Exim, this never
3884 happens - a fresh exec is done - but it can happen when
3885 deliver_drop_privilege is set.
3887 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3888 it now includes up to 150 characters of the unexpected data in the log
3891 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3892 and building data strings. The size of both of these buffers was 10 000
3893 bytes - far larger than anybody would *ever* want, thought I. Needless to
3894 say, somebody hit the limit. I have increased the maximum line length to
3895 20 000 and the maximum data length of concatenated lines to 100 000. I have
3896 also fixed two bugs, because there was no checking on these buffers. Tsk,
3897 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3898 buffer is too small.
3900 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3901 lsearch lookups. Now it does.
3903 46. When parsing a route_list item in a manualroute router, a fixed-length
3904 buffer was used for the list of hosts. I made this 1024 bytes long,
3905 thinking that nobody would ever have a list of hosts that long. Wrong.
3906 Somebody had a whole pile of complicated expansion conditions, and the
3907 string was silently truncated, leading to an expansion error. It turns out
3908 that it is easier to change to an unlimited length (owing to other changes
3909 that have happened since this code was originally written) than to build
3910 structure for giving a limitation error. The length of the item that
3911 expands into the list of hosts is now unlimited.
3913 47. The lsearch lookup could not handle data where the length of text line was
3914 more than 4095 characters. Such lines were truncated, leading to shortened
3915 data being returned. It should now handle lines of any length.
3917 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3918 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3921 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3922 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3923 no longer happens in comment lines. A list of replacements is now placed
3924 at the head of all of the source files, except those whose only change is
3925 to replace PERL_COMMAND in the very first #! line.
3927 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3928 messages on the queue, with a bottom-up merge sort, using code contributed
3929 by Michael Haardt. This should make operations like -bp somewhat faster on
3930 large queues. It won't affect queue runners, except when queue_run_in_order
3933 51. Installed eximstats 1.31 in the distribution.
3935 52. Added support for SRV lookups to the dnslookup router.
3937 53. If an ACL referred to $message_body or $message_body_end, the value was not
3938 reset for any messages that followed in the same SMTP session.
3940 54. The store-handling optimization for building very long strings was not
3941 differentiating between the different store pools. I don't think this
3942 actually made any difference in practice, but I've tidied it.
3944 55. While running the routers to verify a sender address, $sender_address
3945 was still set to the sender address. This is wrong, because when routing to
3946 send a bounce to the sender, it would be empty. Therefore, I have changed
3947 it so that, while verifying a sender address, $sender_address is set to <>.
3948 (There is no change to what happens when verifying a recipient address.)
3950 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3951 target A or AAAA records (if not already returned) without resetting the
3952 qualify_single or search_parents options of the DNS resolver. These are
3953 inappropriate in this case because the targets of MX and SRV records must
3954 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3955 target that, when qualified, matched something in the local domain. These
3956 two options are now turned off when doing these lookups.
3958 57. It seems that at least some releases of Reiserfs (which does not have the
3959 concept of a fixed number of inodes) returns zero and not -1 for the
3960 number of available inodes. This interacted badly with check_spool_inodes,
3961 which assumed that -1 was the "no such thing" setting. What I have done is
3962 to check that the total number of inodes is greater than zero before doing
3963 the test of how many are available.
3965 58. When a "warn" ACL statement has a log_message modifier, the message is
3966 remembered, and not repeated. This is to avoid a lot of repetition when a
3967 message has many recipients that cause the same warning to be written.
3968 Howewer, Exim was preserving the list of already written lines for an
3969 entire SMTP session, which doesn't seem right. The memory is now reset if a
3970 new message is started.
3972 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3973 debugging output unless log_rewrite was also set.
3975 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3976 of (int)(handle) when we know that handle contains (void *)(-1).
3978 61. The Exim daemon panic-logs an error return when it closes the incoming
3979 connection. However "connection reset by peer" seems to be common, and
3980 isn't really an error worthy of noting specially, so that particular error
3983 62. When Exim is trying to find all the local interfaces, it used to panic and
3984 die if the ioctl to get the interface flags failed. However, it seems that
3985 on at least one OS (Solaris 9) it is possible to have an interface that is
3986 included in the list of interfaces, but for which you get a failure error
3987 for this call. This happens when the interface is not "plumbed" into a
3988 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3989 failure of the "get flags" call assumes that the interface is down.
3991 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3992 makes life easier for people who are doing arithmetic on fields extracted
3993 from dates, where you often get leading zeros that should not be
3994 interpreted as octal.
3996 64. Added qualify_domain to the redirect router, to override the global
3999 65. If a pathologically long header line contained very many addresses (the
4000 report of this problem mentioned 10 000) and each of them was rewritten,
4001 Exim could use up a very large amount of memory. (It kept on making new
4002 copies of the header line as it rewrote, and never released the old ones.)
4003 At the expense of a bit more processing, the header rewriting function has
4004 been changed so that it no longer eats memory in this way.
4006 66. The generation of the Received: header has been moved from the time that a
4007 message starts to be received, to the time that it finishes. The timestamp
4008 in the Received: header should now be very close to that of the <= log
4009 line. There are two side-effects of this change:
4011 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4012 logged header lines no longer include the local Received: line, because
4013 it has not yet been created. The same applies to a copy of the message
4014 that is returned to a non-SMTP sender when a message is rejected.
4016 (b) When a filter file is tested using -bf, no additional Received: header
4017 is added to the test message. After some thought, I decided that this
4020 This change does not affect the value of $received_for. It is still set
4021 after address rewriting, but before local_scan() is called.
4023 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4025 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4026 gave an unhelpful panic error message, and a defer error. I have managed to
4027 change this behaviour so that it now rejects any supplied certificate,
4028 which seems right, as the list of acceptable certificates is empty.
4030 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4031 gave an unhelpful defer error. I have not managed to make this reject any
4032 supplied certificates, but the error message it gives is "no certificate
4033 supplied", which is not helpful.
4035 70. exigrep's output now also includes lines that are not associated with any
4036 message, but which match the given pattern. Implemented by a patch from
4037 Martin Sluka, which also tidied up the Perl a bit.
4039 71. Recipient callout verification, like sender verification, was using <> in
4040 the MAIL FROM command. This isn't really the right thing, since the actual
4041 sender may affect whether the remote host accepts the recipient or not. I
4042 have changed it to use the actual sender in the callout; this means that
4043 the cache record is now keyed on a recipient/sender pair, not just the
4044 recipient address. There doesn't seem to be a real danger of callout loops,
4045 since a callout by the remote host to check the sender would use <>.
4046 [SEE ABOVE: changed after hitting problems.]
4048 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4049 temporary errors. However, in the case of such a code being given after
4050 the end of a data transmission (i.e. after ".") Exim was failing to write
4051 a retry record for the message. (Yes, there was some broken host that was
4052 actually sending 8xx at this point.)
4054 73. An unknown lookup type in a host list could cause Exim to panic-die when
4055 the list was checked. (An example that provoked this was putting <; in the
4056 middle of a list instead of at the start.) If this happened during a DATA
4057 ACL check, a -D file could be left lying around. This kind of configuration
4058 error no longer causes Exim to die; instead it causes a defer errror. The
4059 incident is still logged to the main and panic logs.
4061 74. Buglet left over from Exim 3 conversion. The message "too many messages
4062 in one connection" was written to the rejectlog but not the mainlog, except
4063 when address rewriting (yes!) was being logged.
4065 75. Added write_rejectlog option.
4067 76. When a system filter was run not as root (that is, when system_filter_user
4068 was set), the values of the $n variables were not being returned to the
4069 main process; thus, they were not subsequently available in the $sn
4072 77. Added +return_path_on_delivery log selector.
4074 78. A connection timeout was being treated differently from recipients deferred
4075 when testing hosts_max_try with a message that was older than the host's
4076 retry timeout. (The host should not be counted, thus allowing all hosts to
4077 be tried at least once before bouncing.) This may have been the cause of an
4078 occasionally reported bug whereby a message would remain on the queue
4079 longer than the retry timeout, but would be bounced if a delivery was
4080 forced. I say "may" because I never totally pinned down the problem;
4081 setting up timeout/retry tests is difficult. See also the next item.
4083 79. The ultimate address timeout was not being applied to errors that involved
4084 a combination of host plus message (for example, a timeout on a MAIL
4085 command). When an address resolved to a number of possible hosts, and they
4086 were not all tried for each delivery (e.g. because of hosts_max_try), a
4087 message could remain on the queue longer than the retry timeout.
4089 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4092 81. Fixed an obscure SMTP outgoing bug which required at least the following
4093 conditions: (a) there was another message waiting for the same server;
4094 (b) the server returned 5xx to all RCPT commands in the first message so
4095 that the message was not completed; (c) the server dropped the connection
4096 or gave a negative response to the RSET that Exim sends to abort the
4097 transaction. The observed case was a dropped connection after DATA that had
4098 been sent in pipelining mode. That is, the server had advertised PIPELINING
4099 but was not implementing it correctly. The effect of the bug was incorrect
4100 behaviour, such as trying another host, and this could lead to a crash.
4106 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4107 and daemon.c were passed as pointers to ints; they should have been
4108 pointers to socklen_t variables (which are typically unsigned ints).
4110 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4113 3. Fixed a really odd bug that affected only the testing scheme; patching a
4114 certain fixed string in the binary changed the value of another string that
4115 happened to be identical to the end of the original first string.
4117 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4118 name", it returns that address as the IP address. On some operating
4119 systems (e.g. Solaris), it also passes back the IP address string as the
4120 "host name". However, on others (e.g. Linux), it passes back an empty
4121 string. Exim wasn't checking for this, and was changing the host name to an
4122 empty string, assuming it had been canonicized.
4124 5. Although rare, it is permitted to have more than one PTR record for a given
4125 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4126 all the names associated with an address, because they do in Solaris.
4127 However, it seems that they do not in Linux for data that comes from the
4128 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4129 I found this out when I moved to a new Linux workstation and tried to run
4130 the Exim test suite.
4132 To get round this problem I have changed the code so that it now does its
4133 own call to the DNS to look up PTR records when searching for a host name.
4134 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4135 addresses that are only in /etc/hosts are still found.
4137 This behaviour is, however, controlled by an option called host_lookup_
4138 order, which defaults to "bydns:byaddr". If people want to use the other
4139 order, or indeed, just use one or the other means of lookup, they can
4140 specify it in this variable.
4142 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4143 some operating systems, this comes back from gethostbyaddr() as an empty
4144 string, and this is what Exim used to test for. However, it seems that in
4145 other systems, "." is yielded. Exim now tests for this case too.
4147 7. The values of check_spool_space and check_log_space are now held internally
4148 as a number of kilobytes instead of an absolute number of bytes. If a
4149 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4150 kilobyte. This means that much larger values can be stored.
4152 8. Exim monitor: an attempt to get the action menu when not actually pointing
4153 at a message produces an empty menu entitled "No message selected". This
4154 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4155 no entries in it ("Shell widget menu has zero width and/or height"). So I
4156 have added a single, blank menu entry in this case.
4158 9. Added ${quote_local_part.
4160 10. MIME decoding is now applied to the contents of Subject: header lines when
4163 11. Now that a reference to $sender_host_address automatically causes a reverse
4164 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4165 host lookup before query-style lookups in lists that might use this
4166 variable. This has therefore been abolished, and the "net-" prefix is no
4167 longer necessary for query-style lookups.
4169 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4170 have been a typo for LFLAGS, so it has been changed.
4172 13. The install script calls Exim with "-C /dev/null" in order to find the
4173 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4174 to be output. Howeve, since Exim outputs its version number before the
4175 error, it didn't break the script. It just looked ugly. I fixed this by
4176 always allowing "-C /dev/null" if the caller is root.
4178 14. Ignore overlarge ACL variable number when reading spool file - insurance
4179 against a later release with more variables having written the file.
4181 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4182 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4183 this, as well as the form without the "IPv6" on the front.
4185 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4186 OS/Makefile-Darwin file.
4188 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4189 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4191 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4192 variable. However, some OS use ints for inodes. I've added cast to long int
4193 to get rid of the compiler warning.
4195 19. I had forgotten to lock out "/../" in configuration file names when
4196 ALT_CONFIG_PREFIX was set.
4198 20. Routers used for verification do not need to specify transports. However,
4199 if such a router generated a host list, and callout was configured, Exim
4200 crashed, because it could not find a port number from the (non-existent)
4201 transport. It now assumes port 25 in this circumstance.
4203 21. Added the -t option to exigrep.
4205 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4206 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4207 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4208 them is set, LOOKUP_LSEARCH is forced.
4210 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4212 24. Added sender and host information to the "rejected by local_scan()" log
4213 line; previously there was no indication of these.
4215 25. Added .include_if_exists.
4217 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4218 Linux. It turns out that not all file systems support this. Apparently some
4219 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4220 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4221 not supported on the file descriptor, is now ignored when Exim is trying to
4222 sync a directory. This applies only to Linux.
4224 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4226 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4227 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4230 29. Added support for maildirsize files from supplied patch (modified a bit).
4232 30. The use of :fail: followed by an empty string could lead Exim to respond to
4233 sender verification failures with (e.g.):
4235 550 Verification failed for <xxx>
4236 550 Sender verify failed
4238 where the first response line was missing the '-' that indicates it is not
4239 the final line of the response.
4241 31. The loop for finding the name of the user that called Exim had a hardwired
4242 limit of 10; it now uses the value of finduser_retries, which is used for
4243 all other user lookups.
4245 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4248 33. Exim was neglecting to zero errno before one call of strtol() when
4249 expanding a string and expecting an integer value. On some systems this
4250 resulted in spurious "integer overflow" errors. Also, it was casting the
4251 result into an int without checking.
4253 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4254 did not work. The code looks as if it has *never* worked, though it appears
4255 to have been documented since at least releast 1.62. I have made it work.
4257 35. The "timeout_DNS" error in retry rules, also documented since at least
4258 1.62, also never worked. As it isn't clear exactly what this means, and
4259 clearly it isn't a major issue, I have abolished the feature by treating it
4260 as "timeout", and writing a warning to the main and panic logs.
4262 36. The display of retry rules for -brt wasn't always showing the error code
4265 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4266 timeout_connect_A, timeout_connect_MX.
4268 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4269 to the empty sender.
4271 39. The daemon was not analysing the content of -oX till after it had closed
4272 stderr and disconnected from the controlling terminal. This meant that any
4273 syntax errors were only noted on the panic log, and the return code from
4274 the command was 0. By re-arranging the code a little, I've made the
4275 decoding happen first, so such errors now appear on stderr, and the return
4276 code is 1. However, the actual setting up of the sockets still happens in
4277 the disconnected process, so errors there are still only recorded on the
4280 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4281 connections (as happens on some IP stacks) was logged at start up time as
4282 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4283 it from "IPv6 and IPv4", which means that two separate sockets are being
4286 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4287 says whether AF_INET or AF_INET6 was passed as an argument.
4289 42. Exiwhat output was messed up when time zones were included in log
4292 43. Exiwhat now gives more information about the daemon's listening ports,
4293 and whether -tls-on-connect was used.
4295 44. The "port" option of the smtp transport is now expanded.
4297 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4298 silently ignored. Now an error message is written to the main and panic
4301 46. There's a new ACL modifier called "logwrite" which writes to a log file
4302 as soon as it is encountered.
4304 47. Added $local_user_uid and $local_user_gid at routing time.
4306 48. Exim crashed when trying to verify a sender address that was being
4309 49. Exim was recognizing only a space character after ".include". It now also
4310 recognizes a tab character.
4312 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4313 extracting the relevant information from the specification. The man page no
4314 longer contains scrambled data for the -d option, and I've added a section
4315 at the front about calling Exim under different names.
4317 51. Added "extra_headers" argument to the "mail" command in filter files.
4319 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4322 53. Installed eximstats 1.29.
4324 54. Added transport_filter_timeout as a generic transport option.
4326 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4327 Cc: header lines. This was required by RFC 822, but it not required by RFC
4330 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4331 incoming messages that did not have them. Now it does so only if the
4332 message originates locally, that is, if there is no associated remote host
4333 address. When Resent- header lines are present, this applies to the Resent-
4334 lines rather than the non-Resent- lines.
4336 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4337 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4339 58. Messages for configuration errors now include the name of the main
4340 configuration file - useful now that there may be more than one file in a
4341 list (.included file names were always shown).
4343 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4344 for those rare installations that do not start the daemon as root or run it
4345 setuid root. I've cut out the call to initgroups() if the daemon is not
4348 60. The Exim user and group can now be bound into the binary as text strings
4349 that are looked up at the start of Exim's processing.
4351 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4353 62. Added $mailstore_basename variable.
4355 63. Installed patch to sieve.c from Michael Haardt.
4357 64. When Exim failed to open the panic log after failing to open the main log,
4358 the original message it was trying to log was written to stderr and debug
4359 output, but if they were not available (the usual case in production), it
4360 was lost. Now it is written to syslog before the two lines that record the
4361 failures to open the logs.
4363 65. Users' Exim filters run in subprocesses under the user's uid. It is
4364 possible for a "deliver" command or an alias in a "personal" command to
4365 provoke an address rewrite. If logging of address rewriting is configured,
4366 this fails because the process is not running as root or exim. There may be
4367 a better way of dealing with this, but for the moment (because 4.30 needs
4368 to be released), I have disabled address rewrite logging when running a
4369 filter in a non-root, non-exim process.
4375 1. The buildconfig auxiliary program wasn't quoting the value set for
4376 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4377 not defined. This bug was masked in 4.22 by the effect that was fixed in
4380 2. Some messages that were rejected after a message id was allocated were
4381 shown as "incomplete" by exigrep. It no longer does this for messages that
4382 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4384 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4385 have allow_domain_literals set, the ID did not get logged in the <= line.
4386 Domain literals are now always recognized in Message-ID: header lines.
4388 4. The first argument for a ${extract expansion item is the key name or field
4389 number. Leading and trailing spaces in this item were not being ignored,
4390 causing some misleading effects.
4392 5. When deliver_drop_privilege was set, single queue runner processes started
4393 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4394 same command in the process it spins off) were not dropping privilege.
4396 6. When the daemon running as "exim" started a queue runner, it always
4397 re-executed Exim in the spun-off process. This is a waste of effort when
4398 deliver_drop_privilege is set. The new process now just calls the
4399 queue-runner function directly.
4405 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4408 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4409 ignored. Though the use of -oP was forcing the writing of a pid file, it
4410 was always written to the default place.
4412 3. If the message "no IP address found for host xxxx" is generated during
4413 incoming verification, it is now followed by identification of the incoming
4414 connection (so you can more easily find what provoked it).
4416 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4418 5. Added some features to "harden" Exim a bit more against certain attacks:
4420 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4421 be put in Local/Makefile. This is like the never_users runtime option,
4422 but it cannot be overridden. The default setting is "root".
4424 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4425 prefix string with which any file named in a -C command line option
4428 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4429 is retained for -C and -D only if the caller of Exim is root. Without
4430 it, the exim user may also use -C and -D and retain privilege.
4432 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4433 command line option is disabled.
4435 6. Macro names set by the -D option must start with an upper case letter, just
4436 like macro names defined in the configuration file.
4438 7. Added "dereference=" facility to LDAP.
4440 8. Two instances of the typo "uknown" in the source files are fixed.
4442 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4443 the Configure-Makefile script screwed up while processing it.
4445 10. Incorporated PCRE 4.4.
4447 11. The SMTP synchronization check was not operating right at the start of an
4448 SMTP session. For example, it could not catch a HELO sent before the client
4449 waited for the greeting. There is now a check for outstanding input at the
4450 point when the greeting is written. Because of the duplex, asynchronous
4451 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4452 way, but not yet received, when the check is performed.
4454 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4455 on TCP/IP sockets, because this apparently causes some broken clients to
4458 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4459 unchanged) from the Cygwin maintainer.
4461 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4462 was supported instead of testing for mbx. Effectively a typo.
4464 15. The spa authenticator server code was not checking that the input it
4465 received was valid base64.
4467 16. The debug output line for the "set" modifier in ACLs was not showing the
4468 name of the variable that was being set.
4470 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4472 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4473 lookup on its own. Something else was needed to trigger the lookup. For
4474 example, a match in host_lookup or the need for a host name in a host list.
4475 Now, if $sender_host_name is referenced and the host name has not yet been
4476 looked up, a lookup is performed. If the lookup fails, the variable remains
4477 empty, and $host_lookup_failed is set to "1".
4479 19. Added "eqi" as a case-independent comparison operator.
4481 20. The saslauthd authentication condition could segfault if neither service
4482 nor realm was specified.
4484 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4485 error message that was logged was misleading, and incoming SMTP
4486 connections were dropped. The message is now more accurate, and temporary
4487 errors are given to SMTP connections.
4489 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4490 (except RSET) with a 421 error, until QUIT is received. However, it was
4491 failing to send a response to QUIT.
4493 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4494 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4495 ACL is now run after the helo_try_verify_hosts code.
4497 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4498 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4499 case-independent, so other case variants are also recognized). Apparently
4500 some systems use these upper case variants.
4502 25. If more than two messages were waiting for the same host, and a transport
4503 filter was specified for the transport, Exim sent two messages over the
4504 same TCP/IP connection, and then failed with "socket operation on non-
4505 socket" when it tried to send the third.
4507 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4509 27. The extern definition of crypt16() in expand.c was not being excluded when
4510 the OS had its own crypt16() function.
4512 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4513 as a preferred synonym for return_size_limit, both as an option and as an
4516 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4518 30. Changed the default configuration ACL to relax the local part checking rule
4519 for addresses that are not in any local domains. For these addresses,
4520 slashes and pipe symbols are allowed within local parts, but the sequence
4521 /../ is explicitly forbidden.
4523 31. SPA server authentication was not clearing the challenge buffer before
4526 32. log_message in a "warn" ACL statement was writing to the reject log as
4527 well as to the main log, which contradicts the documentation and doesn't
4528 seem right (because no rejection is happening). So I have stopped it.
4530 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4531 However, I am unable to do any testing of this.
4533 34. Fixed an infelicity in the appendfile transport. When checking directories
4534 for a mailbox, to see if any needed to be created, it was accidentally
4535 using path names with one or more superfluous leading slashes; tracing
4536 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4538 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4539 contents are added to the log line that is written for every discarded
4540 recipient. (Previously a log_message setting was ignored.)
4542 36. The ${quote: operator now quotes the string if it is empty.
4544 37. The install script runs exim in order to find its version number. If for
4545 some reason other than non-existence or emptiness, which it checks, it
4546 could not run './exim', it was installing it with an empty version number,
4547 i.e. as "exim-". This error state is now caught, and the installation is
4550 38. An argument was missing from the function that creates an error message
4551 when Exim fails to connect to the socket for saslauthd authentication.
4552 This could cause Exim to crash, or give a corrupted message.
4554 39. Added isip, isip4, and isip6 to ${if conditions.
4556 40. The ACL variables $acl_xx are now saved with the message, and can be
4557 accessed later in routers, transports, and filters.
4559 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4560 strings in the file are not string-expanded.
4562 42. If a MAIL command specified a SIZE value that was too large to fit into an
4563 int variable, the check against message_size_limit failed. Such values are
4564 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4565 day this will have to be increased, but I don't think I want to be around
4566 when emails are that large.
4573 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4574 iconv() is not standard in FreeBSD.
4576 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4577 IPv6 enabled. The observed symptom was a segmentation fault on return from
4578 the function os_common_find_running_interfaces() in src/os.c.
4580 3. In the check_special_case() function in daemon.c I had used "errno" as an
4581 argument name, which causes warnings on some systems. This was basically a
4582 typo, since it was named "eno" in the comments!
4584 4. The code that waits for the clock to tick (at a resolution of some fraction
4585 of a second) so as to ensure message-id uniqueness was always waiting for
4586 at least one whole tick, when it could have waited for less. [This is
4587 almost certainly not relevant at current processor speeds, where it is
4588 unlikely to ever wait at all. But we try to future-proof.]
4590 5. The function that sleeps for a time interval that includes fractions of a
4591 second contained a race. It did not block SIGALRM between setting the
4592 timer, and suspending (a couple of lines later). If the interval was short
4593 and the sigsuspend() was delayed until after it had expired, the suspension
4594 never ended. On busy systems this could lead to processes getting stuck for
4597 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4598 process, before it forks any delivery processes. The open lookup caching
4599 mechanism meant that the open file or database connection was passed into
4600 the delivery process. The problem was that delivery processes always tidy
4601 up cached lookup data. This could cause a problem for the next delivery
4602 process started by the queue runner, because the external queue runner
4603 process does not know about the closure. So the next delivery process
4604 still has data in the lookup cache. In the case of a file lookup, there was
4605 no problem because closing a file descriptor in a subprocess doesn't affect
4606 the parent. However, if the lookup was caching a connection to a database,
4607 the connection was closed, and the second delivery process was likely to
4608 see errors such as "PGSQL: query failed: server closed the connection
4609 unexpectedly". The problem has been fixed by closing all cached lookups
4610 in a queue runner before running a delivery process.
4612 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4613 seem to have the "const" qualifier which it has on other OS. I've
4616 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4617 *of the same type* (client or server) with the same public name that an
4618 error should be diagnosed.
4620 9. When Exim looked up a host name for an IP address, but failed to find the
4621 original IP address when looking up the host name (a safety check), it
4622 output the message "<ip address> does not match any IP for NULL", which was
4623 confusing, to say the least. The bug was that the host name should have
4624 appeared instead of "NULL".
4626 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4627 user that is built into the binary, and the -C or -D options is used, root
4628 privilege is dropped before the configuration file is read. In addition,
4629 logging is switched to stderr instead of the normal log files. If the
4630 configuration then re-defines the Exim user, the unprivileged environment
4631 is probably not what is expected, so Exim logs a panic warning message (but
4634 However, if deliver_drop_privilege is set, the unprivileged state may well
4635 be exactly what is intended, so the warning has been cut out in that case,
4636 and Exim is allowed to try to write to its normal log files.
4642 1. smtp_return_error_details was not giving details for temporary sender
4643 or receiver verification errors.
4645 2. Diagnose a configuration error if two authenticators have the same public
4648 3. Exim used not to create the message log file for a message until the first
4649 delivery attempt. This could be confusing when incoming messages were held
4650 for policy or load reasons. The message log file is now created at the time
4651 the message is received, and an initial "Received" line is written to it.
4653 4. The automatically generated man page for command line options had a minor
4654 bug that caused no ill effects; however, a more serious problem was that
4655 the procedure for building the man page automatically didn't always
4656 operate. Consequently, release 4.20 contains an out-of-date version. This
4657 shouldn't happen again.
4659 5. When building Exim with embedded Perl support, the script that builds the
4660 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4661 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4663 6. The freeze_tell option was not being used for messages that were frozen on
4664 arrival, either by an ACL or by local_scan().
4666 7. Added the smtp_incomplete_transaction log selector.
4668 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4669 was accepting AUTH without a new EHLO.
4671 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4672 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4675 10. Logging of TCP/IP connections (when configured) now happens in the main
4676 daemon process instead of the child process, so that the TCP/IP connection
4677 count is more accurate (but it can never be perfect).
4679 11. The use of "drop" in a nested ACL was not being handled correctly in the
4680 outer ACL. Now, if condition failure induced by the nested "drop" causes
4681 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4682 or "require"), the connection is dropped.
4684 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4685 that yield "discard" can now be used with an "accept" or a "discard" verb,
4686 but an error is generated for any others (because I can't see a useful way
4687 to define what should happen).
4689 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4690 are now processed in the same way as lines in the Exim configuration file.
4691 In particular, continuation lines are supported.
4693 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4695 15. Added -ti meaning -t -i.
4697 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4698 domains, and warn by logging if others are found.
4700 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4701 returned from GIFCONF when Exim is trying to find the list of interfaces on
4702 a host. The code in os.c has been modified to copy each ifreq to an aligned
4703 structure in all cases.
4705 Also, in some cases, the returned ifreq's were being copied to a 'struct
4706 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4707 means the last couple of bytes of an IPv6 address could be chopped if the
4708 ifreq contained only a normal sockaddr (14 bytes storage).
4710 18. Named domain lists were not supported in the hosts_treat_as_local option.
4711 An entry such as +xxxx was not recognized, and was treated as a literal
4714 19. Ensure that header lines added by a DATA ACL are included in the reject log
4715 if the ACL subsequently rejects the message.
4717 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4718 MD5 (which is deprecated).
4720 21. When testing a filter file using -bf, Exim was writing a message when it
4721 took the sender from a "From " line in the message, but it was not doing so
4722 when it took $return_path from a Return-Path: header line. It now does.
4724 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4725 with a valid header line field name (a series of printing characters
4726 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4728 23. Changed "disc" in the source to "disk" to conform to the documentation and
4729 the book and for uniformity.
4731 24. Ignore Sendmail's -Ooption=value command line item.
4733 25. When execve() failed while trying to run a command in a pipe transport,
4734 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4735 could be confused with a return value of 69 from the command itself. This
4736 has been changed to 127, the value the shell returns if it is asked to run
4737 a non-existent command. The wording for the related log line suggests a
4738 non-existent command as the problem.
4740 26. If received_header_text expands to an empty string, do not add a Received:
4741 header line to the message. (Well, it adds a token one on the spool, but
4742 marks it "old" so that it doesn't get used or transmitted.)
4744 27. Installed eximstats 1.28 (addition of -nt option).
4746 28. There was no check for failure on the call to getsockname() in the daemon
4747 code. This can fail if there is a shortage of resources on the system, with
4748 ENOMEM, for example. A temporary error is now given on failure.
4750 29. Contrary to the C standard, it seems that in some environments, the
4751 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4752 program. Exim now does this explicitly; it affects the formatting of
4753 timestamps using strftime().
4755 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4756 complaints. I've now initialized all the variables, to avoid this.
4758 32. Header lines added by a system filter were not being "seen" during
4759 transport-time rewrites.
4761 33. The info_callback() function passed to OpenSSL is set up with type void
4762 (*)(SSL *, int, int), as described somewhere. However, when calling the
4763 function (actually a macro) that sets it up, the type void(*)() is
4764 expected. I've put in a cast to prevent warnings from picky compilers.
4766 34. If a DNS black list lookup found a CNAME record, but there were no A
4767 records associated with the domain it pointed at, Exim crashed.
4769 35. If a DNS black list lookup returned more than one A record, Exim ignored
4770 all but the first. It now scans all returned addresses if a particular IP
4771 value is being sought. In this situation, the contents of the
4772 $dnslist_value variable are a list of all the addresses, separated by a
4775 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4776 to accept a host name and all its aliases if the forward lookup for any of
4777 them yielded the IP address of the incoming connection. Now it accepts only
4778 those names whose forward lookup yields the correct IP address. Any other
4779 names are discarded. This closes a loophole whereby a rogue DNS
4780 administrator could create reverse DNS records to break through a
4781 wildcarded host restriction in an ACL.
4783 37. If a user filter or a system filter that ran in a subprocess used any of
4784 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4785 the wrong values were passed to the pipe command ($thisaddress had the
4786 value of $0, $0 had the value of $1, etc). This bug was introduced by
4787 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4789 38. Improved the line breaking for long SMTP error messages from ACLs.
4790 Previously, if there was no break point between 40 and 75 characters, Exim
4791 left the rest of the message alone. Two changes have been made: (a) I've
4792 reduced the minimum length to 35 characters; (b) if it can't find a break
4793 point between 35 and 75 characters, it looks ahead and uses the first one
4794 that it finds. This may give the occasional overlong line, but at least the
4795 remaining text gets split now.
4797 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4798 file descriptors might be low, and that setting 1000 would always raise it.
4799 It turns out that in some environments, the limit is already over 1000 and
4800 that lowering it causes trouble. So now Exim takes care not to decrease it.
4802 40. When delivering a message, the value of $return_path is set to $sender_
4803 address at the start of routing (routers may change the value). By an
4804 oversight, this default was not being set up when an address was tested by
4805 -bt or -bv, which affected the outcome if any router or filter referred to
4808 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4809 log only when "message" or "log_message" are set. However, if one of the
4810 conditions was an address verification, or a call to a nested ACL, the
4811 messages generated by the underlying test were being passed through. This
4812 no longer happens. The underlying message is available in $acl_verify_
4813 message for both "message" and "log_message" expansions, so it can be
4814 passed through if needed.
4816 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4817 new expansion $bh_ to give the encoded byte string without charset
4818 translation. Translation happens only if iconv() is available; HAVE_ICONV
4819 indicates this at build time. HEADERS_CHARSET gives the charset to
4820 translate to; headers_charset can change it in the configuration, and
4821 "headers charset" can change it in an individual filter file.
4823 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4824 that creates RFC 2047 encoded "words" labels them as that charset instead
4825 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4826 expansion operator; (ii) when Exim creates a From: line for a local
4827 message; (iii) when a header line is rewritten to include a "phrase" part.
4829 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4830 buggy, causing it to skip the first lines of messages whose message ID
4831 ended in 'D'. This would not have bitten before Exim release 4.14, because
4832 message IDs were unlikely to end in 'D' before then. The effect was to have
4833 incorrect size information for certain domains.
4835 45. #include "config.h" was missing at the start of the crypt16.c module. This
4836 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4839 46. If there was a timeout during a "random" callout check, Exim treated it as
4840 a failure of the random address, and carried on sending RSET and the real
4841 address. If the delay was just some slowness somewhere, the response to the
4842 original RCPT would be taken as a response to RSET and so on, causing
4843 mayhem of various kinds.
4845 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4846 when I implemented it. It didn't allow for the fact that some option values
4847 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4848 the right test for positive values.
4850 48. Domain names in DNS records are case-independent. Exim always looks them up
4851 in lower case. Some resolvers return domain names in exactly the case they
4852 appear in the zone file, that is, they may contain uppercase letters. Not
4853 all resolvers do this - some return always lower case. Exim was treating a
4854 change of case by a resolver as a change of domain, similar to a widening
4855 of a domain abbreviation. This triggered its re-routing code and so it was
4856 trying to route what was effectively the same domain again. This normally
4857 caused routing to fail (because the router wouldn't handle the domain
4858 twice). Now Exim checks for this case specially, and just changes the
4859 casing of the domain that it ultimately uses when it transmits the message
4862 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4865 50. If a filter generated a file delivery with a non-absolute name (possible if
4866 no home directory exists for the router), the forbid_file option was not
4869 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4870 the existing equality matching.
4872 52. Exim was using ints instead of ino_t variables in some places where it was
4873 dealing with inode numbers.
4875 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4876 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4877 called TMPDIR, and if it finds it is different, it changes its value.
4879 54. The smtp_printf() function is now made available to local_scan() so
4880 additional output lines can be written before returning. There is also an
4881 smtp_fflush() function to enable the detection of a dropped connection.
4882 The variables smtp_input and smtp_batched_input are exported to
4885 55. Changed the default runtime configuration: the message "Unknown user"
4886 has been removed from the ACL, and instead placed on the localuser router,
4887 using the cannot_route_message feature. This means that any verification
4888 failures that generate their own messages won't get overridden. Similarly,
4889 the "Unrouteable address" message that was in the ACL for unverifiable
4890 relay addresses has also been removed.
4892 56. Added hosts_avoid_esmtp to the smtp transport.
4894 57. The exicyclog script was not checking for the esoteric option
4895 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4896 will work only if exicyclog is run under the appropriate euid.
4898 58. Following a discussion on the list, the rules by which Exim recognises line
4899 endings on incoming messages have been changed. The -dropcr and drop_cr
4900 options are now no-ops, retained only for backwards compatibility. The
4901 following line terminators are recognized: LF CRLF CR. However, special
4902 processing applies to CR:
4904 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4905 nor a local message in the state where . is a terminator.
4907 (ii) If a bare CR is encountered in a header line, an extra space is added
4908 after the line terminator so as not to end the header. The reasoning
4909 behind this is that bare CRs in header lines are most likely either
4910 to be mistakes, or people trying to play silly games.
4912 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4913 was being incorrectly given as 18 bytes larger than it should have been.
4914 This is a VOB (very old bug).
4916 60. This may never have affected anything current, but just in case it has:
4917 When the local host is found other than at the start of a list of hosts,
4918 the local host, those with the same MX, and any that follow, are discarded.
4919 When the list in question was part of a longer list of hosts, the following
4920 hosts (not currently being processed) were also being discarded. This no
4921 longer happens. I'm not sure if this situation could ever has previously
4924 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4927 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4928 header. This is something that is recommended in a new Internet Draft, and
4929 is something that is documented as being done by Sendmail. There are two
4930 possible values. For messages generated by the autoreply transport, Exim
4933 Auto-Submitted: auto-replied
4935 whereas for all other generated messages (e.g. bounces) it adds
4937 Auto-Submitted: auto-generated
4939 63. The "personal" condition in filters now includes a test for the
4940 Auto-Submitted: header. If it contains the string "auto-" the message it
4941 not considered personal.
4943 64. Added rcpt_include_affixes as a generic transport option.
4945 65. Added queue_only_override (default true).
4947 66. Added the syslog_duplication option.
4949 67. If what should have been the first header line of a message consisted of
4950 a space followed by a colon, Exim was mis-interpreting it as a header line.
4951 It isn't of course - it is syntactically invalid and should therefore be
4952 treated as the start of the message body. The misbehaviour could have
4953 caused a number of strange effects, including loss of data in subsequent
4954 header lines, and spool format errors.
4956 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4957 client host had authenticated. This control can now be exercised by an ACL
4958 for more flexibility.
4960 69. By default, callouts do not happen when testing with -bh. There is now a
4961 variant, -bhc, which does actually run the callout code, including
4962 consulting and updating the callout cache.
4964 70. Added support for saslauthd authentication, courtesy of Alexander
4967 71. If statvfs() failed on the spool or log directories while checking their
4968 size for availability, Exim confusingly gave the error "space shortage".
4969 Furthermore, in debugging mode it crashed with a floating point exception.
4970 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4971 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4972 serious problem, Exim now writes to the main and panic logs when this
4973 happens, with details of the failure. It then refuses to accept the
4974 incoming message, giving the message "spool directory problem" or "log
4975 directory problem" with a 421 code for SMTP messages.
4977 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4978 0, 1, and 2 exist, because some OS complain for execs without them (see
4979 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4980 descriptors. However, the code omitted to check that the open succeeded,
4981 causing mysterious errors if for some reason the permissions on /dev/null
4982 got screwed. Now Exim writes a message to the main and panic logs, and
4983 bombs out if it can't open /dev/null.
4985 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4986 interact so that it is all more flexible. It is supposed to remain
4987 backwards compatible. Also added extra_local_interfaces.
4989 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4990 to bomb out with an assertion failure - to the client this appears as a
4991 connection drop. This problem occurs in the part of the code that was taken
4992 from the Samba project. Fortunately, the assertion is in a very simple
4993 function, so I have fixed this by reproducing the function inline in the
4994 one place where it is called, and arranging for authentication to fail
4995 instead of killing the process with assert().
4997 75. The SPA client code was not working when the server requested OEM rather
4998 than Unicode encoding.
5000 76. Added code to make require_files with a specific uid setting more usable in
5001 the case where statting the file as root fails - usually a non-root-mounted
5002 NFS file system. When this happens and the failure is EACCES, Exim now
5003 forks a subprocess and does the per-uid checking as the relevant uid.
5005 77. Added process_log_path.
5007 78. If log_file_path was not explicitly set, a setting of check_log_space or
5008 check_log_inodes was ignored.
5010 79. If a space check for the spool or log partitions fails, the incident is now
5011 logged. Of course, in the latter case the data may get lost...
5013 80. Added the %p formatting code to string_format() so that it can be used to
5014 print addresses in debug_print(). Adjusted all the address printing in the
5015 debugging in store.c to use %p rather than %d.
5017 81. There was a concern that a line of code in smtp_in.c could overflow a
5018 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5019 initially expressed, the concern was not well-founded, because trailing
5020 spaces are removed early. However, if the trailing spaces were followed by
5021 a NULL, they did not get removed, so the overflow was possible. Two fixes
5024 (a) I re-wrote the offending code in a cleaner fashion.
5025 (b) If an incoming SMTP command contains a NULL character, it is rejected
5028 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5029 runs initgroups(), so that if the Exim user is in any additional groups,
5030 they will be used during message reception.
5036 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob