1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.600 2010/01/04 18:16:54 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 JJ/01 installed exipick 20100104.1, adding $max_received_linelength, $data_path, and $header_path variables; fixed documentation bugs and typos
14 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
16 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
18 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
20 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
22 NM/04 Bugzilla 915: Flag broken perl installation during build.
28 TK/01 Added patch by Johannes Berg that expands the main option
29 "spamd_address" if it starts with a dollar sign.
31 TK/02 Write list of recipients to X-Envelope-Sender header when building
32 the mbox-format spool file for content scanning (suggested by Jakob
35 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
36 (http://www.dcc-servers.net/) support via dccifd. Activated by
37 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
39 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
40 by Mark Daniel Reidel <mr@df.eu>.
42 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
43 When building exim an external PCRE library is now needed -
44 PCRE is a system library on the majority of modern systems.
45 See entry on PCRE_LIBS in EDITME file.
47 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
48 conversation. Added nologin parameter to request.
49 Patch contributed by Kirill Miazine
51 TF/01 Do not log submission mode rewrites if they do not change the address.
53 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
55 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
56 log files in place. Contributed by Roberto Lima
58 NM/04 Bugzilla 667: close socket used by dovecot authenticator
60 TF/03 Bugzilla 615: When checking the local_parts router precondition
61 after a local_part_suffix or local_part_prefix option, Exim now
62 does not use the address's named list lookup cache, since this
63 contains cached lookups for the whole local part.
65 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
66 Robert Millan. Documentation is in experimental-spec.txt
68 TF/04 Bugzilla 668: Fix parallel build (make -j).
70 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
72 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
73 Patch provided by Jan Srzednicki
75 TF/05 Leading white space used to be stripped from $spam_report which
76 wrecked the formatting. Now it is preserved.
78 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
79 that they are available at delivery time.
81 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
83 TF/08 TLS error reporting now respects the incoming_interface and
84 incoming_port log selectors.
86 TF/09 Produce a more useful error message if an SMTP transport's hosts
87 setting expands to an empty string.
89 NM/06 Bugzilla 744: EXPN did not work under TLS.
90 Patch provided by Phil Pennock
92 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
93 Patch provided by Richard Godbee
95 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
96 acl_smtp_notquit, added index entry.
98 NM/09 Bugzilla 787: Potential buffer overflow in string_format
99 Patch provided by Eugene Bujak
101 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
102 Patch provided by Maxim Dounin
104 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
105 Patch provided by Phil Pennock
107 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
109 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
110 Patch provided by Brad "anomie" Jorsch
112 NM/14 Bugzilla 622: Added timeout setting to transport filter.
113 Patch provided by Dean Brooks
115 TK/05 Add native DKIM support (does not depend on external libraries).
117 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
118 Patch provided by Graeme Fowler
120 NM/16 Bugzilla 851: Documentation example syntax fix.
122 NM/17 Changed NOTICE file to remove references to embedded PCRE.
124 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
126 NM/19 Bugzilla 745: TLS version reporting
127 Patch provided by Phil Pennock
129 NM/20 Bugzilla 167: bool: condition support
130 Patch provided by Phil Pennock
132 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
133 Patch provided by Phil Pennock
135 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
136 Patch provided by Brad "anomie" Jorsch
138 NM/23 Bugzilla 687: Fix misparses in eximstats
139 Patch provided by Heiko Schlittermann
141 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
142 Patch provided by Heiko Schlittermann
144 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
145 plus update to original patch.
147 NM/26 Bugzilla 799: Documentation correction for ratelimit
149 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
150 Patch provided by David Brownlee
152 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
154 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
156 NM/30 Bugzilla 888: TLS documentation bugfixes
158 NM/31 Bugzilla 896: Dovecot buffer overrun fix
160 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
161 Unlike the original bugzilla I have changed all shell scripts in src tree
163 NM/33 Bugzilla 898: Transport filter timeout fix
164 Patch by Todd Rinaldo
166 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
167 Patch by Serge Demonchaux
169 NM/35 Bugzilla 39: Base64 decode bug fixes
170 Patch by Jakob Hirsch
172 NM/36 Bugzilla 909: Correct connect() call in dcc code
174 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
176 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
178 NM/39 Bugzilla 911: Fixed MakeLinks build script
184 TK/01 Add preliminary DKIM support. Currently requires a forked version of
185 ALT-N's libdkim that I have put here:
186 http://duncanthrax.net/exim-experimental/
188 Note to Michael Haardt: I had to rename some vars in sieve.c. They
189 were called 'true' and it seems that C99 defines that as a reserved
190 keyword to be used with 'bool' variable types. That means you could
191 not include C99-style headers which use bools without triggering
192 build errors in sieve.c.
194 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
195 as mailq or other aliases. Changed the --help handling significantly
196 to do whats expected. exim_usage() emits usage/help information.
198 SC/01 Added the -bylocaldomain option to eximstats.
200 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
202 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
204 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
210 PH/01 Another patch from the Sieve maintainer.
212 PH/02 When an IPv6 address is converted to a string for single-key lookup
213 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
214 dots are used instead of colons so that keys in lsearch files need not
215 contain colons. This was done some time before quoting was made available
216 in lsearch files. However, iplsearch files do require colons in IPv6 keys
217 (notated using the quote facility) so as to distinguish them from IPv4
218 keys. This meant that lookups for IP addresses in host lists did not work
219 for iplsearch lookups.
221 This has been fixed by arranging for IPv6 addresses to be expressed with
222 colons if the lookup type is iplsearch. This is not incompatible, because
223 previously such lookups could never work.
225 The situation is now rather anomolous, since one *can* have colons in
226 ordinary lsearch keys. However, making the change in all cases is
227 incompatible and would probably break a number of configurations.
229 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
232 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
233 conversion specification without a maximum field width, thereby enabling
234 a rogue spamd server to cause a buffer overflow. While nobody in their
235 right mind would setup Exim to query an untrusted spamd server, an
236 attacker that gains access to a server running spamd could potentially
237 exploit this vulnerability to run arbitrary code as the Exim user.
239 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
240 $primary_hostname instead of what libspf2 thinks the hosts name is.
242 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
243 a directory entry by the name of the lookup key. Previously, if a
244 symlink pointed to a non-existing file or a file in a directory that
245 Exim lacked permissions to read, a lookup for a key matching that
246 symlink would fail. Now it is enough that a matching directory entry
247 exists, symlink or not. (Bugzilla 503.)
249 PH/03 The body_linecount and body_zerocount variables are now exported in the
252 PH/04 Added the $dnslist_matched variable.
254 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
255 This means they are set thereafter only if the connection becomes
258 PH/06 Added the client_condition to authenticators so that some can be skipped
259 by clients under certain conditions.
261 PH/07 The error message for a badly-placed control=no_multiline_responses left
262 "_responses" off the end of the name.
264 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
266 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
267 (without spaces) instead of just copying the configuration text.
269 PH/10 Added the /noupdate option to the ratelimit ACL condition.
271 PH/11 Added $max_received_linelength.
273 PH/12 Added +ignore_defer and +include_defer to host lists.
275 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
276 way in which PCRE > 7.0 is built.
278 PH/14 Implemented queue_only_load_latch.
280 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
281 MAIL command. The effect was to mangle the value on 64-bit systems.
283 PH/16 Another patch from the Sieve maintainer.
285 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
287 PH/18 If a system quota error occurred while trying to create the file for
288 a maildir delivery, the message "Mailbox is full" was not appended to the
289 bounce if the delivery eventually timed out. Change 4.67/27 below applied
290 only to a quota excession during the actual writing of the file.
292 PH/19 It seems that peer DN values may contain newlines (and other non-printing
293 characters?) which causes problems in log lines. The DN values are now
294 passed through string_printing() before being added to log lines.
296 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
297 and InterBase are left for another time.)
299 PH/21 Added message_body_newlines option.
301 PH/22 Guard against possible overflow in moan_check_errorcopy().
303 PH/23 POSIX allows open() to be a macro; guard against that.
305 PH/24 If the recipient of an error message contained an @ in the local part
306 (suitably quoted, of course), incorrect values were put in $domain and
307 $local_part during the evaluation of errors_copy.
313 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
314 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
317 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
318 issue a MAIL command.
320 PH/02 In an ACL statement such as
322 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
324 if a client was not listed at all, or was listed with a value other than
325 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
326 the condition was not true (as it should be), so access was not denied.
327 The bug was that the ! inversion was incorrectly passed on to the second
328 item. This has been fixed.
330 PH/03 Added additional dnslists conditions == and =& which are different from
331 = and & when the dns lookup returns more than one IP address.
333 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
334 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
336 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
337 FSYNC, which compiles an option called disable_fsync that allows for
338 bypassing fsync(). The documentation is heavily laced with warnings.
340 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
342 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
343 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
344 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
345 including adding "make clean"; (3) Added -fPIC when compiling the test
346 dynamically loaded module, to get rid of a warning.
348 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
349 message fails, move_frozen_messages = true and ignore_bounce_errors_after
350 = 0s. The bug is otherwise harmless.
352 PH/07 There was a bug in the dovecot authenticator such that the value of
353 $auth1 could be overwritten, and so not correctly preserved, after a
354 successful authentication. This usually meant that the value preserved by
355 the server_setid option was incorrect.
357 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
359 PH/09 Installed PCRE release 7.0.
361 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
362 run for batched SMTP input. It is now run at the start of every message
363 in the batch. While fixing this I discovered that the process information
364 (output by running exiwhat) was not always getting set for -bs and -bS
365 input. This is fixed, and it now also says "batched" for BSMTP.
367 PH/11 Added control=no_pipelining.
369 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
370 patch, slightly modified), and move the expansion of helo_data till after
371 the connection is made in the smtp transport (so it can use these
374 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
376 PH/14 Added log_selector = +pid.
378 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
380 PH/16 Add ${if forany and ${if forall.
382 PH/17 Added dsn_from option to vary the From: line in DSNs.
384 PH/18 Flush SMTP output before performing a callout, unless control =
385 no_callout_flush is set.
387 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
388 was true (the default) a successful delivery failed to delete the retry
389 item, thus causing premature timeout of the address. The bug is now
392 PH/20 Added hosts_avoid_pipelining to the smtp transport.
394 PH/21 Long custom messages for fakedefer and fakereject are now split up
395 into multiline reponses in the same way that messages for "deny" and
396 other ACL rejections are.
398 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
399 with slight modification.
401 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
402 draft, changing the syntax and factoring some duplicate code".
404 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
405 for deliveries of the second and subsequent messages over the same SMTP
408 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
409 ${reduce, with only minor "tidies".
411 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
413 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
414 expansion side effects.
416 PH/27 When a message times out after an over-quota error from an Exim-imposed
417 quota, the bounce message says "mailbox is full". This message was not
418 being given when it was a system quota that was exceeded. It now should
421 MH/03 Made $recipients available in local_scan(). local_scan() already has
422 better access to the recipient list through recipients_list[], but
423 $recipients can be useful in postmaster-provided expansion strings.
425 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
426 in the case of a MAIL command with additional options following the
427 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
428 were accidentally chopped off.
430 PH/29 SMTP synchronization checks are implemented when a command is read -
431 there is a check that no more input is waiting when there shouldn't be
432 any. However, for some commands, a delay in an ACL can mean that it is
433 some time before the response is written. In this time, more input might
434 arrive, invalidly. So now there are extra checks after an ACL has run for
435 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
436 pipelining has not been advertised.
438 PH/30 MH's patch to allow iscntrl() characters to be list separators.
440 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
441 returned in the SMTP response when smtp_return_error_details was false.
444 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
445 instead of the C I/O that was originally supplied, because problems were
448 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
449 Exim which did not show up earlier: it was assuming that a call to
450 SSL_CTX_set_info_callback() might give an error value. In fact, there is
451 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
452 was a macro that became an assignment, so it seemed to work. This has
453 changed to a proper function call with a void return, hence the compile
454 error. Exim's code has been fixed.
456 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
459 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
461 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
463 PH/37 If a message is not accepted after it has had an id assigned (e.g.
464 because it turns out to be too big or there is a timeout) there is no
465 "Completed" line in the log. When some messages of this type were
466 selected by exigrep, they were listed as "not completed". Others were
467 picked up by some special patterns. I have improved the selection
468 criteria to be more general.
470 PH/38 The host_find_failed option in the manualroute router can now be set
471 to "ignore", to completely ignore a host whose IP address cannot be
472 found. If all hosts are ignored, the behaviour is controlled by the new
473 host_all_ignored option.
475 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
476 homing or because of multiple MX records with /mx) generated more than
477 one IP address, and the following item turned out to be the local host,
478 all the secondary addresses of the first item were incorrectly removed
479 from the list, along with the local host and any following hosts (which
480 is what is supposed to happen).
482 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
483 whoever called Exim into the -H file. In the case of the daemon it was
484 behaving confusingly. When first started, it used values for whoever
485 started the daemon, but after a SIGHUP it used the Exim user (because it
486 calls itself on a restart). I have changed the code so that it now always
489 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
490 message are rejected with the same error (e.g. no authentication or bad
491 sender address), and a DATA command is nevertheless sent (as can happen
492 with PIPELINING or a stupid MUA), the error message that was given to the
493 RCPT commands is included in the rejection of the DATA command. This is
494 intended to be helpful for MUAs that show only the final error to their
497 PH/42 Another patch from the Sieve maintainer.
499 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
500 Eximstats - Fixed some broken HTML links and added missing column headers
502 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
503 columns for Rejects, Temp Rejects, Ham, and Spam rows.
505 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
507 PH/43 Yet another patch from the Sieve maintainer.
509 PH/44 I found a way to check for a TCP/IP connection going away before sending
510 the response to the final '.' that terminates a message, but only in the
511 case where the client has not sent further data following the '.'
512 (unfortunately, this is allowed). However, in many cases there won't be
513 any further data because there won't be any more messages to send. A call
514 to select() can be used: if it shows that the input is "ready", there is
515 either input waiting, or the socket has been closed. An attempt to read
516 the next input character can distinguish the two cases. Previously, Exim
517 would have sent an OK response which the client would never have see.
518 This could lead to message repetition. This fix should cure that, at
519 least in a lot of common cases.
521 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
522 advertised in response to EHLO.
528 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
529 fixed by 4.65/MH/01 (is this a record?) are fixed:
531 (i) An empty string was always treated as zero by the numeric comparison
532 operators. This behaviour has been restored.
534 (ii) It is documented that the numeric comparison operators always treat
535 their arguments as decimal numbers. This was broken in that numbers
536 starting with 0 were being interpreted as octal.
538 While fixing these problems I realized that there was another issue that
539 hadn't been noticed. Values of message_size_limit (both the global option
540 and the transport option) were treated as octal if they started with 0.
541 The documentation was vague. These values are now always treated as
542 decimal, and I will make that clear in the documentation.
548 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
549 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
552 MH/01 Don't check that the operands of numeric comparison operators are
553 integers when their expansion is in "skipping" mode (fixes bug
554 introduced by 4.64-PH/07).
556 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
557 child addresses, Exim now panics and dies. Previously, because the count
558 is held in a short int, deliveries were likely to be lost. As such a
559 large number of recipients for a single message is ridiculous
560 (performance will be very, very poor), I have chosen to impose a limit
561 rather than extend the field.
567 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
568 leftover -K file (the existence of which was triggered by #402).
569 While we were at it, introduced process PID as part of the -K
570 filename. This should rule out race conditions when creating
573 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
574 processing considerably. Previous code took too long for large mails,
575 triggering a timeout which in turn triggers #401.
577 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
578 in the DK code in transports.c. sendfile() is not really portable,
579 hence the _LINUX specificness.
581 TF/01 In the add_headers option to the mail command in an Exim filter,
582 there was a bug that Exim would claim a syntax error in any
583 header after the first one which had an odd number of characters
586 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
587 callout verification, Exim cached a "reject" for the entire domain. This
588 is correct for most verifications, but it is not correct for a recipient
589 verification with use_sender or use_postmaster set, because in that case
590 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
591 case of MAIL FROM:<> rejection from other early rejections (e.g.
592 rejection of HELO). When verifying a recipient using a non-null MAIL
593 address, the cache is ignored if it shows MAIL FROM:<> rejection.
594 Whatever the result of the callout, the value of the domain cache is
595 left unchanged (for any other kind of callout, getting as far as trying
596 RCPT means that the domain itself is ok).
598 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
601 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
602 manifest itself as EPIPE rather than ECONNECT. When tidying away a
603 session, the daemon ignores ECONNECT errors and logs others; it now
604 ignores EPIPE as well.
606 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
607 (quoted-printable decoding).
609 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
610 later the small subsequent patch to fix an introduced bug.
612 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
614 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
616 PH/08 An error is now given if message_size_limit is specified negative.
618 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
619 to be given (somewhat) arbitrary names.
621 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
624 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
625 miscellaneous code fixes
627 PH/10 Added the log_reject_target ACL modifier to specify where to log
630 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
631 hostname. This is wrong, because it relates to the incoming message (and
632 probably the interface on which it is arriving) and not to the outgoing
633 callout (which could be using a different interface). This has been
634 changed to use the value of the helo_data option from the smtp transport
635 instead - this is what is used when a message is actually being sent. If
636 there is no remote transport (possible with a router that sets up host
637 addresses), $smtp_active_hostname is used.
639 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
640 tweaks were necessary in order to get it to work (see also 21 below):
641 (a) The code assumed that strncpy() returns a negative number on buffer
642 overflow, which isn't the case. Replaced with Exim's string_format()
644 (b) There were several signed/unsigned issues. I just did the minimum
645 hacking in of casts. There is scope for a larger refactoring.
646 (c) The code used strcasecmp() which is not a standard C function.
647 Replaced with Exim's strcmpic() function.
648 (d) The code set only $1; it now sets $auth1 as well.
649 (e) A simple test gave the error "authentication client didn't specify
650 service in request". It would seem that Dovecot has changed its
651 interface. Fortunately there's a specification; I followed it and
652 changed what the client sends and it appears to be working now.
654 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
657 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
658 address A is aliased to B and C, where B exists and C does not. Without
659 -v the output is "A verified" because verification stops after a
660 successful redirection if more than one address is generated. However,
661 with -v the child addresses are also verified. Exim was outputting "A
662 failed to verify" and then showing the successful verification for C,
663 with its parentage. It now outputs "B failed to verify", showing B's
664 parentage before showing the successful verification of C.
666 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
667 look up a TXT record in a specific list after matching in a combined
670 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
671 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
672 they consult the DNS. I had assumed they would set it the way they
673 wanted; and indeed my experiments on Linux seem to show that in some
674 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
675 To be on the safe side, however, I have now made the interface to
676 host_find_byname() similar to host_find_bydns(), with an argument
677 containing the DNS resolver options. The host_find_byname() function now
678 sets these options at its start, just as host_find_bydns() does. The smtp
679 transport options dns_qualify_single and dns_search_parents are passed to
680 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
681 of host_find_byname() use the default settings of RES_DEFNAMES
682 (qualify_single) but not RES_DNSRCH (search_parents).
684 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
685 spool_read_header() do less string testing, by means of a preliminary
686 switch on the second character of optional "-foo" lines. (This is
687 overdue, caused by the large number of possibilities that now exist.
688 Originally there were few.) While I was there, I also converted the
689 str(n)cmp tests so they don't re-test the leading "-" and the first
690 character, in the hope this might squeeze out yet more improvement.
692 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
693 flag allowing group syntax was set by the header_syntax check but not
694 turned off, possible causing trouble later; (2) The flag was not being
695 set at all for the header_verify test, causing "group"-style headers to
696 be rejected. I have now set it in this case, and also caused header_
697 verify to ignore an empty address taken from a group. While doing this, I
698 came across some other cases where the code for allowing group syntax
699 while scanning a header line wasn't quite right (mostly, not resetting
700 the flag correctly in the right place). These bugs could have caused
701 trouble for malformed header lines. I hope it is now all correct.
703 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
704 with the "reply" argument non-NULL. The code, however (which originally
705 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
706 but it didn't always do it. This confused somebody who was copying the
707 code for some other use. I have removed all the tests.
709 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
710 feature that was used to support insecure browsers during the U.S. crypto
711 embargo. It requires special client support, and Exim is probably the
712 only MTA that supported it -- and would never use it because real RSA is
713 always available. This code has been removed, because it had the bad
714 effect of slowing Exim down by computing (never used) parameters for the
715 RSA_EXPORT functionality.
717 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
718 authenticator to fail if there's a tab character in the incoming data
719 (there should never be unless someone is messing about, as it's supposed
720 to be base64-encoded). Also added, on Timo's advice, the "secured" option
721 if the connection is using TLS or if the remote IP is the same as the
722 local IP, and the "valid-client-cert option" if a client certificate has
725 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
726 authenticators. This can be used for authorization after authentication
727 succeeds. (In the case of plaintext, it servers for both authentication
730 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
731 if any retry times were supplied.
733 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
734 connection, where there is no client IP address to check. In this
735 situation, the verify now always succeeds.
737 PH/25 Applied John Jetmore's -Mset patch.
739 PH/26 Added -bem to be like -Mset, but loading a message from a file.
741 PH/27 In a string expansion for a processed (not raw) header when multiple
742 headers of the same name were present, leading whitespace was being
743 removed from all of them, but trailing whitespace was being removed only
744 from the last one. Now trailing whitespace is removed from each header
745 before concatenation. Completely empty headers in a concatenation (as
748 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
749 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
751 PH/29 [Removed. This was a change that I later backed out, and forgot to
752 correct the ChangeLog entry (that I had efficiently created) before
753 committing the later change.]
755 PH/30 Exim was sometimes attempting to deliver messages that had suffered
756 address errors (4xx response to RCPT) over the same connection as other
757 messages routed to the same hosts. Such deliveries are always "forced",
758 so retry times are not inspected. This resulted in far too many retries
759 for the affected addresses. The effect occurred only when there were more
760 hosts than the hosts_max_try setting in the smtp transport when it had
761 the 4xx errors. Those hosts that it had tried were not added to the list
762 of hosts for which the message was waiting, so if all were tried, there
763 was no problem. Two fixes have been applied:
765 (i) If there are any address or message errors in an SMTP delivery, none
766 of the hosts (tried or untried) are now added to the list of hosts
767 for which the message is waiting, so the message should not be a
768 candidate for sending over the same connection that was used for a
769 successful delivery of some other message. This seems entirely
770 reasonable: after all the message is NOT "waiting for some host".
771 This is so "obvious" that I'm not sure why it wasn't done
772 previously. Hope I haven't missed anything, but it can't do any
773 harm, as the worst effect is to miss an optimization.
775 (ii) If, despite (i), such a delivery is accidentally attempted, the
776 routing retry time is respected, so at least it doesn't keep
777 hammering the server.
779 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
780 in ${readsocket because some servers need this prod.
782 PH/32 Added some extra debug output when updating a wait-xxx database.
784 PH/33 The hint "could be header name not terminated by colon", which has been
785 given for certain expansion errors for a long time, was not being given
786 for the ${if def:h_colon_omitted{... case.
788 PH/34 The spec says: "With one important exception, whenever a domain list is
789 being scanned, $domain contains the subject domain." There was at least
790 one case where this was not true.
792 PH/35 The error "getsockname() failed: connection reset by peer" was being
793 written to the panic log as well as the main log, but it isn't really
794 panic-worthy as it just means the connection died rather early on. I have
795 removed the panic log writing for the ECONNRESET error when getsockname()
798 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
799 runs only) independently of the message's sender address. This meant
800 that, if the 4xx error was in fact related to the sender, a different
801 message to the same recipient with a different sender could confuse
802 things. In particualar, this can happen when sending to a greylisting
803 server, but other circumstances could also provoke similar problems.
804 I have changed the default so that the retry time for these errors is now
805 based a combination of the sender and recipient addresses. This change
806 can be overridden by setting address_retry_include_sender=false in the
809 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
810 remote server are returned as part of bounce messages. This was not
811 happening for LMTP over a pipe (the lmtp transport), but now it is the
812 same for both kinds of LMTP.
814 PH/38 Despite being documented as not happening, Exim was rewriting addresses
815 in header lines that were in fact CNAMEs. This is no longer the case.
817 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
818 and queue runs started by the daemon processed all messages. This has
819 been fixed so that -R and -S can now usefully be given with -q<time>.
821 PH/40 Import PCRE release 6.7 (fixes some bugs).
823 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
825 PH/42 Give an error if -q is specified more than once.
827 PH/43 Renamed the variables $interface_address and $interface_port as
828 $received_ip_address and $received_port, to make it clear that these
829 values apply to message reception, and not to the outgoing interface when
830 a message is delivered. (The old names remain recognized, of course.)
832 PH/44 There was no timeout on the connect() call when using a Unix domain
833 socket in the ${readsocket expansion. There now is.
835 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
836 be meaningful with "accept".
838 SC/01 Eximstats V1.43
839 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
841 SC/02 Eximstats V1.44
842 Use a glob alias rather than an array ref in the generated
843 parser. This improves both readability and performance.
845 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
846 Collect SpamAssassin and rejection statistics.
847 Don't display local sender or destination tables unless
848 there is data to show.
849 Added average volumes into the top table text output.
851 SC/04 Eximstats V1.46
852 Collect data on the number of addresses (recipients)
853 as well as the number of messages.
855 SC/05 Eximstats V1.47
856 Added 'Message too big' to the list of mail rejection
857 reasons (thanks to Marco Gaiarin).
859 SC/06 Eximstats V1.48
860 Mainlog lines which have GMT offsets and are too short to
861 have a flag are now skipped.
863 SC/07 Eximstats V1.49 (Alain Williams)
864 Added the -emptyok flag.
866 SC/08 Eximstats V1.50
867 Fixes for obtaining the IP address from reject messages.
869 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
870 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
871 whitesspace changes from 4.64-PH/27
873 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
876 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
877 are found, allow negative numbers in numeric criteria)
879 JJ/06 exipick.20061117.2, added new $message_body_missing variable
881 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
882 to match changes made in 4.64-PH/43
884 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
886 PH/47 Put in an explicit test for a DNS lookup of an address record where the
887 "domain" is actually an IP address, and force a failure. This locks out
888 those revolvers/nameservers that support "A-for-A" lookups, in
889 contravention of the specifications.
891 PH/48 When a host name was looked up from an IP address, and the subsequent
892 forward lookup of the name timed out, the host name was left in
893 $sender_host_name, contrary to the specification.
895 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
896 restricted to single-key lookups, Exim was not diagnosing an error if
897 * or *@ was used with a query-style lookup.
899 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
901 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
902 long ago, but noone interested enough thought of it. Let's just say that
903 the "1.1" means that there are some new functions that weren't there at
904 some point in the past.
906 PH/51 Error processing for expansion failure of helo_data from an smtp
907 transport during callout processing was broken.
909 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
910 tested/used via the -bh/-bhc/-bs options.
912 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
913 bug, fixed in subsequent PCRE releases).
915 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
916 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
918 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
924 SC/01 Use a glob alias rather than an array ref in eximstats generated
925 parser. This improves both readability and performance.
927 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
928 Don't display local sender or destination tables in eximstats unless
929 there is data to show.
930 Added average volumes into the eximstats top table text output.
932 SC/03 Collect data on the number of addresses (recipients) as well
933 as the number of messages in eximstats.
935 TF/01 Correct an error in the documentation for the redirect router. Exim
936 does (usually) call initgroups() when daemonizing.
938 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
939 with consistent privilege compared to when running as a daemon.
941 TF/03 Note in the spec that $authenticated_id is not set for local
942 submissions from trusted users.
944 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
945 Thanks to Dean Brooks <dean@iglou.com> for the patch.
947 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
948 by adding some example configuration directives to the default
949 configuration file. A little bit of work is required to uncomment the
950 directives and define how usernames and passwords are checked, but
951 there is now a framework to start from.
953 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
954 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
955 without this. I don't know how relevant this is to other LDAP libraries.
957 PH/02 Add the verb name to the "unknown ACL verb" error.
959 PH/03 Magnus Holmgren's patch for filter_prepend_home.
961 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
963 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
964 directory not expanded when it should be if an expanded home directory
965 was set for the address (which is overridden by the transport).
967 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
970 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
971 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
972 because it is too late at that time, and has no effect.
974 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
975 security issue with \' (bugzilla #107). I could not use the
976 PQescapeStringConn() function, because it needs a PGconn value as one of
979 PH/08 When testing addresses using -bt, indicate those final addresses that
980 are duplicates that would not cause an additional delivery. At least one
981 person was confused, thinking that -bt output corresponded to deliveries.
982 (Suppressing duplicates isn't a good idea as you lose the information
983 about possibly different redirections that led to the duplicates.)
985 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
986 systems where poll() doesn't work, in particular OS X.
988 PH/10 Added more information to debugging output for retry time not reached.
990 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
991 operations in malware.c.
993 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
996 PH/13 If write_rejectlog was set false when logging was sent to syslog with
997 syslog_duplication set false, log lines that would normally be written
998 both the the main log and to the reject log were not written to syslog at
1001 PH/14 In the default configuration, change the use of "message" in ACL warn
1002 statements to "add_header".
1004 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1005 not followed by a command (e.g. "seen endif").
1007 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1008 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1011 PH/17 Added extra conditions to the default value of delay_warning_condition
1015 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1016 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1017 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1020 The Auto-Submitted: and various List- headers are standardised, whereas I
1021 don't think Precedence: ever was.
1023 PH/18 Refactored debugging code in route_finduser() to show more information,
1024 in particular, the error code if getpwnam() issues one.
1026 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1027 This is apparently needed in addition to the PH/07 change above to avoid
1028 any possible encoding problems.
1030 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1031 but not after initializing Perl.
1033 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1034 output them only if debugging. By default they are written stderr,
1035 apparently, which is not desirable.
1037 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1040 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1043 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1045 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1046 authenticated or an ident call has been made. Suppress the default
1047 values for $authenticated_id and $authenticated_sender (but permit -oMai
1048 and -oMas) when testing with -bh.
1050 PH/24 Re-jigged the order of the tests in the default configuration so that the
1051 tests for valid domains and recipients precede the DNS black list and CSA
1052 tests, on the grounds that those ones are more expensive.
1054 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1055 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1056 This bug exists in every version of Exim that I still have, right back to
1059 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1060 However, an attempt to turn on case-sensitivity in a regex key by
1061 including (?-i) didn't work because the subject string was already
1062 lowercased, and the effects were non-intuitive. It turns out that a
1063 one-line patch can be used to allow (?-i) to work as expected.
1069 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1070 other effects) broke the use of negated acl sub-conditions.
1072 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1075 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1076 "Deny" causes Exim to reject the incoming connection with a 554 error.
1077 Unfortunately, if there is a major crisis, such as a disk failure,
1078 tcp-wrappers gives "deny", whereas what one would like would be some
1079 kind of temporary error. A kludge has been added to help with this.
1080 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1081 554 error is used if errno is still zero or contains ENOENT (which occurs
1082 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1085 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1087 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1088 errors. Otherwise a message that provokes a temporary error (when other
1089 messages do not) can cause a whole host to time out.
1091 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1092 addresses were routed directly to files or pipes from a redirect router.
1093 File deliveries just didn't batch; pipe deliveries might have suffered
1096 PH/06 A failure to get a lock for a hints database would erroneously always say
1097 "Failed to get write lock", even when it was really a read lock.
1099 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1100 of 0600. This has been changed to use the value of the lockfile_mode
1101 option (which defaults to 0600).
1103 PH/08 Applied small patch from the Sieve maintainer.
1105 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1106 folder from quota calculations, a direct delivery into this folder messed
1107 up the contents of the maildirsize file. This was because the regex was
1108 used only to exclude .Trash (or whatever) when the size of the mailbox
1109 was calculated. There was no check that a delivery was happening into an
1110 excluded directory. This bug has been fixed by ignoring all quota
1111 processing for deliveries into excluded directories.
1113 PH/10 Added the maildirfolder_create_regex option to appendfile.
1119 PH/01 The code for finding all the local interface addresses on a FreeBSD
1120 system running IPv6 was broken. This may well have applied to all BSD
1121 systems, as well as to others that have similar system calls. The broken
1122 code found IPv4 interfaces correctly, but gave incorrect values for the
1123 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1124 that it would not match correctly against @[] and not recognize the IPv6
1127 PH/02 The ipliteral router was not recognizing addresses of the form user@
1128 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1130 PH/03 Added disable_ipv6.
1132 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1133 decoded form, because it is most often used to construct To: headers
1134 lines in autoreplies, and the decoded form may well be syntactically
1135 invalid. However, $reply_address has leading white space removed, and all
1136 newlines turned into spaces so that the autoreply transport does not
1139 PH/05 If group was specified without a user on a router, and no group or user
1140 was specified on a transport, the group from the router was ignored.
1142 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1143 for visible compile-time settings that can be used to change these
1144 numbers, for those that want even more. Backwards compatibility with old
1145 spool files has been maintained. However, going back to a previous Exim
1146 release will lost any variables that are in spool files.
1148 PH/07 Two small changes when running in the test harness: increase delay when
1149 passing a TCP/IP connection to a new process, in case the original
1150 process has to generate a bounce, and remove special handling of
1151 127.0.0.2 (sic), which is no longer necessary.
1153 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1154 be the same on different OS.
1156 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1159 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1160 whether --show-vars was specified or not
1162 JJ/02 exipick: Added support for new ACL variable spool format introduced
1165 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1166 syntactically invalid From: or Reply-to: line, and a filter used this to
1167 generate an autoreply, and therefore failed to obtain an address for the
1168 autoreply, Exim could try to deliver to a non-existent relative file
1169 name, causing unrelated and misleading errors. What now happens is that
1170 it logs this as a hard delivery error, but does not attempt to create a
1173 PH/11 The exinext utility has a -C option for testing purposes, but although
1174 the given file was scanned by exinext itself; it wasn't being passed on
1175 when Exim was called.
1177 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1178 an end-of-file indication when reading a command response.
1180 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1181 compiled. In many other places in Exim, IPv6 addresses are always
1182 recognized, so I have changed this. It also means that IPv4 domain
1183 literals of the form [IPV4:n.n.n.n] are now always recognized.
1185 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1186 used if the router is not running as root, for example, when verifying at
1187 ACL time, or when using -bh. The debugging output from this situation was
1188 non-existent - all you got was a failure to exec. I have made two
1191 (a) Failures to set uid/gid, the current directory, or a process leader
1192 in a subprocess such as that created by queryprogram now generate
1193 suitable debugging ouput when -d is set.
1195 (b) The queryprogram router detects when it is not running as root,
1196 outputs suitable debugging information if -d is set, and then runs
1197 the subprocess without attempting to change uid/gid.
1199 PH/15 Minor change to Makefile for building test_host (undocumented testing
1202 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1203 additional section of a DNS packet that returns MX or SRV records.
1204 Instead, it always explicitly searches for A/AAAA records. This avoids
1205 major problems that occur when a DNS server includes only records of one
1206 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1207 fixed another bug: if SRV records were looked up and the corresponding
1208 address records were *not* found in the additional section, the port
1209 values from the SRV records were lost.
1211 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1212 using the correct key (the original address) when searching the retry
1213 rules in order to find which one to use for generating the retry hint.
1215 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1216 adding the default one. Similarly, if it contains a Reply-To: header, the
1217 errors_reply_to option, if set, is not used.
1219 PH/19 When calculating a retry time, Exim used to measure the "time since
1220 failure" by looking at the "first failed" field in the retry record. Now
1221 it does not use this if it is later than than the arrival time of the
1222 message. Instead it uses the arrival time. This makes for better
1223 behaviour in cases where some deliveries succeed, thus re-setting the
1224 "first failed" field. An example is a quota failure for a huge message
1225 when small messages continue to be delivered. Without this change, the
1226 "time since failure" will always be short, possible causing more frequent
1227 delivery attempts for the huge message than are intended.
1228 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1230 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1231 $1, $2, $3) because the numerical variables can be reset during some
1232 expansion items (e.g. "match"), thereby losing the authentication data.
1234 PH/21 Make -bV show the size of off_t variables so that the test suite can
1235 decide whether to run tests for quotas > 2G.
1237 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1238 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1239 filecount value is greater than 2G or if a quota value is greater than 2G
1240 on a system where the size of off_t is not greater than 4, a panic error
1243 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1244 never match. The debug and -bh output now contains an explicit error
1245 message indicating a malformed IPv4 address or mask.
1247 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1248 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1249 PH/23 above applies.
1251 PH/25 Do not write to syslog when running in the test harness. The only
1252 occasion when this arises is a failure to open the main or panic logs
1253 (for which there is an explicit test).
1255 PH/26 Added the /no_tell option to "control=freeze".
1257 PH/27 If a host name lookup failed very early in a connection, for example, if
1258 the IP address matched host_lookup and the reverse lookup yielded a name
1259 that did not have a forward lookup, an error message of the form "no IP
1260 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1261 could be logged. Now it outputs the IP address instead of "NULL".
1263 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1264 allows the sender and the authenticated sender to be set when
1265 submitting a message from within Exim. Since child_open_exim() is
1266 documented for local_scan(), the new function should be too.
1268 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1269 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1270 results in an empty string is now treated as unset.
1272 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1274 PH/31 Added sender_verify_fail logging option.
1276 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1277 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1278 tidied the source and removed it altogether.
1280 PH/33 When a queue run was abandoned because the load average was too high, a
1281 log line was always written; now it is written only if the queue_run log
1282 selector is set. In addition, the log line for abandonment now contains
1283 information about the queue run such as the pid. This is always present
1284 in "start" and "stop" lines but was omitted from the "abandon" line.
1286 PH/34 Omit spaces between a header name and the colon in the error message that
1287 is given when verify = headers_syntax fails (if there are lots of them,
1288 the message gets confusing).
1290 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1291 names, as there are now some PTR records that contain slashes. This check
1292 is only to protect against broken name servers that fall over on strange
1293 characters, so the fact that it applies to all lookups doesn't matter.
1295 PH/36 Now that the new test suite is complete, we can remove some of the
1296 special code in Exim that was needed for the old test suite. For example,
1297 sorting DNS records because real resolvers return them in an arbitrary
1298 order. The new test suite's fake resolver always returns records in the
1301 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1302 bounces) except when queue_only is set, to avoid logging races between
1303 the different processes.
1305 PH/38 Panic-die if .include specifies a non-absolute path.
1307 PH/39 A tweak to the "H" retry rule from its user.
1309 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1310 a label. They prevented compilation on older perls.
1312 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1313 a warning to be raised on newish perls.
1315 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1316 on queue. Changes to match documented behaviour of showing count of
1317 messages matching specified criteria.
1319 PH/40 Changed the default ident timeout from 30s to 5s.
1321 PH/41 Added support for the use of login_cap features, on those BSD systems
1322 that have them, for controlling the resources used by pipe deliveries.
1324 PH/42 The content-scanning code uses fopen() to create files in which to put
1325 message data. Previously it was not paying any attention to the mode of
1326 the files. Exim runs with umask(0) because the rest of the code creates
1327 files with open(), and sets the required mode explicitly. Thus, these
1328 files were ending up world-writeable. This was not a big issue, because,
1329 being within the spool directory, they were not world-accessible. I have
1330 created a function called modefopen, which takes an additional mode
1331 argument. It sets umask(777), creates the file, chmods it to the required
1332 mode, then resets the umask. All the relevant calls to fopen() in the
1333 content scanning code have been changed to use this function.
1335 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1336 to 24 hours. This avoids potential overflow problems when processing G
1337 and H retry rules. I suspect nobody ever tinkers with this value.
1339 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1341 PH/45 When the plaintext authenticator is running as a client, the server's
1342 challenges are checked to ensure they are valid base64 strings. By
1343 default, the authentication attempt is cancelled if an invalid string is
1344 received. Setting client_ignore_invalid_base64 true ignores these errors.
1345 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1346 they are received. Thus, the responses can be made to depend on the
1347 challenges. If an invalid string is ignored, an empty string is placed in
1350 PH/46 Messages that are created by the autoreply transport now contains a
1351 References: header, in accordance with RFCs 2822 and 3834.
1353 PH/47 Added authenticated_sender_force to the smtp transport.
1355 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1357 PH/49 Installed latest patch from the Sieve maintainer.
1359 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1360 was also set, the appendfile transport was unnecessarily scanning a
1361 directory of message files (e.g. for maildir delivery) to find the count
1362 of files (along with the size), even though it did not need this
1363 information. It now does the scan only if it needs to find either the
1364 size of the count of files.
1366 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1368 PH/52 Two bugs concerned with error handling when the smtp transport is
1371 (i) Exim was not creating retry information for temporary errors given
1372 for individual recipients after the DATA command when the smtp transport
1373 was used in LMTP mode. This meant that they could be retried too
1374 frequently, and not timed out correctly.
1376 (ii) Exim was setting the flag that allows error details to be returned
1377 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1378 recipients that were returned after the DATA command.
1380 PH/53 This is related to PH/52, but is more general: for any failing address,
1381 when detailed error information was permitted to be returned to the
1382 sender, but the error was temporary, then after the final timeout, only
1383 "retry timeout exceeded" was returned. Now it returns the full error as
1384 well as "retry timeout exceeded".
1386 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1387 do this, and (what is worse) MTAs that accept it.
1389 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1390 will now be deprecated.
1392 PH/56 New os.c-cygwin from the Cygwin maintainer.
1394 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1395 formats (previously only available in exim formats via -bpr, -bpru,
1396 and -bpra. Now also available in native and exiqgrep formats)
1398 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1399 with very large, slow to parse queues
1401 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1403 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1405 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1406 responses to authentication challenges, though it was showing the
1407 challenges; (ii) I've removed the CR characters from the debug output for
1410 PH/58 Allow for the insertion of a newline as well as a space when a string
1411 is turned into more than one encoded-word during RFC 2047 encoding. The
1412 Sieve code now uses this.
1414 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1415 data_4xx, lost_connection, tls_required.
1417 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1418 message was being sent as an SMTP response.
1420 PH/61 Add -l and -k options to exicyclog.
1422 PH/62 When verifying, if an address was redirected to one new address, so that
1423 verification continued, and the new address failed or deferred after
1424 having set something in $address_data, the value of $address_data was not
1425 passed back to the ACL. This was different to the case when no
1426 redirection occurred. The value is now passed back in both cases.
1428 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1429 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1430 use login_cap.h, so on its own it isn't the distinguishing feature. The
1431 new name refers directly to the setclassresources() function.
1433 PH/65 Added configuration files for NetBSD3.
1435 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1437 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1438 is preferred over IPv4.
1440 PH/68 The bounce_return_message and bounce_return_body options were not being
1441 honoured for bounces generated during the reception of non-SMTP messages.
1442 In particular, this applied to messages rejected by the ACL. This bug has
1443 been fixed. However, if bounce_return_message is true and bounce_return_
1444 body is false, the headers that are returned for a non-SMTP message
1445 include only those that have been read before the error was detected.
1446 (In the case of an ACL rejection, they have all been read.)
1448 PH/69 The HTML version of the specification is now built in a directory called
1449 spec_html instead of spec.html, because the latter looks like a path with
1450 a MIME-type, and this confuses some software.
1452 PH/70 Catch two compiler warnings in sieve.c.
1454 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1455 function verify_get_ident() calls ip_connect() to connect a socket, but
1456 if the "connect()" function timed out, ip_connect() used to close the
1457 socket. However, verify_get_ident() also closes the socket later, and in
1458 between Exim writes to the log, which may get opened at this point. When
1459 the socket was closed in ip_connect(), the log could get the same file
1460 descriptor number as the socket. This naturally causes chaos. The fix is
1461 not to close the socket in ip_connect(); the socket should be closed by
1462 the function that creates it. There was only one place in the code where
1463 this was missing, in the iplookup router, which I don't think anybody now
1464 uses, but I've fixed it anyway.
1466 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1467 well as to direct DNS lookups. Otherwise the handling of names in host
1468 lists is inconsistent and therefore confusing.
1474 PH/01 Two changes to the default runtime configuration:
1476 (1) Move the checks for relay_from_hosts and authenticated clients from
1477 after to before the (commented out) DNS black list checks.
1479 (2) Add control=submission to the relay_from_hosts and authenticated
1480 clients checks, on the grounds that messages accepted by these
1481 statements are most likely to be submissions.
1483 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1485 (1) Generate an error if the third argument for the ${prvs expansion is
1488 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1491 (3) Reset the variables that are obtained from the first argument of
1492 ${prvscheck and used in the second argument before leaving the code,
1493 because their memory is reclaimed, so using them afterwards may do
1496 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1497 one (it's much easier than Tom thought :-).
1499 (5) Because of (4), we can now allow for the use of $prvscheck_result
1500 inside the third argument.
1502 PH/03 For some reason, the default setting of PATH when running a command from
1503 a pipe transport was just "/usr/bin". I have changed it to
1506 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1507 anything to be listed in the output from -bV.
1509 PH/05 When a filter generated an autoreply, the entire To: header line was
1510 quoted in the delivery log line, like this:
1512 => >A.N.Other <ano@some.domain> <original@ddress> ...
1514 This has been changed so that it extracts the operative address. There
1515 may be more than one such address. If so, they are comma-separated, like
1518 => >ano@some.domain,ona@other.domain <original@ddress> ...
1520 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1521 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1522 not being looked up in the rDNS to get a host name, Exim was showing the
1523 IP address twice in Received: lines, even though the IP addresses were
1524 identical. For example:
1526 Received: from [1.2.3.4] (helo=[1.2.3.4])
1528 However, if the real host name was known, it was omitting the HELO data
1529 if it matched the actual IP address. This has been tidied up so that it
1530 doesn't show the same IP address twice.
1532 PH/07 When both +timestamp and +memory debugging was on, the value given by
1533 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1534 called by the debug printing, thereby overwriting the timestamp buffer.
1535 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1537 PH/08 When the original message was included in an autoreply transport, it
1538 always said "this is a copy of the message, including all the headers",
1539 even if body_only or headers_only was set. It now gives an appropriate
1542 PH/09 Applied a patch from the Sieve maintainer which:
1544 o fixes some comments
1545 o adds the (disabled) notify extension core
1546 o adds some debug output for the result of if/elsif tests
1547 o points to the current vacation draft in the documentation
1548 and documents the missing references header update
1552 o fixes a bug in processing the envelope test (when testing
1553 multiple envelope elements, the last element determinted the
1556 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1557 Electronic Mail") by including:
1559 Auto-submitted: auto-generated
1561 in the messages that it generates (bounce messages and others, such as
1562 warnings). In the case of bounce messages for non-SMTP mesages, there was
1563 also a typo: it was using "Auto_submitted" (underscore instead of
1564 hyphen). Since every message generated by Exim is necessarily in response
1565 to another message, thes have all been changed to:
1567 Auto-Submitted: auto-replied
1569 in accordance with these statements in the RFC:
1571 The auto-replied keyword:
1573 - SHOULD be used on messages sent in direct response to another
1574 message by an automatic process,
1576 - MUST NOT be used on manually-generated messages,
1578 - MAY be used on Delivery Status Notifications (DSNs) and Message
1579 Disposition Notifications (MDNs),
1581 - MUST NOT be used on messages generated by automatic or periodic
1582 processes, except for messages which are automatic responses to
1585 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1586 to the default Received: header definition.
1588 PH/12 Added log selector acl_warn_skipped (default on).
1590 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1591 variables because (a) they are in the wrong storage pool and (b) even if
1592 they were copied, it wouldn't work properly because of the caching.
1594 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1595 checking when decoding. Apparently there are clients that generate
1596 overlong encoded strings. Why am I not surprised?
1598 PH/15 If the first argument of "${if match_address" was not empty, but did not
1599 contain an "@" character, Exim crashed. Now it writes a panic log message
1600 and treats the condition as false.
1602 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1604 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1605 extension "envelope-auth". The code is finished and in agreement with
1606 other implementations, but there is no documentation so far and in fact,
1607 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1608 not changing the active code.
1610 Print executed "if" and "elsif" statements when debugging is used. This
1611 helps a great deal to understand what a filter does.
1613 Document more things not specified clearly in RFC3028. I had all this
1614 sorted out, when out of a sudden new issues came to my mind. Oops."
1616 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1619 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1620 canonical form (as documented). However, after a host name lookup from
1621 the IP address, check_host() was doing a simple string comparison with
1622 addresses acquired from the DNS when checking that the found name did
1623 have the original IP as one of its addresses. Since any found IPv6
1624 addresses are likely to be in abbreviated form, the comparison could
1625 fail. Luckily, there already exists a function for doing the comparison
1626 by converting both addresses to binary, so now that is used instead of
1627 the text comparison.
1629 PH/20 There was another similar case to PH/19, when a complete host name was
1630 given in a host list; looking up its IP address could give an abbreviated
1631 form, whereas the current host's name might or might not be abbreviated.
1632 The same fix has been applied.
1638 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1639 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1642 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1643 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1645 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1647 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1648 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1649 Also, for "Auto-Submitted", treat anything other than "no" as
1650 non-personal, in accordance with RFC 3834. (Previously it treated
1651 anything starting "auto-" as non-personal.)
1653 TF/01 The control=submission/name=... option had a problem with syntax
1654 errors if the name included a slash character. The /name= option
1655 now slurps the rest of the string, so it can include any characters
1656 but it must come last in the list of options (after /sender_retain
1659 PH/05 Some modifications to the interface to the fake nameserver for the new
1667 TK/01 Added the "success_on_redirect" address verification option. See
1668 NewStuff for rationale and an example.
1670 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1672 PH/02 Patch to exigrep to allow it to work on syslog lines.
1674 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1675 fread() to skip over the body file's header line, because in Cygwin the
1676 header line is locked and is inaccessible.
1678 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1679 co-exist for some time) to make it clear that it is the Exim ID that is
1680 referenced, not the Message-ID: header line.
1682 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1683 string_format() function, because snprintf() does not exist on all
1686 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1687 ${stat: expansion item.
1689 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1690 protocol synchronization error", to keep the pedants happy.
1692 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1693 well as for IRIX systems, when gcc is being used. See the host.c source
1696 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1698 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1701 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1702 transport and to the smtp transport in LMTP mode.
1704 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1706 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1707 run a filter in a subprocess. This could lead to confusion in subsequent
1708 lookups in the parent process. There should also be a search_tidyup() at
1709 the end of the subprocess.
1711 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1712 only if the host matched helo_try_verify_hosts, which caused the
1713 verification to occur when the EHLO/HELO command was issued. The ACL just
1714 tested the remembered result. Now, if a previous verification attempt has
1715 not happened, "verify = helo" does it there and then.
1717 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1719 TK/03 Fix log output including CR from clamd.
1721 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1722 exist provoked a memory error which could cause a segfault.
1724 PH/15 Installed PCRE 6.2
1726 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1728 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1729 of the problem. Specifically, suggested +O2 rather than +O1 for the
1732 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1734 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1735 record, the port was not correctly logged when the outgoing_port log
1736 selector was set (it logged the transort's default port).
1738 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1739 fallback_hosts, and "hosts" in the smtp transport.
1741 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1742 host errors such as "Connection refused".
1744 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1745 authentication with radiusclient 0.4.9:
1747 - Error returned from rc_read_config was caught wrongly
1748 - Username/password not passed on to radius server due to wrong length.
1750 The presumption is that some radiusclient API changes for 4.51/PH/17
1751 were not taken care of correctly. The code is still untested by me (my
1752 Linux distribution still has 0.3.2 of radiusclient), but it was
1753 contributed by a Radius user.
1755 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1756 expanding the "port" option of the smtp transport.
1758 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1759 while reading a MIME header. Thanks to Tom Hughes for a patch.
1761 PH/24 Include config.h inside local_scan.h so that configuration settings are
1764 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1765 that in an ACL for RCPT (for example), you can examine exactly what was
1768 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1769 commands, but it was not correctly comparing the address with the actual
1770 client host address. Thus, it would show the EHLO address in Received:
1771 header lines when this was not necessary.
1773 PH/27 Added the % operator to ${eval:}.
1775 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1776 it should be ignoring failures (because with -C, for example, it has lost
1777 privilege). It wasn't ignoring creation failures other than "already
1780 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1783 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1784 because an input line was too long, either on its own, or by virtue of
1785 too many continuations, the temporary file was not being removed, and the
1786 return code was incorrect.
1788 PH/31 Missing "BOOL" in function definition in filtertest.c.
1790 PH/32 Applied Sieve patches from the maintainer.
1792 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1794 PH/33 Added "verify = not_blind".
1796 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1797 Local/Makefile (with some defaults set). These are used in built scripts
1798 such as exicyclog, but they have never been used in the exim_install
1799 script (though there are many overriding facilities there). I have
1800 arranged that the exim_install script now takes note of these two
1803 PH/35 Installed configuration files for Dragonfly.
1805 PH/36 When a locally submitted message by a trusted user did not contain a
1806 From: header, and the sender address was obtained from -f or from an SMTP
1807 MAIL command, and the trusted user did not use -F to supply a sender
1808 name, $originator_name was incorrectly used when constructing a From:
1809 header. Furthermore, $originator_name was used for submission mode
1810 messages from external hosts without From: headers in a similar way,
1811 which is clearly wrong.
1813 PH/37 Added control=suppress_local_fixups.
1815 PH/38 When log_selector = +received_sender was set, and the addition of the
1816 sender made the log line's construction buffer exactly full, or one byte
1817 less than full, an overflow happened when the terminating "\n" was
1820 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1821 when the result of a list match is failure because a DNS lookup failed.
1823 PH/40 RM_COMMAND is now used in the building process.
1825 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1826 the "build-* directories that it finds.
1828 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1829 address in a domain literal was a prefix of an interface address.
1831 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1832 when verifying a sender address, unless rewrite_headers is false.
1834 PH/44 Wrote a long comment about why errors_to addresses are verified as
1835 recipients, not senders.
1837 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1838 the ratelimit ACL was added.
1840 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1842 PH/46 Added extra information about PostgreSQL errors to the error string.
1844 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1845 avoiding the need to install special zones in a real server. This is
1846 backwards compatible; if it can't find the fake resolver, it drops back.
1847 Thus, both old and new test suites can be run.
1849 TF/02 Added util/ratelimit.pl
1851 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1852 clock is set back in time.
1854 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1855 Candler <B.Candler@pobox.com>.
1857 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1858 OK for addresses that are the result of redirections.
1860 PH/48 A number of further additions for the benefit of the new test suite,
1861 including a fake gethostbyname() that interfaces to the fake DNS resolver
1864 TF/06 The fix for widen_domains has also been applied to qualify_single and
1865 search_parents which are the other dnslookup options that can cause
1868 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1871 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1873 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1874 service" if it encountered an MX record with an empty target hostname.
1875 The message is now "an MX or SRV record indicated no SMTP service".
1877 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1878 if the DNS of the sending site is misconfigured. This is quite a
1879 common situation. This change restores the behaviour of treating a
1880 helo verification defer as a failure.
1882 PH/51 If self=fail was set on a router, the bounce message did not include the
1883 actual error message.
1889 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1891 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1892 command itself ended in error, the underlying message about the transport
1893 filter timeout was being overwritten with the pipe command error. Now the
1894 underlying error message should be appended to the second error message.
1896 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1898 PH/02 Reduce the amount of output that "make" produces by default. Full output
1899 can still be requested.
1901 PH/03 The warning log line about a condition test deferring for a "warn" verb
1902 was being output only once per connection, rather than after each
1903 occurrence (because it was using the same function as for successful
1904 "warn" verbs). This seems wrong, so I have changed it.
1906 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1907 it should not have, which might have caused a crash in the right
1908 circumstances, but probably never did.
1910 PH/04 Installed a modified version of Tony Finch's patch to make submission
1911 mode fix the return path as well as the Sender: header line, and to
1912 add a /name= option so that you can make the user's friendly name appear
1915 TF/03 Added the control = fakedefer ACL modifier.
1917 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1918 Mark Lowes for thorough testing.
1920 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1922 TK/03 Merged latest SRS patch from Miles Wilton.
1924 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1925 with the definition in sysexits.h (which is #included earlier).
1926 Fortunately, Exim does not actually use EX_OK. The code used to try to
1927 preserve the sysexits.h value, by assumimg that macro definitions were
1928 scanned for macro replacements. I have been disabused of this notion,
1929 so now the code just undefines EX_OK before #including unistd.h.
1931 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1932 in the smtp transport. When a block could not be written in a single
1933 write() function, the timeout was being re-applied to each part-write.
1934 This seems wrong - if the receiver was accepting one byte at a time it
1935 would take for ever. The timeout is now adjusted when this happens. It
1936 doesn't have to be particularly precise.
1938 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1939 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1941 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1942 without a domain if the check to <postmaster@domain> fails.
1944 SC/01 Eximstats: added -xls and the ability to specify output files
1945 (patch written by Frank Heydlauf).
1947 SC/02 Eximstats: use FileHandles for outputing results.
1949 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1951 SC/04 Eximstats: fixed display of large numbers with -nvr option
1953 SC/05 Eximstats: fixed merging of reports with empty tables.
1955 SC/06 Eximstats: added the -include_original_destination flag
1957 SC/07 Eximstats: removed tabs and trailing whitespace.
1959 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1961 TK/06 MBOX spool code: Add real "From " MBOX separator line
1962 so the .eml file is really in mbox format (even though
1963 most programs do not really care). Patch from Alex Miller.
1965 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1966 The latter is generated from $received_to and is only set if the
1967 message has one envelope recipient. SA can use these headers,
1968 obviously out-of-the-box. Patch from Alex Miller.
1970 PH/08 The ${def test on a variable was returning false if the variable's
1971 value was "0", contrary to what the specification has always said!
1972 The result should be true unless the variable is empty.
1974 PH/09 The syntax error of a character other than { following "${if
1975 def:variable_name" (after optional whitespace) was not being diagnosed.
1976 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1977 accidental colon was present, for example, could give incorrect results.
1979 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1980 result is used (not including appendfile, where other changes are about
1983 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1984 This involved changing a lot of size variables from int to off_t. It
1985 should work with maildirs and everything.
1987 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1988 spamd dying while we are connected to it.
1990 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1993 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1994 function for BDB 4.3.
1996 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1998 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1999 the sources for PCRE have been re-arranged and more files are now
2002 PH/15 The code I had for printing potentially long long variables in PH/11
2003 above was not the best (it lost precision). The length of off_t variables
2004 is now inspected at build time, and an appropriate printing format (%ld
2005 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2006 to be "long long int" or "long int". This is needed for the internal
2007 formatting function string_vformat().
2009 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2010 the configuration file to be ":syslog", then the script "guesses" where
2011 the logs files are, rather than using the compiled in default. In our
2012 case the guess is not the same as the compiled default, so the script
2013 suddenly stopped working when I started to use syslog. The patch checks
2014 to see if log_file_path is "". If so, it attempts to read it from exim
2015 with no configuration file to get the compiled in version, before it
2016 falls back to the previous guessing code."
2018 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2019 implementing BATV in an Exim configuration. See NewStuff for the gory
2022 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2023 Makefile that are specific to HP-UX.
2025 PH/18 If the "use_postmaster" option was set for a recipient callout together
2026 with the "random" option, the postmaster address was used as the MAIL
2027 FROM address for the random test, but not for the subsequent recipient
2028 test. It is now used for both.
2030 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2031 patch removes a few documentation additions to RFC 3028, because the
2032 latest draft now contains them. It adds the new en;ascii-case comparator
2033 and a new error check for 8bit text in MIME parts. Comparator and
2034 require names are now matched exactly. I enabled the subaddress
2035 extension, but it is not well tested yet (read: it works for me)."
2037 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2038 rework some of the code of TK/09 above to avoid the hardwired use of
2039 "%lld" and "long long". Replaced the call to snprintf() with a call to
2042 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2043 records point to non-existent hosts", "retry timeout exceeded", and
2044 "retry time not reached for any host after a long failure period".
2046 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2047 experimental DomainKeys support:
2049 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2050 (2) On an error such as an illegally used "control", the wrong name for
2051 the control was given.
2053 These problems did NOT occur unless DomainKeys support was compiled.
2055 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2057 PH/24 Added ${if match_ip condition.
2059 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2060 databases so that it will be absolutely obvious if a crash occurs in the
2061 DB library. This is a regular occurrence (often caused by mis-matched
2064 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2065 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2066 picked up on a user's system that detects such things. There doesn't seem
2067 to be a gcc warning option for this - only an attribute that has to be
2068 put on the function's prototype. It seems that in Fedora Core 4 they have
2069 set this on a number of new functions. No doubt there will be more in due
2072 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2073 specify a transport. However, if an address that was verified by such a
2074 router was the subject of a callout, Exim crashed because it tried to
2075 read the rcpt_include_affixes from the non-existent transport. Now it
2076 just assumes that the setting of that option is false. This bug was
2077 introduced by 4.51/PH/31.
2079 PH/28 Changed -d+all to exclude +memory, because that information is very
2080 rarely of interest, but it makes the output a lot bigger. People tend to
2081 do -d+all out of habit.
2083 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2084 code in os-type was giving problems when libc.so lives in lib64, like on
2087 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2088 aren't the modern standard, and it seems that some systems' include files
2089 don't always have them. Exim was already checking for some of the newer
2090 ones like T_AAAA, and defining it itself. I've added checks for all the
2091 record types that Exim uses.
2093 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2094 not automatically generating a new one, as it is supposed to. This
2095 prevented TLS from working. If the file did exist, but contained invalid
2096 data, a new version was generated, as expected. It was only the case of a
2097 non-existent file that was broken.
2099 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2100 with a change in libdomainkeys > 0.64.
2102 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2103 from DNS. If the selector record carries the flag, it now has
2104 precedence over the domain-wide flag.
2106 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2108 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2109 the use of a port name that isn't defined in /etc/services) occurred, the
2110 message was deferred as in a normal delivery, and thus remained on the
2111 spool, instead of being failed because of the mua_wrapper setting. This
2112 is now fixed, and I tidied up some of the mua_wrapper messages at the
2115 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2116 the messages in a hash of arrays rather than using individual hashes.
2117 This is a bit cleaner and results in dramatic memory savings, albeit
2118 at a slight CPU cost.
2120 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2121 as requested by Marc Sherman.
2123 SC/10 Eximstats: added histograms for user specified patterns as requested
2126 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2128 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2129 fopen() in the content-scanning modules that did not already have it.
2135 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2136 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2138 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2140 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2142 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2143 to test Sieve filters that use "vacation".
2145 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2146 that changes the way the GnuTLS parameters are stored in the cache file.
2147 The new format can be generated externally. For backward compatibility,
2148 if the data in the cache doesn't make sense, Exim assumes it has read an
2149 old-format file, and it generates new data and writes a new file. This
2150 means that you can't go back to an older release without removing the
2153 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2154 work if there are any delivery delays because "one_time" forces the
2155 parent to be marked "delivered", so its unseen clone is never tried
2156 again. For this reason, Exim now forbids the simultaneous setting of
2159 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2160 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2161 entry if you want to know the details. The fix, however, neglected to
2162 consider the case when local delivery batching is involved. The test for
2163 "previously delivered" was not happening when checking to see if an
2164 address could be batched with a previous (undelivered) one; under
2165 certain circumstances this could lead to multiple deliveries to the same
2168 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2169 in its include files, and this causes problems building Exim.
2171 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2172 header_syntax) but Exim was just ignoring anything given after a slash.
2173 In particular, this caused confusion with an attempt to use "verify =
2174 reverse_host_lookup/defer_ok". An error is now given when options are
2175 supplied for verify items that do not have them. (Maybe reverse_host_
2176 lookup should have a defer_ok option, but that's a different point.)
2178 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2179 defined by RFC 821) to 2048, because there were problems with some AUTH
2180 commands, and RFC 1869 says the size should be increased for extended
2181 SMTP commands that take arguments.
2183 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2186 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2187 "unknown" error; now it says that the functionality isn't in the binary.
2189 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2190 an address' error message when a string expansion fails (syntax or
2191 whatever). Otherwise the password may appear in the log. Following change
2192 PH/42 below, there is no longer a chance of it appearing in a bounce
2195 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2197 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2198 of its addresses was ever tried. (Bugzilla bug #2.)
2200 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2201 the result incorrectly in the debug output. (It correctly added a newline
2202 to what was transported.)
2204 TF/01 Added $received_time.
2206 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2207 commented out examples of how to interface to a virus scanner and to
2208 SpamAssassin. Also added commented examples of av_scanner and
2209 spamd_address settings.
2211 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2212 and controls are allowed in which ACLs. There were a couple of minor
2213 errors. Some of the entries in the conditions table (which is a table of
2214 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2215 negation of where the condition IS allowed.
2217 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2219 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2220 header file does not have a version number, so I've had to invent a new
2221 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2222 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2223 radiusclient), but it was contributed by a Radius user.
2225 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2226 files or directories, for OpenSSL.
2228 PH/19 When an Exim process that is running as root has to create an Exim log
2229 file, it does so in a subprocess that runs as exim:exim so as to get the
2230 ownership right at creation (otherwise, other Exim processes might see
2231 the file with the wrong ownership). There was no test for failure of this
2232 fork() call, which would lead to the process getting stuck as it waited
2233 for a non-existent subprocess. Forks do occasionally fail when resources
2234 run out. I reviewed all the other calls to fork(); they all seem to check
2237 PH/20 When checking for unexpected SMTP input at connect time (before writing
2238 the banner), Exim was not dealing correctly with a non-positive return
2239 from the read() function. If the client had disconnected by this time,
2240 the result was a log entry for a synchronization error with an empty
2241 string after "input=" when read() returned zero. If read() returned -1
2242 (an event I could not check), uninitialized data bytes were printed.
2243 There were reports of junk text (parts of files, etc) appearing after
2246 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2248 PH/22 Added support for macro redefinition, and (re)definition in between
2249 driver and ACL definitions.
2251 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2252 forgetting to use the resulting value; it was using the unexpanded value.
2254 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2255 hadn't been configured. The fix is from Juergen Kreileder, who
2256 understands it better than I do:
2258 "Here's what I see happening with three configured cyrus_sasl
2259 authenticators configured (plain, login, cram-md5):
2261 On startup auth_cyrus_sasl_init() gets called for each of these.
2262 This means three calls to sasl_listmech() without a specified mech_list.
2263 => SASL tests which mechs of all available mechs actually work
2264 => three warnings about OTP not working
2265 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2267 With the patch, sasl_listmech() also gets called three times. But now
2268 SASL's mech_list option is set to the server_mech specified in the the
2269 authenticator. Or in other words, the answer from sasl_listmech()
2270 gets limited to just the mech you're testing for (which is different
2272 => the return list contains just 'plain' or 'login', 'cram-md5' or
2273 nothing depending on the value of ob->server_mech.
2275 I've just tested the patch: Authentication still works fine,
2276 unavailable mechs specified in the exim configuration are still
2277 caught, and the auth.log warnings about OTP are gone."
2279 PH/25 When debugging is enabled, the contents of the command line are added
2280 to the debugging output, even when log_selector=+arguments is not
2283 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2284 answer is "GNU", and only if the return is "GNU/something" is the answer
2287 PH/27 $acl_verify_message is now set immediately after the failure of a
2288 verification in an ACL, and so is available in subsequent modifiers. In
2289 particular, the message can be preserved by coding like this:
2291 warn !verify = sender
2292 set acl_m0 = $acl_verify_message
2294 Previously, $acl_verify_message was set only while expanding "message"
2295 and "log_message" when a very denied access.
2297 PH/28 Modified OS/os.c-Linux with
2299 -#ifndef OS_LOAD_AVERAGE
2300 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2302 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2303 nomenclature these days.)
2305 PH/29 Installed patch from the Sieve maintainer that adds the options
2306 sieve_useraddress and sieve_subaddress to the redirect router.
2308 PH/30 In these circumstances:
2309 . Two addresses routed to the same list of hosts;
2310 . First host does not offer TLS;
2311 . First host accepts first address;
2312 . First host gives temporary error to second address;
2313 . Second host offers TLS and a TLS session is established;
2314 . Second host accepts second address.
2315 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2316 and peerdn, if requested) that were in fact used only for the second
2319 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2320 attention to any local part prefix or suffix that was matched by the
2321 router that accepted the address. It now behaves in the same way as it
2322 does for delivery: the affixes are removed from the local part unless
2323 rcpt_include_affixes is set on the transport.
2325 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2326 timeout during the DATA phase of an incoming message.
2328 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2329 applied a patch sanctioned by the Sieve maintainer.
2331 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2332 the uid or gid is negative. A case of a negative gid caused this to be
2333 noticed. The fix allows for either to be negative.
2335 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2336 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2339 PH/36 The change PH/12 above was broken. Fixed it.
2341 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2342 the grounds that routing the same address twice would always produce the
2343 same answer. This might have been true once, but it is certainly no
2344 longer true now. Routing a child address may depend on the previous
2345 routing that produced that child. Some complicated redirection strategies
2346 went wrong when messages had multiple recipients, and made Exim's
2347 behaviour dependent on the order in which the addresses were given.
2349 I have moved the duplicate checking until after the routing is complete.
2350 Exim scans the addresses that are assigned to local and remote
2351 transports, and removes any duplicates. This means that more work will be
2352 done, as duplicates will always all be routed, but duplicates are
2353 presumably rare, so I don't expect this is of any significance.
2355 For deliveries to pipes, files, and autoreplies, the duplicate checking
2356 still happens during the routing process, since they are not going to be
2359 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2360 It corrects a timeout issue with spamd. This is Ian's comment: "The
2361 background is that sometimes spamd either never reads data from a
2362 connection it has accepted, or it never writes response data. The exiscan
2363 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2364 blindly assumes that writes won't block so it may never time out."
2366 PH/39 Allow G after quota size as well as K and M.
2368 PH/40 The value set for $authenticated_id in an authenticator may not contain
2369 binary zeroes or newlines because the value is written to log lines and
2370 to spool files. There was no check on this. Now the value is run through
2371 the string_printing() function so that such characters are converted to
2372 printable escape sequences.
2374 PH/41 $message_linecount is a new variable that contains the total number of
2375 lines in the message. Compare $body_linecount, which is the count for the
2378 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2379 bounce and delay warning messages, except in certain special cases, which
2382 (a) An SMTP error message from a remote host;
2383 (b) A message specified in a :fail: redirection;
2384 (c) A message specified in a "fail" command in a system filter;
2385 (d) A message specified in a FAIL return from the queryprogram router;
2386 (e) A message specified by the cannot_route_message router option.
2388 In these cases only, Exim does include the error details in bounce and
2389 warning messages. There are also a few cases where bland messages such
2390 as "unrouteable address" or "local delivery error" are given.
2392 PH/43 $value is now also set for the "else" part of a ${run expansion.
2394 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2395 being worked on, but at least Exim now implements the latest version to
2398 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2399 process to complete was treated as a delivery failure, a timeout while
2400 writing the message to the pipe was logged, but erroneously treated as a
2401 successful delivery. Such timeouts include transport filter timeouts. For
2402 consistency with the overall process timeout, these timeouts are now
2403 treated as errors, giving rise to delivery failures by default. However,
2404 there is now a new Boolean option for the pipe transport called
2405 timeout_defer, which, if set TRUE, converts the failures into defers for
2406 both kinds of timeout. A transport filter timeout is now identified in
2409 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2410 systems where "make" and "gmake" are different, calling "gmake" at top
2411 level broke things. I've arranged for the value of $(MAKE) to be passed
2412 from the Makefile to this script so that it can call the same version of
2416 A note about Exim versions 4.44 and 4.50
2417 ----------------------------------------
2419 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2420 changes of various kinds. As a consequence, a big documentation update was
2421 needed. This delayed the release for rather longer than seemed good, especially
2422 in the light of a couple of (minor) security issues. Therefore, the changes
2423 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2424 release. So 4.44 and 4.50 are in effect two different branches that both start
2427 I have left the 4.50 change log unchanged; it contains all the changes since
2428 4.43. The change log for 4.44 is below; many of its items are identical to
2429 those for 4.50. This seems to be the most sensible way to preserve the
2430 historical information.
2436 1. Minor wording change to the doc/README.SIEVE file.
2438 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2439 computation of the current number of files was incorrect.
2441 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2442 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2445 4. Give more explanation in the error message when the command for a transport
2446 filter fails to execute.
2448 5. There are several places where Exim runs a non-Exim command in a
2449 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2450 was being done only for the command run by the queryprogram router. It is
2451 now done for all such subprocesses. The other cases are: ${run, transport
2452 filters, and the commands run by the lmtp and pipe transports.
2454 6. Added CONFIGURE_GROUP build-time option.
2456 7. Some older OS have a limit of 256 on the maximum number of file
2457 descriptors. Exim was using setrlimit() to set 1000 as a large value
2458 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2459 systems. I've change it so that if it can't get 1000, it tries for 256.
2461 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2462 was an oversight, and furthermore, ever since the addition of extra
2463 controls (e.g. 4.43/32), the checks on when to allow different forms of
2464 "control" were broken. There should now be diagnostics for all cases when a
2465 control that does not make sense is encountered.
2467 9. Added the /retain_sender option to "control=submission".
2469 10. $recipients is now available in the predata ACL (oversight).
2471 11. Tidy the search cache before the fork to do a delivery from a message
2472 received from the command line. Otherwise the child will trigger a lookup
2473 failure and thereby defer the delivery if it tries to use (for example) a
2474 cached ldap connection that the parent has called unbind on.
2476 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2477 of $address_data from the recipient verification was clobbered by the
2478 sender verification.
2480 13. The value of address_data from a sender verification is now available in
2481 $sender_address_data in subsequent conditions in the ACL statement.
2483 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2485 15. Added a new option "connect=<time>" to callout options, to set a different
2488 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2489 was its contents. (It was OK if the option was not defined at all.)
2491 17. A "Completed" log line is now written for messages that are removed from
2492 the spool by the -Mrm option.
2494 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2495 information about exactly what failed.
2497 19. Added -dd to debug only the daemon process.
2499 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2500 handles timeouts, both on the server side and network timeouts. Renamed the
2501 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2503 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2504 It is now set to "smtps".
2506 22. $host_address is now set to the target address during the checking of
2507 ignore_target_hosts.
2509 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2510 being passed; this would have caused $sender_host_name to have been used if
2511 matching the list had actually called for a host name (not very likely,
2512 since this list is usually IP addresses). A host name is now passed as
2515 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2516 code that specifies a non-restarting handler (typically sigaction() in
2517 modern systems) in an attempt to fix a rare and obscure crash bug.
2519 25. Narrowed the window for a race in the daemon that could cause it to ignore
2520 SIGCHLD signals. This is not a major problem, because they are used only to
2521 wake it up if nothing else does.
2523 26. A malformed maildirsize file could cause Exim to calculate negative values
2524 for the mailbox size or file count. Odd effects could occur as a result.
2525 The maildirsize information is now recalculated if the size or filecount
2528 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2529 support for a long time. Removed HAVE_SYS_VFS_H.
2531 28. Installed the latest version of exipick from John Jetmore.
2533 29. In an address list, if the pattern was not a regular expression, an empty
2534 subject address (from a bounce message) matched only if the pattern was an
2535 empty string. Non-empty patterns were not even tested. This was the wrong
2536 because it is perfectly reasonable to use an empty address as part of a
2537 database query. An empty address is now tested by patterns that are
2538 lookups. However, all the other forms of pattern expect the subject to
2539 contain a local part and a domain, and therefore, for them, an empty
2540 address still always fails if the pattern is not itself empty.
2542 30. Exim went into a mad DNS loop when attempting to do a callout where the
2543 host was specified on an smtp transport, and looking it up yielded more
2544 than one IP address.
2546 31. Re-factored the code for checking spool and log partition space into a
2547 function that finds that data and another that does the check. The former
2548 is then used to implement four new variables: $spool_space, $log_space,
2549 $spool_inodes, and $log_inodes.
2551 32. The RFC2047 encoding function was originally intended for short strings
2552 such as real names; it was not keeping to the 75-character limit for
2553 encoded words that the RFC imposes. It now respects the limit, and
2554 generates multiple encoded words if necessary. To be on the safe side, I
2555 have increased the buffer size for the ${rfc2047: expansion operator from
2558 33. It is now permitted to omit both strings after an "if" condition; if the
2559 condition is true, the result is "true". As before, when the second string
2560 is omitted, a false condition yields an empty string. This makes it less
2561 cumbersome to write custom ACL and router conditions.
2563 34. Failure to deliver a bounce message always caused it to be frozen, even if
2564 there was an errors_to setting on the router. The errors_to setting is now
2567 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2568 canonical form (fully expanded) before being placed in
2569 $sender_host_address.
2571 36. The table in the code that translates DNS record types into text (T_A to
2572 "A" for instance) was missing entries for NS and CNAME. It is just possible
2573 that this could have caused confusion if both these types were looked up
2574 for the same domain, because the text type is used as part of Exim's
2575 per-process caching. But the chance of anyone hitting this buglet seems
2578 37. The dnsdb lookup has been extended in a number of ways.
2580 (1) There is a new type, "zns", which walks up the domain tree until it
2581 finds some nameserver records. It should be used with care.
2583 (2) There is a new type, "mxh", which is like "mx" except that it returns
2584 just the host names, not the priorities.
2586 (3) It is now possible to give a list of domains (or IP addresses) to be
2587 looked up. The behaviour when one of the lookups defers can be
2588 controlled by a keyword.
2590 (4) It is now possible to specify the separator character for use when
2591 multiple records are returned.
2593 38. The dnslists ACL condition has been extended: it is now possible to supply
2594 a list of IP addresses and/or domains to be looked up in a particular DNS
2597 39. Added log_selector=+queue_time_overall.
2599 40. When running the queue in the test harness, wait just a tad after forking a
2600 delivery process, to get repeatability of debugging output.
2602 41. Include certificate and key file names in error message when GnuTLS fails
2603 to set them up, because the GnuTLS error message doesn't include the name
2604 of the failing file when there is a problem reading it.
2606 42. Allow both -bf and -bF in the same test run.
2608 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2610 44. The "Exiscan patch" is now merged into the mainline Exim source.
2612 45. Sometimes the final signoff response after QUIT could fail to get
2613 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2614 < 0 before doing a fflush(). This bug looks as though it goes back to the
2615 introduction of TLS in release 3.20, but "sometimes" must have been rare
2616 because the tests only now provoked it.
2618 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2619 (this can affect the format of dates).
2621 47. exim_tidydb, when checking for the continued existence of a message for
2622 which it has found a message-specific retry record, was not finding
2623 messages that were in split spool directories. Consequently, it was
2624 deleting retry records that should have stayed in existence.
2626 48. Steve fixed some bugs in eximstats.
2628 49. The SPA authentication driver was not abandoning authentication and moving
2629 on to the next authenticator when an expansion was forced to fail,
2630 contradicting the general specification for all authenticators. Instead it
2631 was generating a temporary error. It now behaves as specified.
2633 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2634 (the order specifies the preference for clients). The order is now AES256,
2635 AES128, 3DES, ARCFOUR128.
2637 51. Small patch to Sieve code - explicitly set From: when generating an
2640 52. Exim crashed if a remote delivery caused a very long error message to be
2641 recorded - for instance if somebody sent an entire SpamAssassin report back
2642 as a large number of 550 error lines. This bug was coincidentally fixed by
2643 increasing the size of one of Exim's internal buffers (big_buffer) that
2644 happened as part of the Exiscan merge. However, to be on the safe side, I
2645 have made the code more robust (and fixed the comments that describe what
2648 53. Now that there can be additional text after "Completed" in log lines (if
2649 the queue_time_overall log selector is set), a one-byte patch to exigrep
2650 was needed to allow it to recognize "Completed" as not the last thing in
2653 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2654 patch that reportedly fixes this has been added. I am not expert enough to
2655 create a test for it. This is what the patch creator wrote:
2657 "I found a little strange behaviour of ldap code when working with
2658 Windows 2003 AD Domain, where users was placed in more than one
2659 Organization Units. When I tried to give exim partial DN, the exit code
2660 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2661 But simultaneously result of request was absolutely normal ldap result,
2662 so I produce this patch..."
2664 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2665 REFERENCE, so I have modified the code to exclude the patch when that macro
2668 55. Some experimental protocols are using DNS PTR records for new purposes. The
2669 keys for these records are domain names, not reversed IP addresses. The
2670 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2671 leaves it alone. Component reversal etc. now happens only for IP addresses.
2674 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2676 57. Double the size of the debug message buffer (to 2048) so that more of very
2677 long debug lines gets shown.
2679 58. The exicyclog utility now does better if the number of log files to keep
2680 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2682 59. Two changes related to the smtp_active_hostname option:
2684 (1) $smtp_active_hostname is now available as a variable.
2685 (2) The default for smtp_banner uses $smtp_active_hostname instead
2686 of $primary_hostname.
2688 60. The host_aton() function is supposed to be passed a string that is known
2689 to be a valid IP address. However, in the case of IPv6 addresses, it was
2690 not checking this. This is a hostage to fortune. Exim now panics and dies
2691 if the condition is not met. A case was found where this could be provoked
2692 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2693 components; fortuitously, this particular loophole had already been fixed
2694 by change 4.50/55 above.
2696 If there are any other similar loopholes, the new check in host_aton()
2697 itself should stop them being exploited. The report I received stated that
2698 data on the command line could provoke the exploit when Exim was running as
2699 exim, but did not say which command line option was involved. All I could
2700 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2701 running as the user.
2704 61. There was a buffer overflow vulnerability in the SPA authentication code
2705 (which came originally from the Samba project). I have added a test to the
2706 spa_base64_to_bits() function which I hope fixes it.
2709 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2710 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2712 63. The daemon start-up calls getloadavg() while still root for those OS that
2713 need the first call to be done as root, but it missed one case: when
2714 deliver_queue_load_max is set with deliver_drop_privilege. This is
2715 necessary for the benefit of the queue runner, because there is no re-exec
2716 when deliver_drop_privilege is set.
2718 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2719 This has been fixed.
2721 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2722 was in use, was not putting the data itself into the right store pool;
2723 consequently, it could be overwritten for a subsequent message in the same
2724 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2727 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2729 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2730 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2731 the function were treating the return as a boolean value, which happened to
2732 work because 0=false and not-0=true, but is not correct code.
2734 68. The host_aton() function was not handling scoped IPv6 addresses (those
2735 with, for example, "%eth0" on the end) correctly.
2737 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2738 negated items (that is, ~something) in unsigned ints. Some compilers
2739 apparently mutter when there is no cast.
2741 70. If an address verification called from an ACL failed, and did not produce a
2742 user-specific message (i.e. there was only a "system" message), nothing was
2743 put in $acl_verify_message. In this situation, it now puts the system
2746 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2747 session; change 4.31/43 added the unwanted input to the log line - except
2748 that it did not do this in the start of session case. It now does.
2750 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2751 This is wrong and can cause the other end to generate a synchronization
2752 error if it is another Exim or anything else that does the synchronization
2753 check. A QUIT command is no longer sent after a timeout.
2755 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2756 during host lookups.
2758 74. The defer_ok option of callout verification was not working if it was used
2759 when verifying addresses in header lines, that is, for this case:
2761 verify = header_sender/callout=defer_ok
2763 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2764 those file descriptors could be used for SMTP connections. If anything
2765 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2766 could be sent to the SMTP client, causing chaos. The daemon now opens
2767 stdin, stdout, and stderr to /dev/null when it puts itself into the
2770 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2771 log by default. The user can override this with suitable Perl magic.
2773 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2774 the log message when discard triggers, was not working for the DATA ACL or
2775 for the non-SMTP ACL.
2777 78. Error message wording change in sieve.c.
2779 79. If smtp_accept_max_per_host was set, the number of connections could be
2780 restricted to fewer than expected, because the daemon was trying to set up
2781 a new connection before checking whether the processes handling previous
2782 connections had finished. The check for completed processes is now done
2783 earlier. On busy systems, this bug wouldn't be noticed because something
2784 else would have woken the daemon, and it would have reaped the completed
2787 80. If a message was submitted locally by a user whose login name contained one
2788 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2789 It caused a spool format error. I have fixed the spool reading code. A
2790 related problem was that the "from" clause in the Received: line became
2791 illegal because of the space(s). It is now covered by ${quote_local_part.
2793 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2796 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2798 83. Patch from Sieve maintainer to fix unterminated string problem in
2799 "vacation" handling.
2801 84. Some minor changes to the Linux configuration files to help with other
2802 OS variants using glibc.
2804 85. One more patch for Sieve to update vacation handling to latest spec.
2807 ----------------------------------------------------
2808 See the note above about the 4.44 and 4.50 releases.
2809 ----------------------------------------------------
2815 1. Change 4.43/35 introduced a bug that caused file counts to be
2816 incorrectly computed when quota_filecount was set in an appendfile
2819 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2820 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2823 3. Give more explanation in the error message when the command for a transport
2824 filter fails to execute.
2826 4. There are several places where Exim runs a non-Exim command in a
2827 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2828 was being done only for the command run by the queryprogram router. It is
2829 now done for all such subprocesses. The other cases are: ${run, transport
2830 filters, and the commands run by the lmtp and pipe transports.
2832 5. Some older OS have a limit of 256 on the maximum number of file
2833 descriptors. Exim was using setrlimit() to set 1000 as a large value
2834 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2835 systems. I've change it so that if it can't get 1000, it tries for 256.
2837 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2838 was an oversight, and furthermore, ever since the addition of extra
2839 controls (e.g. 4.43/32), the checks on when to allow different forms of
2840 "control" were broken. There should now be diagnostics for all cases when a
2841 control that does not make sense is encountered.
2843 7. $recipients is now available in the predata ACL (oversight).
2845 8. Tidy the search cache before the fork to do a delivery from a message
2846 received from the command line. Otherwise the child will trigger a lookup
2847 failure and thereby defer the delivery if it tries to use (for example) a
2848 cached ldap connection that the parent has called unbind on.
2850 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2851 of $address_data from the recipient verification was clobbered by the
2852 sender verification.
2854 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2855 was its contents. (It was OK if the option was not defined at all.)
2857 11. A "Completed" log line is now written for messages that are removed from
2858 the spool by the -Mrm option.
2860 12. $host_address is now set to the target address during the checking of
2861 ignore_target_hosts.
2863 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2864 being passed; this would have caused $sender_host_name to have been used if
2865 matching the list had actually called for a host name (not very likely,
2866 since this list is usually IP addresses). A host name is now passed as
2869 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2870 code that specifies a non-restarting handler (typically sigaction() in
2871 modern systems) in an attempt to fix a rare and obscure crash bug.
2873 15. Narrowed the window for a race in the daemon that could cause it to ignore
2874 SIGCHLD signals. This is not a major problem, because they are used only to
2875 wake it up if nothing else does.
2877 16. A malformed maildirsize file could cause Exim to calculate negative values
2878 for the mailbox size or file count. Odd effects could occur as a result.
2879 The maildirsize information is now recalculated if the size or filecount
2882 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2883 support for a long time. Removed HAVE_SYS_VFS_H.
2885 18. Updated exipick to current release from John Jetmore.
2887 19. Allow an empty sender to be matched against a lookup in an address list.
2888 Previously the only cases considered were a regular expression, or an
2891 20. Exim went into a mad DNS lookup loop when doing a callout where the
2892 host was specified on the transport, if the DNS lookup yielded more than
2895 21. The RFC2047 encoding function was originally intended for short strings
2896 such as real names; it was not keeping to the 75-character limit for
2897 encoded words that the RFC imposes. It now respects the limit, and
2898 generates multiple encoded words if necessary. To be on the safe side, I
2899 have increased the buffer size for the ${rfc2047: expansion operator from
2902 22. Failure to deliver a bounce message always caused it to be frozen, even if
2903 there was an errors_to setting on the router. The errors_to setting is now
2906 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2907 canonical form (fully expanded) before being placed in
2908 $sender_host_address.
2910 24. Updated eximstats to version 1.33
2912 25. Include certificate and key file names in error message when GnuTLS fails
2913 to set them up, because the GnuTLS error message doesn't include the name
2914 of the failing file when there is a problem reading it.
2916 26. Expand error message when OpenSSL has problems setting up cert/key files.
2919 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2920 (this can affect the format of dates).
2922 28. exim_tidydb, when checking for the continued existence of a message for
2923 which it has found a message-specific retry record, was not finding
2924 messages that were in split spool directories. Consequently, it was
2925 deleting retry records that should have stayed in existence.
2927 29. eximstats updated to version 1.35
2928 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2929 1.35 - bugfix such that pie charts by volume are generated correctly
2931 30. The SPA authentication driver was not abandoning authentication and moving
2932 on to the next authenticator when an expansion was forced to fail,
2933 contradicting the general specification for all authenticators. Instead it
2934 was generating a temporary error. It now behaves as specified.
2936 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2937 (the order specifies the preference for clients). The order is now AES256,
2938 AES128, 3DES, ARCFOUR128.
2940 31. Small patch to Sieve code - explicitly set From: when generating an
2943 32. Exim crashed if a remote delivery caused a very long error message to be
2944 recorded - for instance if somebody sent an entire SpamAssassin report back
2945 as a large number of 550 error lines. This bug was coincidentally fixed by
2946 increasing the size of one of Exim's internal buffers (big_buffer) that
2947 happened as part of the Exiscan merge. However, to be on the safe side, I
2948 have made the code more robust (and fixed the comments that describe what
2951 33. Some experimental protocols are using DNS PTR records for new purposes. The
2952 keys for these records are domain names, not reversed IP addresses. The
2953 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2954 leaves it alone. Component reversal etc. now happens only for IP addresses.
2957 34. The host_aton() function is supposed to be passed a string that is known
2958 to be a valid IP address. However, in the case of IPv6 addresses, it was
2959 not checking this. This is a hostage to fortune. Exim now panics and dies
2960 if the condition is not met. A case was found where this could be provoked
2961 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2962 components; fortuitously, this particular loophole had already been fixed
2963 by change 4.50/55 or 4.44/33 above.
2965 If there are any other similar loopholes, the new check in host_aton()
2966 itself should stop them being exploited. The report I received stated that
2967 data on the command line could provoke the exploit when Exim was running as
2968 exim, but did not say which command line option was involved. All I could
2969 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2970 running as the user.
2973 35. There was a buffer overflow vulnerability in the SPA authentication code
2974 (which came originally from the Samba project). I have added a test to the
2975 spa_base64_to_bits() function which I hope fixes it.
2978 36. The daemon start-up calls getloadavg() while still root for those OS that
2979 need the first call to be done as root, but it missed one case: when
2980 deliver_queue_load_max is set with deliver_drop_privilege. This is
2981 necessary for the benefit of the queue runner, because there is no re-exec
2982 when deliver_drop_privilege is set.
2984 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2985 was in use, was not putting the data itself into the right store pool;
2986 consequently, it could be overwritten for a subsequent message in the same
2987 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2990 38. Sometimes the final signoff response after QUIT could fail to get
2991 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2992 < 0 before doing a fflush(). This bug looks as though it goes back to the
2993 introduction of TLS in release 3.20, but "sometimes" must have been rare
2994 because the tests only now provoked it.
3000 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3001 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3002 Now it returns an integer. A number of calls were still expecting a T/F
3003 return. Fortuitously, in all cases, the tests worked in OK situations,
3004 which is the norm. However, things would have gone wrong on any write
3005 failures on the smtp file descriptor. This function is used when sending
3006 messages over SMTP and also when doing verify callouts.
3008 2. When Exim is called to do synchronous delivery of a locally submitted
3009 message (the -odf or -odi options), it no longer closes stderr before doing
3012 3. Implemented the mua_wrapper option.
3014 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3016 5. Implemented the functions header_remove(), header_testname(),
3017 header_add_at_position(), and receive_remove_recipient(), and exported them
3020 6. If an ACL "warn" statement specified the addition of headers, Exim already
3021 inserted X-ACL-Warn: at the start if there was no header name. However, it
3022 was not making this test for the second and subsequent header lines if
3023 there were newlines in the string. This meant that an invalid header could
3024 be inserted if Exim was badly configured.
3026 7. Allow an ACL "warn" statement to add header lines at the start or after all
3027 the Received: headers, as well as at the end.
3029 8. Added the rcpt_4xx retry error code.
3031 9. Added postmaster_mailfrom=xxx to callout verification option.
3033 10. Added mailfrom=xxxx to the callout verification option, for verify=
3036 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3037 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3039 12. Inserted some casts to stop certain compilers warning when using pointer
3040 differences as field lengths or precisions in printf-type calls (mostly
3041 affecting debugging statements).
3043 13. Added optional readline() support for -be (dynamically loaded).
3045 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3046 same clock tick as a message's arrival, so that its received time was the
3047 same as the "first fail" time on the retry record, and that message
3048 remained on the queue past the ultimate address timeout, every queue runner
3049 would try a delivery (because it was past the ultimate address timeout) but
3050 after another failure, the ultimate address timeout, which should have then
3051 bounced the address, did not kick in. This was a "< instead of <=" error;
3052 in most cases the first failure would have been in the next clock tick
3053 after the received time, and all would be well.
3055 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3056 being recognized when the domain list was tested by the match_domain
3057 condition in an expansion string.
3059 16. Added the ${str2b64: operator.
3061 17. Exim was always calling setrlimit() to set a large limit for the number of
3062 processes, without checking whether the existing limit was already
3063 adequate. (It did check for the limit on file descriptors.) Furthermore,
3064 errors from getrlimit() and setrlimit() were being ignored. Now they are
3065 logged to the main and panic logs, but Exim does carry on, to try to do its
3066 job under whatever limits there are.
3068 18. Imported PCRE 5.0.
3070 19. Trivial typo in log message " temporarily refused connection" (the leading
3073 20. If the log selector return_path_on_delivery was set and an address was
3074 redirected to /dev/null, the delivery process crashed because it assumed
3075 that a return path would always be set for a "successful" delivery. In this
3076 case, the whole delivery is bypassed as an optimization, and therefore no
3079 21. Internal re-arrangement: the function for sending a challenge and reading
3080 a response while authentication was assuming a zero-terminated challenge
3081 string. It's now changed to take a pointer and a length, to allow for
3082 binary data in such strings.
3084 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3086 23. Exim was not respecting finduser_retries when seeking the login of the
3087 uid under which it was called; it was always trying 10 times. (The default
3088 setting of finduser_retries is zero.) Also, it was sleeping after the final
3089 failure, which is pointless.
3091 24. Implemented tls_on_connect_ports.
3093 25. Implemented acl_smtp_predata.
3095 26. If the domain in control=submission is set empty, Exim assumes that the
3096 authenticated id is a complete email address when it generates From: or
3097 Sender: header lines.
3099 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3100 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3101 chown and chgrp in /bin and hostname in /usr/bin.
3103 28. Exim was keeping the "process log" file open after each use, just as it
3104 does for the main log. This opens the possibility of it remaining open for
3105 long periods when the USR1 signal hits a daemon. Occasional processlog
3106 errors were reported, that could have been caused by this. Anyway, it seems
3107 much more sensible not to leave this file open at all, so that is what now
3110 29. The long-running daemon process does not normally write to the log once it
3111 has entered its main loop, and it closes the log before doing so. This is
3112 so that log files can straightforwardly be renamed and moved. However,
3113 there are a couple of unusual error situations where the daemon does write
3114 log entries, and I had neglected to close the log afterwards.
3116 30. The text of an SMTP error response that was received during a remote
3117 delivery was being truncated at 512 bytes. This is too short for some of
3118 the long messages that one sometimes sees. I've increased the limit to
3121 31. It is now possible to make retry rules that apply only when a message has a
3122 specific sender, in particular, an empty sender.
3124 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3125 it possible to be selective about when SMTP synchronization is enforced.
3127 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3129 32. Implemented hosts_connection_nolog.
3131 33. Added an ACL for QUIT.
3133 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3136 35. Added mailbox_size and mailbox_filecount to appendfile.
3138 36. Added control = no_multiline_responses to ACLs.
3140 37. There was a bug in the logic of the code that waits for the clock to tick
3141 in the case where the clock went backwards by a substantial amount such
3142 that the microsecond fraction of "now" was more than the microsecond
3143 fraction of "then" (but the whole seconds number was less).
3145 38. Added support for the libradius Radius client library this is found on
3146 FreeBSD (previously only the radiusclient library was supported).
3152 1. When certain lookups returned multiple values in the form name=value, the
3153 quoting of the values was not always being done properly. Specifically:
3154 (a) If the value started with a double quote, but contained no whitespace,
3156 (b) If the value contained whitespace other than a space character (i.e.
3157 tabs or newlines or carriage returns) it was not quoted.
3158 This fix has been applied to the mysql and pgsql lookups by writing a
3159 separate quoting function and calling it from the lookup code. The fix
3160 should probably also be applied to nisplus, ibase and oracle lookups, but
3161 since I cannot test any of those, I have not disturbed their existing code.
3163 2. A hit in the callout cache for a specific address caused a log line with no
3164 reason for rejecting RCPT. Now it says "Previous (cached) callout
3165 verification failure".
3167 3. There was an off-by-one bug in the queryprogram router. An over-long
3168 return line was truncated at 256 instead of 255 characters, thereby
3169 overflowing its buffer with the terminating zero. As well as fixing this, I
3170 have increased the buffer size to 1024 (and made a note to document this).
3172 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3173 when Exim is waiting for an SMTP response from a remote server, Exim
3174 restarts its select() call on the socket, thereby resetting its timeout.
3175 This is not a problem when such interrupts are rare. Somebody set up a cron
3176 job to run exiwhat every 2 minutes, which is less than the normal select()
3177 timeout (5 or 10 minutes). This meant that the select() timeout never
3178 kicked in because it was always reset. I have fixed this by comparing the
3179 time when an interrupt arrives with the time at the start of the first call
3180 to select(). If more time than the timeout has elapsed, the interrupt is
3181 treated as a timeout.
3183 5. Some internal re-factoring in preparation for the addition of Sieve
3184 extensions (by MH). In particular, the "personal" test is moved to a
3185 separate function, and given an option for scanning Cc: and Bcc: (which is
3186 not set for Exim filters).
3188 6. When Exim created an email address using the login of the caller as the
3189 local part (e.g. when creating a From: or Sender: header line), it was not
3190 quoting the local part when it contained special characters such as @.
3192 7. Installed new OpenBSD configuration files.
3194 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3195 try to make them clearer.
3197 9. Callout options, other than the timeout value, were being ignored when
3198 verifying sender addresses in header lines. For example, when using
3200 verify = header_sender/callout=no_cache
3202 the cache was (incorrectly) being used.
3204 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3205 only the Cygwin environment.
3207 11. When return_path_on_delivery was set as a log selector, if different remote
3208 addresses in the same message used different return paths and parallel
3209 remote delivery occurred, the wrong values would sometimes be logged.
3210 (Whenever a remote delivery process finished, the return path value from
3211 the most recently started remote delivery process was logged.)
3213 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3214 lines when AUTH and/or TLS are in use. This is the "received protocol"
3215 field. Exim used to use "asmtp" for authenticated SMTP, without any
3216 indication (in the protocol name) for TLS use. Now it follows the RFC and
3217 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3218 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3219 names appear in log lines as well as in Received: header lines.
3221 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3222 extensions, and comparison tests, and to fix some bugs.
3224 14. Changes to the "personal" filter test:
3226 (1) The test was buggy in that it was just doing the equivalent of
3227 "contains" tests on header lines. For example, if a user's address was
3228 anne@some.where, the "personal" test would incorrectly be true for
3230 To: susanne@some.where
3232 This test is now done by extracting each address from the header in turn,
3233 and checking the entire address. Other tests that are part of "personal"
3234 are now done using regular expressions (for example, to check local parts
3235 of addresses in From: header lines).
3237 (2) The list of non-personal local parts in From: addresses has been
3238 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3239 taken from the Sieve specification recommendations.
3241 (3) If the message contains any header line starting with "List-" it is
3242 treated as non-personal.
3244 (4) The test for "circular" in the Subject: header line has been removed
3245 because it now seems ill-conceived.
3247 15. Minor typos in src/EDITME comments corrected.
3249 16. Installed latest exipick from John Jetmore.
3251 17. If headers_add on a router specified a text string that was too long for
3252 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3253 of string_sprintf() is now avoided.
3255 18. $message_body_size was not set (it was always zero) when running the DATA
3256 ACL and the local_scan() function.
3258 19. For the "mail" command in an Exim filter, no default was being set for
3259 the once_repeat time, causing a random time value to be used if "once" was
3260 specified. (If the value happened to be <= 0, no repeat happened.) The
3261 default is now 0s, meaning "never repeat". The "vacation" command was OK
3262 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3263 (I found it when inspecting the code).
3265 20. There is now an overall timeout for performing a callout verification. It
3266 defaults to 4 times the callout timeout, which applies to individual SMTP
3267 commands during the callout. The overall timeout applies when there is more
3268 than one host that can be tried. The timeout is checked before trying the
3269 next host. This prevents very long delays if there are a large number of
3270 hosts and all are timing out (e.g. when the network connections are timing
3271 out). The value of the overall timeout can be changed by specifying an
3272 additional sub-option for "callout", called "maxwait". For example:
3274 verify = sender/callout=5s,maxwait=20s
3276 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3277 to the end before writing, but this should make it even safer).
3279 22. Exim was forgetting that it had advertised PIPELINING for the second and
3280 subsequent messages on an SMTP connection. It was also not resetting its
3281 memory on STARTTLS and an internal HELO.
3283 23. When Exim logs an SMTP synchronization error within a session, it now
3284 records whether PIPELINING has been advertised or not.
3286 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3287 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3288 rather than long int.
3290 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3292 26. Added the never_mail option to autoreply.
3298 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3299 crash if the getsockname() call failed; this can happen if a connection is
3300 closed very soon after it is established. The problem was simply in the
3301 order in which certain operations were done, causing Exim to try to write
3302 to the SMTP stream before it had set up the file descriptor. The bug has
3303 been fixed by making things happen in the correct order.
3309 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3310 before the connection was closed, thus losing the rejection response.
3312 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3313 some early Solaris releases, but causes trouble in current releases where
3314 socklen_t is defined.
3316 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3319 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3322 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3323 that releasing the top of it at the end releases what was used for sub-
3324 expansions (unless the block got too big). However, discard this block if
3325 the first thing is a variable or header, so that we can use its block when
3326 it is dynamic (useful for very large $message_headers, for example).
3328 6. Lookups now cache *every* query, not just the most recent. A new, separate
3329 store pool is used for this. It can be recovered when all lookup caches are
3330 flushed. Lookups now release memory at the end of their result strings.
3331 This has involved some general refactoring of the lookup sources.
3333 7. Some code has been added to the store_xxx() functions to reduce the amount
3334 of flapping under certain conditions.
3336 8. log_incoming_interface used to affect only the <= reception log lines. Now
3337 it causes the local interface and port to be added to several more SMTP log
3338 lines, for example "SMTP connection from", and rejection lines.
3340 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3342 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3344 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3345 could be overwritten at the end of the current message (or the start of a
3346 new message if it was set in a HELO ACL). The value is now preserved for
3347 the duration of the SMTP connection.
3349 12. If a transport had a headers_rewrite setting, and a matching header line
3350 contained an unqualified address, that address was qualified, even if it
3351 did not match any rewriting rules. The underlying bug was that the values
3352 of the flags that permit the existence of unqualified sender and recipient
3353 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3354 non-local messages, and by -bnq for local messages) were not being
3355 preserved with the message after it was received.
3357 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3358 "next input=" as part of the text comprising the host identity instead of
3359 the correct text. The code was using the same buffer for two different
3360 strings. However, depending on which order the printing function evaluated
3361 its arguments, the bug did not always show up. Under Linux, for example, my
3362 test suite worked just fine.
3364 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3365 doesn't work with some older versions of Perl. It has been changed to "my",
3366 which in any case is probably the better facility to use.
3368 15. A really picky compiler found some instances of statements for creating
3369 error messages that either had too many or two few arguments for the format
3372 16. The size of the buffer for calls to the DNS resolver has been increased
3373 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3374 for addresses that have a lot of PTR records. This alleviates a problem; it
3375 does not fully solve it.
3377 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3378 buffer now truncates the list and logs the incident, which is the same
3379 action as happens when Exim is looking up a host name and its aliases.
3380 Previously in this situation something unpredictable would happen;
3381 sometimes it was "internal error: store_reset failed".
3383 18. If a server dropped the connection unexpectedly when an Exim client was
3384 using GnuTLS and trying to read a response, the client delivery process
3385 crashed while trying to generate an error log message.
3387 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3388 string, for example:
3390 warn message = H1: something\nH2: something
3392 the text was added as a single header line from Exim's point of view
3393 though it ended up OK in the delivered message. However, searching for the
3394 second and subsequent header lines using $h_h2: did not work. This has been
3395 fixed. Similarly, if a system filter added multiple headers in this way,
3396 the routers could not see them.
3398 20. Expanded the error message when iplsearch is called with an invalid key to
3399 suggest using net-iplsearch in a host list.
3401 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3402 ACLs are no longer actually imposed (and a message to that effect is
3405 22. If a "gecos" field in a passwd entry contained escaped characters, in
3406 particular, if it contained a \" sequence, Exim got it wrong when building
3407 a From: or a Sender: header from that name. A second bug also caused
3408 incorrect handling when an unquoted " was present following a character
3409 that needed quoting.
3411 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3412 was not being matched caselessly.
3414 24. Arranged for all hyphens in the exim.8 source to be escaped with
3417 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3418 properly. Recipient callout cache records were still being keyed to include
3419 the sender, even when use_sender was set false. This led to far more
3420 callouts that were necessary. The sender is no longer included in the key
3421 when use_sender is false.
3423 26. Added "control = submission" modifier to ACLs.
3425 27. Added the ${base62d: operator to decode base 62 numbers.
3427 28. dnsdb lookups can now access SRV records.
3429 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3430 the configuration file.
3432 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3433 (-v) mode. This makes the output for a verbose queue run more intelligible.
3435 31. Added a use_postmaster feature to recipient callouts.
3437 32. Added the $body_zerocount variable, containing the number of binary zero
3438 bytes in the message body.
3440 33. The time of last modification of the "new" subdirectory is now used as the
3441 "mailbox time last read" when there is a quota error for a maildir
3444 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3446 35. Added +ignore_unknown as a special item in host lists.
3448 36. Code for decoding IPv6 addresses in host lists is now included, even if
3449 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3450 address was recognized as an IP address, but was then not correctly decoded
3451 into binary, causing unexpected and incorrect effects when compared with
3458 1. Very minor rewording of debugging text in manualroute to say "list of
3459 hosts" instead of "hostlist".
3461 2. If verify=header_syntax was set, and a header line with an unqualified
3462 address (no domain) and a large number of spaces between the end of the
3463 name and the colon was received, the reception process suffered a buffer
3464 overflow, and (when I tested it) crashed. This was caused by some obsolete
3465 code that should have been removed. The fix is to remove it!
3467 3. When running in the test harness, delay a bit after writing a bounce
3468 message to get a bit more predictability in the log output.
3470 4. Added a call to search_tidyup() just before forking a reception process. In
3471 theory, someone could use a lookup in the expansion of smtp_accept_max_
3472 per_host which, without the tidyup, could leave open a database connection.
3474 5. Added the variables $recipient_data and $sender_data which get set from a
3475 lookup success in an ACL "recipients" or "senders" condition, or a router
3476 "senders" option, similar to $domain_data and $local_part_data.
3478 6. Moved the writing of debug_print from before to after the "senders" test
3481 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3482 problems for message scanning, either using a data ACL, or using
3483 local_scan() because the Received: header was not generated till after they
3484 were called (in order to set the time as the time of reception completion).
3485 I have revised the way this works. The header is now generated after the
3486 body is received, but before the ACL or local_scan() are called. After they
3487 are run, the timestamp in the header is updated.
3493 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3494 before starting a queue runner without re-exec. This happened only when
3495 deliver_drop_privilege was set or when the Exim user was set to root. The
3496 effect of the bug was that timeouts during subsequent deliveries caused
3497 crashes instead of being properly handled. The handler is now left at its
3498 default (and expected) setting.
3500 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3501 message, again when deliver_drop_privilege is set or Exim is run as root.
3502 The bug described in (1) was not present in this case, but the tidying up
3503 of the other signals was missing. I have made the two cases consistent.
3505 3. The ignore_target_hosts setting on a manualroute router was being ignored
3506 for hosts that were looked up using the /MX notation.
3508 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3511 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3512 operated on the sender address. After changing the $sender_address to <>
3513 for the sender address verify, Exim was re-instated it as the original
3514 (before rewriting) address, but remembering that it had rewritten it, so it
3515 wasn't rewriting it again. This bug also had the effect of breaking the
3516 sender address verification caching when the sender address was rewritten.
3518 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3519 This has been changed so that if the ip literal address matches
3520 ignore_target_hosts, the router declines.
3522 7. Added expansion conditions match_domain, match_address, and match_local_
3523 part (NOT match_host).
3525 8. The placeholder for the Received: header didn't have a length field set.
3527 9. Added code to Exim itself and to exim_lock to test for a specific race
3528 condition that could lead to file corruption when using MBX delivery. The
3529 issue is with the lockfile that is created in /tmp. If this file is removed
3530 after a process has opened it but before that process has acquired a lock,
3531 there is the potential for a second process to recreate the file and also
3532 acquire a lock. This could lead to two Exim processes writing to the file
3533 at the same time. The added code performs the same test as UW imapd; it
3534 checks after acquiring the lock that its file descriptor still refers to
3535 the same named file.
3537 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3538 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3541 11. Added the smtp_active_hostname option. If used, this will typically be made
3542 to depend on the incoming interface address. Because $interface_address is
3543 not set up until the daemon has forked a reception process, error responses
3544 that can happen earlier (such as "too many connections") no longer contain
3547 12. If an expansion in a condition on a "warn" statement fails because a lookup
3548 defers, the "warn" statement is abandoned, and the next ACL statement is
3549 processed. Previously this caused the whole ACL to be aborted.
3551 13. Added the iplsearch lookup type.
3553 14. Added ident_timeout as a log selector.
3555 15. Added tls_certificate_verified as a log selector.
3557 16. Added a global option tls_require_ciphers (compare the smtp transport
3558 option of the same name). This controls incoming TLS connections.
3560 17. I finally figured out how to make tls_require_ciphers do a similar thing
3561 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3562 before starting the TLS session.
3564 18. Tabs are now shown as \t in -bP output.
3566 19. If the log selector return_path_on_delivery was set, Exim crashed when
3567 bouncing a message because it had too many Received: header lines.
3569 20. If two routers both had headers_remove settings, and the first one included
3570 a superfluous trailing colon, the final name in the first list and the
3571 first name in the second list were incorrectly joined into one item (with a
3572 colon in the middle).
3578 1. Added -C and -D options to the exinext utility, mainly to make it easier
3579 to include in the automated testing, but these could be helpful when
3580 multiple configurations are in use.
3582 2. The exinext utility was not formatting the output nicely when there was
3583 an alternate port involved in the retry record key, nor when there was a
3584 message id as well (for retries that were specific to a specific message
3585 and a specific host). It was also confused by IPv6 addresses, because of
3586 the additional colons they contain. I have fixed the IPv4 problem, and
3587 patched it up to do a reasonable job for IPv6.
3589 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3590 delivery, the log line now contains "pipelined" if PIPELINING was used.
3592 4. An SMTP transport process used to panic and die if the bind() call to set
3593 an explicit outgoing interface failed. This has been changed; it is now
3594 treated in the same way as a connect() failure.
3596 5. A reference to $sender_host_name in the part of a conditional expansion
3597 that was being skipped was still causing a DNS lookup. This no longer
3600 6. The def: expansion condition was not recognizing references to header lines
3601 that used bh_ and bheader_.
3603 7. Added the _cache feature to named lists.
3605 8. The code for checking quota_filecount in the appendfile transport was
3606 allowing one more file than it should have been.
3608 9. For compatibility with Sendmail, the command line option
3616 and sets the incoming protocol and host name (for trusted callers). The
3617 host name and its colon can be omitted when only the protocol is to be set.
3618 Note the Exim already has two private options, -pd and -ps, that refer to
3619 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3620 "s", but I don't think that's a major issue.
3622 10. A number of refactoring changes to the code, none of which should affect
3625 (a) The number of logging options was getting close to filling up the
3626 32-bit word that was used as a bit map. I have split them into two classes:
3627 those that are passed in the argument to log_write(), and those that are
3628 only ever tested independently outside of that function. These are now in
3629 separate 32-bit words, so there is plenty of room for expansion again.
3630 There is no change in the user interface or the logging behaviour.
3632 (b) When building, for example, log lines, the code previously used a
3633 macro that called string_cat() twice, in order to add two strings. This is
3634 not really sufficiently general. Furthermore, there was one instance where
3635 it was actually wrong because one of the argument was used twice, and in
3636 one call a function was used. (As it happened, calling the function twice
3637 did not affect the overall behaviour.) The macro has been replaced by a
3638 function that can join an arbitrary number of extra strings onto a growing
3641 (c) The code for expansion conditions now uses a table and a binary chop
3642 instead of a serial search (which was left over from when there were very
3643 few conditions). Also, it now recognizes conditions like "pam" even when
3644 the relevant support is not compiled in: a suitably worded error message is
3645 given if an attempt is made to use such a condition.
3647 11. Added ${time_interval:xxxxx}.
3649 12. A bug was causing one of the ddress fields not to be passed back correctly
3650 from remote delivery subprocesses. The field in question was not being
3651 subsequently used, so this caused to problems in practice.
3653 13. Added new log selectors queue_time and deliver_time.
3655 14. Might have fixed a bug in maildirsizefile handling that threw up
3656 "unexpected character" debug warnings, and recalculated the data
3657 unnecessarily. In any case, I expanded the warning message to give more
3660 15. Added the message "Restricted characters in address" to the statements in
3661 the default ACL that block characters like @ and % in local parts.
3663 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3664 Three changes have been made:
3666 (a) There was a serious bug; a negative response to MAIL caused the whole
3667 recipient domain to be cached as invalid, thereby blocking all messages
3668 to all local parts at the same domain, from all senders. This bug has
3669 been fixed. The domain is no longer cached after a negative response to
3670 MAIL if the sender used is not empty.
3672 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3675 (c) A new callout option, "use_sender" has been added for people who want
3676 the modified behaviour.
3682 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3685 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3686 indeed breaks things for older releases.
3688 3. Added additional logging to the case where there is a problem reading data
3689 from a filter that is running in a subprocess using a pipe, in order to
3690 try to track down a specific problem.
3692 4. Testing facility fudge: when running in the test harness and attempting
3693 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3694 getting "No route to host". Convert this to a timeout.
3696 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3699 6. Some OS don't have socklen_t but use size_t instead. This affects the
3700 fifth argument of getsockopt() amongst other things. This is now
3701 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3702 can be set for individual OS. I have set it for SunOS5, OSF1, and
3703 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3704 some earlier ones do not.
3706 7. Change 4.30/15 was not doing the test caselessly.
3708 8. The standard form for an IPv6 address literal was being rejected by address
3709 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3710 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3711 this, as well as the form without the "IPv6" on the front (but only when
3712 address literals are enabled, of course).
3714 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3716 10. Exim crashed if a message with an empty sender address specified by -f
3717 encountered a router with an errors_to setting. This could be provoked only
3718 by a command such as
3722 where an empty string was supplied; "<>" did not hit this bug.
3724 11. Installed PCRE release 4.5.
3726 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3727 remained set. It is now erased.
3729 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3730 times from message ids (which are base 36 rather than the normal 62).
3732 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3733 were being counted as actual protocol errors, and logged if the log
3734 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3735 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3736 and DATA following a set of rejected RCPTs do not count as protocol errors.
3737 In other words, Exim assumes they were pipelined, though this may not
3738 actually be the case. Of course, in all cases the client gets an
3739 appropriate error code.
3741 15. If a lookup fails in an ACL condition, a message about the failure may
3742 be available; it is used if testing the ACL cannot continue, because most
3743 such messages specify what the cause of the deferral is. However, some
3744 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3745 that caused an old message to be retained and used if a later statement
3746 caused a defer, replacing the real cause of the deferral.
3748 16. If an IP address had so many PTR records that the DNS lookup buffer
3749 was not large enough to hold them, Exim could crash while trying to process
3750 the truncated data. It now detects and logs this case.
3752 17. Further to 4.21/58, another change has been made: if (and only if) the
3753 first line of a message (the first header line) ends with CRLF, a bare LF
3754 in a subsequent header line has a space inserted after it, so as not to
3755 terminate the header.
3757 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3758 unnecessarily, used atoi() instead of strtol(), and didn't check the
3759 termination when getting file sizes from file names by regex.
3761 19. Completely re-implemented the support for maildirsize files, in the light
3762 of a number of problems with the previous contributed implementation
3763 (4.30/29). In particular:
3765 . If the quota is zero, the maildirsize file is maintained, but no quota is
3768 . If the maildir directory does not exist, it is created before any attempt
3769 to write a maildirsize file.
3771 . The quota value in the file is just a cache; if the quota is changed in
3772 the transport, the new value overrides.
3774 . A regular expression is available for excluding directories from the
3777 20. The autoreply transport checks the characters in options that define the
3778 message's headers; it allows continued headers, but it was checking with
3779 isspace() after an embedded newline instead of explicitly looking for a
3782 21. If all the "regular" hosts to which an address was routed had passed their
3783 expiry times, and had not reached their retry times, the address was
3784 bounced, even if fallback hosts were defined. Now Exim should go on to try
3787 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3788 equivalent code in the SMTP transport. Some hosts send humungous responses
3789 to HELO/EHLO, more than 1024 it seems.
3791 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3792 gives compiler warnings in some environments. I've now done it "properly",
3795 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3796 (because of problems with the built-in one) was declared to return uschar *
3797 instead of char *, causing compiler failure.
3799 25. Fixed a file descriptor leak when processing alias/forward files.
3801 26. Fixed a minor format string issue in dbfn.c.
3803 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3805 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3806 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3809 29. When a sender address is verified, it is cached, to save repeating the test
3810 when there is more than one recipient in a message. However, when the
3811 verification involves a callout, it is possible for different callout
3812 options to be set for different recipients. It is too complicated to keep
3813 track of this in the cache, so now Exim always runs a verification when a
3814 callout is required, relying on the callout cache for the optimization.
3815 The overhead is duplication of the address routing, but this should not be
3818 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3819 to be verified with callout=postmaster, and the main callout worked but the
3820 postmaster check failed, the verification correctly failed. However, if a
3821 subsequent RCPT command asked for sender verification *without* the
3822 postmaster check, incorrect caching caused this verification also to fail,
3825 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3826 it was not caching the DNS options (qualify_single, search_parents) that
3827 were used when the lookup failed. A subsequent lookup with different
3828 options therefore always gave the same answer, though there were cases
3829 where it should not have. (Example: a "domains = !$mx_any" option on a
3830 dnslookup router: the "domains" option is always processed without any
3831 widening, but the router might have qualify_single set.) Now Exim uses the
3832 cached value only when the same options are set.
3834 32. Added John Jetmore's "exipick" utility to the distribution.
3836 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3837 than a timeout (e.g. a certificate is required, and is not provided), an
3838 Exim server now closes the connection immediately. Previously it waited for
3839 the client to close - but if the client is SSL, it seems that they each
3840 wait for each other, leading to a delay before one of them times out.
3842 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3843 maintained 0.8.x compatibility because I don't think many are using it, and
3844 it is clearly obsolete.
3846 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3849 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3850 client certificate was expired. A simple patch fixes this, though I don't
3851 understand the full logic of why the verify callback is called multiple
3854 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3855 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3856 which causes problems with some clients (such as the Certicom SSL Plus
3857 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3858 disables the coutermeasure allowing Eudora to connect."
3860 38. Exim was not checking that a write() to a log file succeeded. This could
3861 lead to Bad Things if a log got too big, in particular if it hit a file
3862 size limit. Exim now panics and dies if it cannot write to a log file, just
3863 as it does if it cannot open a log file.
3865 39. Modified OS/Makefile-Linux so that it now contains
3867 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3869 The two -D definitions ensure that Exim is compiled with large file
3870 support, which makes it possible to handle log files that are bigger than
3873 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3874 instance) a domain was checked against a named list that involved a lookup,
3875 causing $domain_data to be set, then another domain was checked against the
3876 same list, then the first domain was re-checked, the value of $domain_data
3877 after the final check could be wrong. In particular, if the second check
3878 failed, it could be set empty. This bug probably also applied to
3881 41. The strip_trailing_dot option was not being applied to the address given
3882 with the -f command-line option.
3884 42. The code for reading a message's header from the spool was incrementing
3885 $received_count, but never initializing it. This meant that the value was
3886 incorrect (doubled) while delivering a message in the same process in which
3887 it was received. In the most common configuration of Exim, this never
3888 happens - a fresh exec is done - but it can happen when
3889 deliver_drop_privilege is set.
3891 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3892 it now includes up to 150 characters of the unexpected data in the log
3895 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3896 and building data strings. The size of both of these buffers was 10 000
3897 bytes - far larger than anybody would *ever* want, thought I. Needless to
3898 say, somebody hit the limit. I have increased the maximum line length to
3899 20 000 and the maximum data length of concatenated lines to 100 000. I have
3900 also fixed two bugs, because there was no checking on these buffers. Tsk,
3901 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3902 buffer is too small.
3904 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3905 lsearch lookups. Now it does.
3907 46. When parsing a route_list item in a manualroute router, a fixed-length
3908 buffer was used for the list of hosts. I made this 1024 bytes long,
3909 thinking that nobody would ever have a list of hosts that long. Wrong.
3910 Somebody had a whole pile of complicated expansion conditions, and the
3911 string was silently truncated, leading to an expansion error. It turns out
3912 that it is easier to change to an unlimited length (owing to other changes
3913 that have happened since this code was originally written) than to build
3914 structure for giving a limitation error. The length of the item that
3915 expands into the list of hosts is now unlimited.
3917 47. The lsearch lookup could not handle data where the length of text line was
3918 more than 4095 characters. Such lines were truncated, leading to shortened
3919 data being returned. It should now handle lines of any length.
3921 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3922 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3925 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3926 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3927 no longer happens in comment lines. A list of replacements is now placed
3928 at the head of all of the source files, except those whose only change is
3929 to replace PERL_COMMAND in the very first #! line.
3931 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3932 messages on the queue, with a bottom-up merge sort, using code contributed
3933 by Michael Haardt. This should make operations like -bp somewhat faster on
3934 large queues. It won't affect queue runners, except when queue_run_in_order
3937 51. Installed eximstats 1.31 in the distribution.
3939 52. Added support for SRV lookups to the dnslookup router.
3941 53. If an ACL referred to $message_body or $message_body_end, the value was not
3942 reset for any messages that followed in the same SMTP session.
3944 54. The store-handling optimization for building very long strings was not
3945 differentiating between the different store pools. I don't think this
3946 actually made any difference in practice, but I've tidied it.
3948 55. While running the routers to verify a sender address, $sender_address
3949 was still set to the sender address. This is wrong, because when routing to
3950 send a bounce to the sender, it would be empty. Therefore, I have changed
3951 it so that, while verifying a sender address, $sender_address is set to <>.
3952 (There is no change to what happens when verifying a recipient address.)
3954 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3955 target A or AAAA records (if not already returned) without resetting the
3956 qualify_single or search_parents options of the DNS resolver. These are
3957 inappropriate in this case because the targets of MX and SRV records must
3958 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3959 target that, when qualified, matched something in the local domain. These
3960 two options are now turned off when doing these lookups.
3962 57. It seems that at least some releases of Reiserfs (which does not have the
3963 concept of a fixed number of inodes) returns zero and not -1 for the
3964 number of available inodes. This interacted badly with check_spool_inodes,
3965 which assumed that -1 was the "no such thing" setting. What I have done is
3966 to check that the total number of inodes is greater than zero before doing
3967 the test of how many are available.
3969 58. When a "warn" ACL statement has a log_message modifier, the message is
3970 remembered, and not repeated. This is to avoid a lot of repetition when a
3971 message has many recipients that cause the same warning to be written.
3972 Howewer, Exim was preserving the list of already written lines for an
3973 entire SMTP session, which doesn't seem right. The memory is now reset if a
3974 new message is started.
3976 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3977 debugging output unless log_rewrite was also set.
3979 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3980 of (int)(handle) when we know that handle contains (void *)(-1).
3982 61. The Exim daemon panic-logs an error return when it closes the incoming
3983 connection. However "connection reset by peer" seems to be common, and
3984 isn't really an error worthy of noting specially, so that particular error
3987 62. When Exim is trying to find all the local interfaces, it used to panic and
3988 die if the ioctl to get the interface flags failed. However, it seems that
3989 on at least one OS (Solaris 9) it is possible to have an interface that is
3990 included in the list of interfaces, but for which you get a failure error
3991 for this call. This happens when the interface is not "plumbed" into a
3992 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3993 failure of the "get flags" call assumes that the interface is down.
3995 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3996 makes life easier for people who are doing arithmetic on fields extracted
3997 from dates, where you often get leading zeros that should not be
3998 interpreted as octal.
4000 64. Added qualify_domain to the redirect router, to override the global
4003 65. If a pathologically long header line contained very many addresses (the
4004 report of this problem mentioned 10 000) and each of them was rewritten,
4005 Exim could use up a very large amount of memory. (It kept on making new
4006 copies of the header line as it rewrote, and never released the old ones.)
4007 At the expense of a bit more processing, the header rewriting function has
4008 been changed so that it no longer eats memory in this way.
4010 66. The generation of the Received: header has been moved from the time that a
4011 message starts to be received, to the time that it finishes. The timestamp
4012 in the Received: header should now be very close to that of the <= log
4013 line. There are two side-effects of this change:
4015 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4016 logged header lines no longer include the local Received: line, because
4017 it has not yet been created. The same applies to a copy of the message
4018 that is returned to a non-SMTP sender when a message is rejected.
4020 (b) When a filter file is tested using -bf, no additional Received: header
4021 is added to the test message. After some thought, I decided that this
4024 This change does not affect the value of $received_for. It is still set
4025 after address rewriting, but before local_scan() is called.
4027 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4029 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4030 gave an unhelpful panic error message, and a defer error. I have managed to
4031 change this behaviour so that it now rejects any supplied certificate,
4032 which seems right, as the list of acceptable certificates is empty.
4034 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4035 gave an unhelpful defer error. I have not managed to make this reject any
4036 supplied certificates, but the error message it gives is "no certificate
4037 supplied", which is not helpful.
4039 70. exigrep's output now also includes lines that are not associated with any
4040 message, but which match the given pattern. Implemented by a patch from
4041 Martin Sluka, which also tidied up the Perl a bit.
4043 71. Recipient callout verification, like sender verification, was using <> in
4044 the MAIL FROM command. This isn't really the right thing, since the actual
4045 sender may affect whether the remote host accepts the recipient or not. I
4046 have changed it to use the actual sender in the callout; this means that
4047 the cache record is now keyed on a recipient/sender pair, not just the
4048 recipient address. There doesn't seem to be a real danger of callout loops,
4049 since a callout by the remote host to check the sender would use <>.
4050 [SEE ABOVE: changed after hitting problems.]
4052 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4053 temporary errors. However, in the case of such a code being given after
4054 the end of a data transmission (i.e. after ".") Exim was failing to write
4055 a retry record for the message. (Yes, there was some broken host that was
4056 actually sending 8xx at this point.)
4058 73. An unknown lookup type in a host list could cause Exim to panic-die when
4059 the list was checked. (An example that provoked this was putting <; in the
4060 middle of a list instead of at the start.) If this happened during a DATA
4061 ACL check, a -D file could be left lying around. This kind of configuration
4062 error no longer causes Exim to die; instead it causes a defer errror. The
4063 incident is still logged to the main and panic logs.
4065 74. Buglet left over from Exim 3 conversion. The message "too many messages
4066 in one connection" was written to the rejectlog but not the mainlog, except
4067 when address rewriting (yes!) was being logged.
4069 75. Added write_rejectlog option.
4071 76. When a system filter was run not as root (that is, when system_filter_user
4072 was set), the values of the $n variables were not being returned to the
4073 main process; thus, they were not subsequently available in the $sn
4076 77. Added +return_path_on_delivery log selector.
4078 78. A connection timeout was being treated differently from recipients deferred
4079 when testing hosts_max_try with a message that was older than the host's
4080 retry timeout. (The host should not be counted, thus allowing all hosts to
4081 be tried at least once before bouncing.) This may have been the cause of an
4082 occasionally reported bug whereby a message would remain on the queue
4083 longer than the retry timeout, but would be bounced if a delivery was
4084 forced. I say "may" because I never totally pinned down the problem;
4085 setting up timeout/retry tests is difficult. See also the next item.
4087 79. The ultimate address timeout was not being applied to errors that involved
4088 a combination of host plus message (for example, a timeout on a MAIL
4089 command). When an address resolved to a number of possible hosts, and they
4090 were not all tried for each delivery (e.g. because of hosts_max_try), a
4091 message could remain on the queue longer than the retry timeout.
4093 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4096 81. Fixed an obscure SMTP outgoing bug which required at least the following
4097 conditions: (a) there was another message waiting for the same server;
4098 (b) the server returned 5xx to all RCPT commands in the first message so
4099 that the message was not completed; (c) the server dropped the connection
4100 or gave a negative response to the RSET that Exim sends to abort the
4101 transaction. The observed case was a dropped connection after DATA that had
4102 been sent in pipelining mode. That is, the server had advertised PIPELINING
4103 but was not implementing it correctly. The effect of the bug was incorrect
4104 behaviour, such as trying another host, and this could lead to a crash.
4110 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4111 and daemon.c were passed as pointers to ints; they should have been
4112 pointers to socklen_t variables (which are typically unsigned ints).
4114 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4117 3. Fixed a really odd bug that affected only the testing scheme; patching a
4118 certain fixed string in the binary changed the value of another string that
4119 happened to be identical to the end of the original first string.
4121 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4122 name", it returns that address as the IP address. On some operating
4123 systems (e.g. Solaris), it also passes back the IP address string as the
4124 "host name". However, on others (e.g. Linux), it passes back an empty
4125 string. Exim wasn't checking for this, and was changing the host name to an
4126 empty string, assuming it had been canonicized.
4128 5. Although rare, it is permitted to have more than one PTR record for a given
4129 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4130 all the names associated with an address, because they do in Solaris.
4131 However, it seems that they do not in Linux for data that comes from the
4132 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4133 I found this out when I moved to a new Linux workstation and tried to run
4134 the Exim test suite.
4136 To get round this problem I have changed the code so that it now does its
4137 own call to the DNS to look up PTR records when searching for a host name.
4138 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4139 addresses that are only in /etc/hosts are still found.
4141 This behaviour is, however, controlled by an option called host_lookup_
4142 order, which defaults to "bydns:byaddr". If people want to use the other
4143 order, or indeed, just use one or the other means of lookup, they can
4144 specify it in this variable.
4146 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4147 some operating systems, this comes back from gethostbyaddr() as an empty
4148 string, and this is what Exim used to test for. However, it seems that in
4149 other systems, "." is yielded. Exim now tests for this case too.
4151 7. The values of check_spool_space and check_log_space are now held internally
4152 as a number of kilobytes instead of an absolute number of bytes. If a
4153 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4154 kilobyte. This means that much larger values can be stored.
4156 8. Exim monitor: an attempt to get the action menu when not actually pointing
4157 at a message produces an empty menu entitled "No message selected". This
4158 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4159 no entries in it ("Shell widget menu has zero width and/or height"). So I
4160 have added a single, blank menu entry in this case.
4162 9. Added ${quote_local_part.
4164 10. MIME decoding is now applied to the contents of Subject: header lines when
4167 11. Now that a reference to $sender_host_address automatically causes a reverse
4168 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4169 host lookup before query-style lookups in lists that might use this
4170 variable. This has therefore been abolished, and the "net-" prefix is no
4171 longer necessary for query-style lookups.
4173 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4174 have been a typo for LFLAGS, so it has been changed.
4176 13. The install script calls Exim with "-C /dev/null" in order to find the
4177 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4178 to be output. Howeve, since Exim outputs its version number before the
4179 error, it didn't break the script. It just looked ugly. I fixed this by
4180 always allowing "-C /dev/null" if the caller is root.
4182 14. Ignore overlarge ACL variable number when reading spool file - insurance
4183 against a later release with more variables having written the file.
4185 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4186 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4187 this, as well as the form without the "IPv6" on the front.
4189 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4190 OS/Makefile-Darwin file.
4192 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4193 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4195 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4196 variable. However, some OS use ints for inodes. I've added cast to long int
4197 to get rid of the compiler warning.
4199 19. I had forgotten to lock out "/../" in configuration file names when
4200 ALT_CONFIG_PREFIX was set.
4202 20. Routers used for verification do not need to specify transports. However,
4203 if such a router generated a host list, and callout was configured, Exim
4204 crashed, because it could not find a port number from the (non-existent)
4205 transport. It now assumes port 25 in this circumstance.
4207 21. Added the -t option to exigrep.
4209 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4210 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4211 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4212 them is set, LOOKUP_LSEARCH is forced.
4214 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4216 24. Added sender and host information to the "rejected by local_scan()" log
4217 line; previously there was no indication of these.
4219 25. Added .include_if_exists.
4221 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4222 Linux. It turns out that not all file systems support this. Apparently some
4223 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4224 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4225 not supported on the file descriptor, is now ignored when Exim is trying to
4226 sync a directory. This applies only to Linux.
4228 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4230 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4231 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4234 29. Added support for maildirsize files from supplied patch (modified a bit).
4236 30. The use of :fail: followed by an empty string could lead Exim to respond to
4237 sender verification failures with (e.g.):
4239 550 Verification failed for <xxx>
4240 550 Sender verify failed
4242 where the first response line was missing the '-' that indicates it is not
4243 the final line of the response.
4245 31. The loop for finding the name of the user that called Exim had a hardwired
4246 limit of 10; it now uses the value of finduser_retries, which is used for
4247 all other user lookups.
4249 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4252 33. Exim was neglecting to zero errno before one call of strtol() when
4253 expanding a string and expecting an integer value. On some systems this
4254 resulted in spurious "integer overflow" errors. Also, it was casting the
4255 result into an int without checking.
4257 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4258 did not work. The code looks as if it has *never* worked, though it appears
4259 to have been documented since at least releast 1.62. I have made it work.
4261 35. The "timeout_DNS" error in retry rules, also documented since at least
4262 1.62, also never worked. As it isn't clear exactly what this means, and
4263 clearly it isn't a major issue, I have abolished the feature by treating it
4264 as "timeout", and writing a warning to the main and panic logs.
4266 36. The display of retry rules for -brt wasn't always showing the error code
4269 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4270 timeout_connect_A, timeout_connect_MX.
4272 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4273 to the empty sender.
4275 39. The daemon was not analysing the content of -oX till after it had closed
4276 stderr and disconnected from the controlling terminal. This meant that any
4277 syntax errors were only noted on the panic log, and the return code from
4278 the command was 0. By re-arranging the code a little, I've made the
4279 decoding happen first, so such errors now appear on stderr, and the return
4280 code is 1. However, the actual setting up of the sockets still happens in
4281 the disconnected process, so errors there are still only recorded on the
4284 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4285 connections (as happens on some IP stacks) was logged at start up time as
4286 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4287 it from "IPv6 and IPv4", which means that two separate sockets are being
4290 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4291 says whether AF_INET or AF_INET6 was passed as an argument.
4293 42. Exiwhat output was messed up when time zones were included in log
4296 43. Exiwhat now gives more information about the daemon's listening ports,
4297 and whether -tls-on-connect was used.
4299 44. The "port" option of the smtp transport is now expanded.
4301 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4302 silently ignored. Now an error message is written to the main and panic
4305 46. There's a new ACL modifier called "logwrite" which writes to a log file
4306 as soon as it is encountered.
4308 47. Added $local_user_uid and $local_user_gid at routing time.
4310 48. Exim crashed when trying to verify a sender address that was being
4313 49. Exim was recognizing only a space character after ".include". It now also
4314 recognizes a tab character.
4316 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4317 extracting the relevant information from the specification. The man page no
4318 longer contains scrambled data for the -d option, and I've added a section
4319 at the front about calling Exim under different names.
4321 51. Added "extra_headers" argument to the "mail" command in filter files.
4323 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4326 53. Installed eximstats 1.29.
4328 54. Added transport_filter_timeout as a generic transport option.
4330 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4331 Cc: header lines. This was required by RFC 822, but it not required by RFC
4334 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4335 incoming messages that did not have them. Now it does so only if the
4336 message originates locally, that is, if there is no associated remote host
4337 address. When Resent- header lines are present, this applies to the Resent-
4338 lines rather than the non-Resent- lines.
4340 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4341 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4343 58. Messages for configuration errors now include the name of the main
4344 configuration file - useful now that there may be more than one file in a
4345 list (.included file names were always shown).
4347 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4348 for those rare installations that do not start the daemon as root or run it
4349 setuid root. I've cut out the call to initgroups() if the daemon is not
4352 60. The Exim user and group can now be bound into the binary as text strings
4353 that are looked up at the start of Exim's processing.
4355 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4357 62. Added $mailstore_basename variable.
4359 63. Installed patch to sieve.c from Michael Haardt.
4361 64. When Exim failed to open the panic log after failing to open the main log,
4362 the original message it was trying to log was written to stderr and debug
4363 output, but if they were not available (the usual case in production), it
4364 was lost. Now it is written to syslog before the two lines that record the
4365 failures to open the logs.
4367 65. Users' Exim filters run in subprocesses under the user's uid. It is
4368 possible for a "deliver" command or an alias in a "personal" command to
4369 provoke an address rewrite. If logging of address rewriting is configured,
4370 this fails because the process is not running as root or exim. There may be
4371 a better way of dealing with this, but for the moment (because 4.30 needs
4372 to be released), I have disabled address rewrite logging when running a
4373 filter in a non-root, non-exim process.
4379 1. The buildconfig auxiliary program wasn't quoting the value set for
4380 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4381 not defined. This bug was masked in 4.22 by the effect that was fixed in
4384 2. Some messages that were rejected after a message id was allocated were
4385 shown as "incomplete" by exigrep. It no longer does this for messages that
4386 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4388 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4389 have allow_domain_literals set, the ID did not get logged in the <= line.
4390 Domain literals are now always recognized in Message-ID: header lines.
4392 4. The first argument for a ${extract expansion item is the key name or field
4393 number. Leading and trailing spaces in this item were not being ignored,
4394 causing some misleading effects.
4396 5. When deliver_drop_privilege was set, single queue runner processes started
4397 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4398 same command in the process it spins off) were not dropping privilege.
4400 6. When the daemon running as "exim" started a queue runner, it always
4401 re-executed Exim in the spun-off process. This is a waste of effort when
4402 deliver_drop_privilege is set. The new process now just calls the
4403 queue-runner function directly.
4409 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4412 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4413 ignored. Though the use of -oP was forcing the writing of a pid file, it
4414 was always written to the default place.
4416 3. If the message "no IP address found for host xxxx" is generated during
4417 incoming verification, it is now followed by identification of the incoming
4418 connection (so you can more easily find what provoked it).
4420 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4422 5. Added some features to "harden" Exim a bit more against certain attacks:
4424 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4425 be put in Local/Makefile. This is like the never_users runtime option,
4426 but it cannot be overridden. The default setting is "root".
4428 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4429 prefix string with which any file named in a -C command line option
4432 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4433 is retained for -C and -D only if the caller of Exim is root. Without
4434 it, the exim user may also use -C and -D and retain privilege.
4436 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4437 command line option is disabled.
4439 6. Macro names set by the -D option must start with an upper case letter, just
4440 like macro names defined in the configuration file.
4442 7. Added "dereference=" facility to LDAP.
4444 8. Two instances of the typo "uknown" in the source files are fixed.
4446 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4447 the Configure-Makefile script screwed up while processing it.
4449 10. Incorporated PCRE 4.4.
4451 11. The SMTP synchronization check was not operating right at the start of an
4452 SMTP session. For example, it could not catch a HELO sent before the client
4453 waited for the greeting. There is now a check for outstanding input at the
4454 point when the greeting is written. Because of the duplex, asynchronous
4455 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4456 way, but not yet received, when the check is performed.
4458 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4459 on TCP/IP sockets, because this apparently causes some broken clients to
4462 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4463 unchanged) from the Cygwin maintainer.
4465 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4466 was supported instead of testing for mbx. Effectively a typo.
4468 15. The spa authenticator server code was not checking that the input it
4469 received was valid base64.
4471 16. The debug output line for the "set" modifier in ACLs was not showing the
4472 name of the variable that was being set.
4474 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4476 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4477 lookup on its own. Something else was needed to trigger the lookup. For
4478 example, a match in host_lookup or the need for a host name in a host list.
4479 Now, if $sender_host_name is referenced and the host name has not yet been
4480 looked up, a lookup is performed. If the lookup fails, the variable remains
4481 empty, and $host_lookup_failed is set to "1".
4483 19. Added "eqi" as a case-independent comparison operator.
4485 20. The saslauthd authentication condition could segfault if neither service
4486 nor realm was specified.
4488 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4489 error message that was logged was misleading, and incoming SMTP
4490 connections were dropped. The message is now more accurate, and temporary
4491 errors are given to SMTP connections.
4493 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4494 (except RSET) with a 421 error, until QUIT is received. However, it was
4495 failing to send a response to QUIT.
4497 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4498 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4499 ACL is now run after the helo_try_verify_hosts code.
4501 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4502 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4503 case-independent, so other case variants are also recognized). Apparently
4504 some systems use these upper case variants.
4506 25. If more than two messages were waiting for the same host, and a transport
4507 filter was specified for the transport, Exim sent two messages over the
4508 same TCP/IP connection, and then failed with "socket operation on non-
4509 socket" when it tried to send the third.
4511 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4513 27. The extern definition of crypt16() in expand.c was not being excluded when
4514 the OS had its own crypt16() function.
4516 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4517 as a preferred synonym for return_size_limit, both as an option and as an
4520 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4522 30. Changed the default configuration ACL to relax the local part checking rule
4523 for addresses that are not in any local domains. For these addresses,
4524 slashes and pipe symbols are allowed within local parts, but the sequence
4525 /../ is explicitly forbidden.
4527 31. SPA server authentication was not clearing the challenge buffer before
4530 32. log_message in a "warn" ACL statement was writing to the reject log as
4531 well as to the main log, which contradicts the documentation and doesn't
4532 seem right (because no rejection is happening). So I have stopped it.
4534 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4535 However, I am unable to do any testing of this.
4537 34. Fixed an infelicity in the appendfile transport. When checking directories
4538 for a mailbox, to see if any needed to be created, it was accidentally
4539 using path names with one or more superfluous leading slashes; tracing
4540 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4542 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4543 contents are added to the log line that is written for every discarded
4544 recipient. (Previously a log_message setting was ignored.)
4546 36. The ${quote: operator now quotes the string if it is empty.
4548 37. The install script runs exim in order to find its version number. If for
4549 some reason other than non-existence or emptiness, which it checks, it
4550 could not run './exim', it was installing it with an empty version number,
4551 i.e. as "exim-". This error state is now caught, and the installation is
4554 38. An argument was missing from the function that creates an error message
4555 when Exim fails to connect to the socket for saslauthd authentication.
4556 This could cause Exim to crash, or give a corrupted message.
4558 39. Added isip, isip4, and isip6 to ${if conditions.
4560 40. The ACL variables $acl_xx are now saved with the message, and can be
4561 accessed later in routers, transports, and filters.
4563 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4564 strings in the file are not string-expanded.
4566 42. If a MAIL command specified a SIZE value that was too large to fit into an
4567 int variable, the check against message_size_limit failed. Such values are
4568 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4569 day this will have to be increased, but I don't think I want to be around
4570 when emails are that large.
4577 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4578 iconv() is not standard in FreeBSD.
4580 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4581 IPv6 enabled. The observed symptom was a segmentation fault on return from
4582 the function os_common_find_running_interfaces() in src/os.c.
4584 3. In the check_special_case() function in daemon.c I had used "errno" as an
4585 argument name, which causes warnings on some systems. This was basically a
4586 typo, since it was named "eno" in the comments!
4588 4. The code that waits for the clock to tick (at a resolution of some fraction
4589 of a second) so as to ensure message-id uniqueness was always waiting for
4590 at least one whole tick, when it could have waited for less. [This is
4591 almost certainly not relevant at current processor speeds, where it is
4592 unlikely to ever wait at all. But we try to future-proof.]
4594 5. The function that sleeps for a time interval that includes fractions of a
4595 second contained a race. It did not block SIGALRM between setting the
4596 timer, and suspending (a couple of lines later). If the interval was short
4597 and the sigsuspend() was delayed until after it had expired, the suspension
4598 never ended. On busy systems this could lead to processes getting stuck for
4601 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4602 process, before it forks any delivery processes. The open lookup caching
4603 mechanism meant that the open file or database connection was passed into
4604 the delivery process. The problem was that delivery processes always tidy
4605 up cached lookup data. This could cause a problem for the next delivery
4606 process started by the queue runner, because the external queue runner
4607 process does not know about the closure. So the next delivery process
4608 still has data in the lookup cache. In the case of a file lookup, there was
4609 no problem because closing a file descriptor in a subprocess doesn't affect
4610 the parent. However, if the lookup was caching a connection to a database,
4611 the connection was closed, and the second delivery process was likely to
4612 see errors such as "PGSQL: query failed: server closed the connection
4613 unexpectedly". The problem has been fixed by closing all cached lookups
4614 in a queue runner before running a delivery process.
4616 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4617 seem to have the "const" qualifier which it has on other OS. I've
4620 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4621 *of the same type* (client or server) with the same public name that an
4622 error should be diagnosed.
4624 9. When Exim looked up a host name for an IP address, but failed to find the
4625 original IP address when looking up the host name (a safety check), it
4626 output the message "<ip address> does not match any IP for NULL", which was
4627 confusing, to say the least. The bug was that the host name should have
4628 appeared instead of "NULL".
4630 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4631 user that is built into the binary, and the -C or -D options is used, root
4632 privilege is dropped before the configuration file is read. In addition,
4633 logging is switched to stderr instead of the normal log files. If the
4634 configuration then re-defines the Exim user, the unprivileged environment
4635 is probably not what is expected, so Exim logs a panic warning message (but
4638 However, if deliver_drop_privilege is set, the unprivileged state may well
4639 be exactly what is intended, so the warning has been cut out in that case,
4640 and Exim is allowed to try to write to its normal log files.
4646 1. smtp_return_error_details was not giving details for temporary sender
4647 or receiver verification errors.
4649 2. Diagnose a configuration error if two authenticators have the same public
4652 3. Exim used not to create the message log file for a message until the first
4653 delivery attempt. This could be confusing when incoming messages were held
4654 for policy or load reasons. The message log file is now created at the time
4655 the message is received, and an initial "Received" line is written to it.
4657 4. The automatically generated man page for command line options had a minor
4658 bug that caused no ill effects; however, a more serious problem was that
4659 the procedure for building the man page automatically didn't always
4660 operate. Consequently, release 4.20 contains an out-of-date version. This
4661 shouldn't happen again.
4663 5. When building Exim with embedded Perl support, the script that builds the
4664 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4665 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4667 6. The freeze_tell option was not being used for messages that were frozen on
4668 arrival, either by an ACL or by local_scan().
4670 7. Added the smtp_incomplete_transaction log selector.
4672 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4673 was accepting AUTH without a new EHLO.
4675 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4676 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4679 10. Logging of TCP/IP connections (when configured) now happens in the main
4680 daemon process instead of the child process, so that the TCP/IP connection
4681 count is more accurate (but it can never be perfect).
4683 11. The use of "drop" in a nested ACL was not being handled correctly in the
4684 outer ACL. Now, if condition failure induced by the nested "drop" causes
4685 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4686 or "require"), the connection is dropped.
4688 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4689 that yield "discard" can now be used with an "accept" or a "discard" verb,
4690 but an error is generated for any others (because I can't see a useful way
4691 to define what should happen).
4693 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4694 are now processed in the same way as lines in the Exim configuration file.
4695 In particular, continuation lines are supported.
4697 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4699 15. Added -ti meaning -t -i.
4701 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4702 domains, and warn by logging if others are found.
4704 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4705 returned from GIFCONF when Exim is trying to find the list of interfaces on
4706 a host. The code in os.c has been modified to copy each ifreq to an aligned
4707 structure in all cases.
4709 Also, in some cases, the returned ifreq's were being copied to a 'struct
4710 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4711 means the last couple of bytes of an IPv6 address could be chopped if the
4712 ifreq contained only a normal sockaddr (14 bytes storage).
4714 18. Named domain lists were not supported in the hosts_treat_as_local option.
4715 An entry such as +xxxx was not recognized, and was treated as a literal
4718 19. Ensure that header lines added by a DATA ACL are included in the reject log
4719 if the ACL subsequently rejects the message.
4721 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4722 MD5 (which is deprecated).
4724 21. When testing a filter file using -bf, Exim was writing a message when it
4725 took the sender from a "From " line in the message, but it was not doing so
4726 when it took $return_path from a Return-Path: header line. It now does.
4728 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4729 with a valid header line field name (a series of printing characters
4730 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4732 23. Changed "disc" in the source to "disk" to conform to the documentation and
4733 the book and for uniformity.
4735 24. Ignore Sendmail's -Ooption=value command line item.
4737 25. When execve() failed while trying to run a command in a pipe transport,
4738 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4739 could be confused with a return value of 69 from the command itself. This
4740 has been changed to 127, the value the shell returns if it is asked to run
4741 a non-existent command. The wording for the related log line suggests a
4742 non-existent command as the problem.
4744 26. If received_header_text expands to an empty string, do not add a Received:
4745 header line to the message. (Well, it adds a token one on the spool, but
4746 marks it "old" so that it doesn't get used or transmitted.)
4748 27. Installed eximstats 1.28 (addition of -nt option).
4750 28. There was no check for failure on the call to getsockname() in the daemon
4751 code. This can fail if there is a shortage of resources on the system, with
4752 ENOMEM, for example. A temporary error is now given on failure.
4754 29. Contrary to the C standard, it seems that in some environments, the
4755 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4756 program. Exim now does this explicitly; it affects the formatting of
4757 timestamps using strftime().
4759 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4760 complaints. I've now initialized all the variables, to avoid this.
4762 32. Header lines added by a system filter were not being "seen" during
4763 transport-time rewrites.
4765 33. The info_callback() function passed to OpenSSL is set up with type void
4766 (*)(SSL *, int, int), as described somewhere. However, when calling the
4767 function (actually a macro) that sets it up, the type void(*)() is
4768 expected. I've put in a cast to prevent warnings from picky compilers.
4770 34. If a DNS black list lookup found a CNAME record, but there were no A
4771 records associated with the domain it pointed at, Exim crashed.
4773 35. If a DNS black list lookup returned more than one A record, Exim ignored
4774 all but the first. It now scans all returned addresses if a particular IP
4775 value is being sought. In this situation, the contents of the
4776 $dnslist_value variable are a list of all the addresses, separated by a
4779 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4780 to accept a host name and all its aliases if the forward lookup for any of
4781 them yielded the IP address of the incoming connection. Now it accepts only
4782 those names whose forward lookup yields the correct IP address. Any other
4783 names are discarded. This closes a loophole whereby a rogue DNS
4784 administrator could create reverse DNS records to break through a
4785 wildcarded host restriction in an ACL.
4787 37. If a user filter or a system filter that ran in a subprocess used any of
4788 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4789 the wrong values were passed to the pipe command ($thisaddress had the
4790 value of $0, $0 had the value of $1, etc). This bug was introduced by
4791 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4793 38. Improved the line breaking for long SMTP error messages from ACLs.
4794 Previously, if there was no break point between 40 and 75 characters, Exim
4795 left the rest of the message alone. Two changes have been made: (a) I've
4796 reduced the minimum length to 35 characters; (b) if it can't find a break
4797 point between 35 and 75 characters, it looks ahead and uses the first one
4798 that it finds. This may give the occasional overlong line, but at least the
4799 remaining text gets split now.
4801 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4802 file descriptors might be low, and that setting 1000 would always raise it.
4803 It turns out that in some environments, the limit is already over 1000 and
4804 that lowering it causes trouble. So now Exim takes care not to decrease it.
4806 40. When delivering a message, the value of $return_path is set to $sender_
4807 address at the start of routing (routers may change the value). By an
4808 oversight, this default was not being set up when an address was tested by
4809 -bt or -bv, which affected the outcome if any router or filter referred to
4812 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4813 log only when "message" or "log_message" are set. However, if one of the
4814 conditions was an address verification, or a call to a nested ACL, the
4815 messages generated by the underlying test were being passed through. This
4816 no longer happens. The underlying message is available in $acl_verify_
4817 message for both "message" and "log_message" expansions, so it can be
4818 passed through if needed.
4820 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4821 new expansion $bh_ to give the encoded byte string without charset
4822 translation. Translation happens only if iconv() is available; HAVE_ICONV
4823 indicates this at build time. HEADERS_CHARSET gives the charset to
4824 translate to; headers_charset can change it in the configuration, and
4825 "headers charset" can change it in an individual filter file.
4827 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4828 that creates RFC 2047 encoded "words" labels them as that charset instead
4829 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4830 expansion operator; (ii) when Exim creates a From: line for a local
4831 message; (iii) when a header line is rewritten to include a "phrase" part.
4833 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4834 buggy, causing it to skip the first lines of messages whose message ID
4835 ended in 'D'. This would not have bitten before Exim release 4.14, because
4836 message IDs were unlikely to end in 'D' before then. The effect was to have
4837 incorrect size information for certain domains.
4839 45. #include "config.h" was missing at the start of the crypt16.c module. This
4840 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4843 46. If there was a timeout during a "random" callout check, Exim treated it as
4844 a failure of the random address, and carried on sending RSET and the real
4845 address. If the delay was just some slowness somewhere, the response to the
4846 original RCPT would be taken as a response to RSET and so on, causing
4847 mayhem of various kinds.
4849 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4850 when I implemented it. It didn't allow for the fact that some option values
4851 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4852 the right test for positive values.
4854 48. Domain names in DNS records are case-independent. Exim always looks them up
4855 in lower case. Some resolvers return domain names in exactly the case they
4856 appear in the zone file, that is, they may contain uppercase letters. Not
4857 all resolvers do this - some return always lower case. Exim was treating a
4858 change of case by a resolver as a change of domain, similar to a widening
4859 of a domain abbreviation. This triggered its re-routing code and so it was
4860 trying to route what was effectively the same domain again. This normally
4861 caused routing to fail (because the router wouldn't handle the domain
4862 twice). Now Exim checks for this case specially, and just changes the
4863 casing of the domain that it ultimately uses when it transmits the message
4866 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4869 50. If a filter generated a file delivery with a non-absolute name (possible if
4870 no home directory exists for the router), the forbid_file option was not
4873 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4874 the existing equality matching.
4876 52. Exim was using ints instead of ino_t variables in some places where it was
4877 dealing with inode numbers.
4879 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4880 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4881 called TMPDIR, and if it finds it is different, it changes its value.
4883 54. The smtp_printf() function is now made available to local_scan() so
4884 additional output lines can be written before returning. There is also an
4885 smtp_fflush() function to enable the detection of a dropped connection.
4886 The variables smtp_input and smtp_batched_input are exported to
4889 55. Changed the default runtime configuration: the message "Unknown user"
4890 has been removed from the ACL, and instead placed on the localuser router,
4891 using the cannot_route_message feature. This means that any verification
4892 failures that generate their own messages won't get overridden. Similarly,
4893 the "Unrouteable address" message that was in the ACL for unverifiable
4894 relay addresses has also been removed.
4896 56. Added hosts_avoid_esmtp to the smtp transport.
4898 57. The exicyclog script was not checking for the esoteric option
4899 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4900 will work only if exicyclog is run under the appropriate euid.
4902 58. Following a discussion on the list, the rules by which Exim recognises line
4903 endings on incoming messages have been changed. The -dropcr and drop_cr
4904 options are now no-ops, retained only for backwards compatibility. The
4905 following line terminators are recognized: LF CRLF CR. However, special
4906 processing applies to CR:
4908 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4909 nor a local message in the state where . is a terminator.
4911 (ii) If a bare CR is encountered in a header line, an extra space is added
4912 after the line terminator so as not to end the header. The reasoning
4913 behind this is that bare CRs in header lines are most likely either
4914 to be mistakes, or people trying to play silly games.
4916 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4917 was being incorrectly given as 18 bytes larger than it should have been.
4918 This is a VOB (very old bug).
4920 60. This may never have affected anything current, but just in case it has:
4921 When the local host is found other than at the start of a list of hosts,
4922 the local host, those with the same MX, and any that follow, are discarded.
4923 When the list in question was part of a longer list of hosts, the following
4924 hosts (not currently being processed) were also being discarded. This no
4925 longer happens. I'm not sure if this situation could ever has previously
4928 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4931 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4932 header. This is something that is recommended in a new Internet Draft, and
4933 is something that is documented as being done by Sendmail. There are two
4934 possible values. For messages generated by the autoreply transport, Exim
4937 Auto-Submitted: auto-replied
4939 whereas for all other generated messages (e.g. bounces) it adds
4941 Auto-Submitted: auto-generated
4943 63. The "personal" condition in filters now includes a test for the
4944 Auto-Submitted: header. If it contains the string "auto-" the message it
4945 not considered personal.
4947 64. Added rcpt_include_affixes as a generic transport option.
4949 65. Added queue_only_override (default true).
4951 66. Added the syslog_duplication option.
4953 67. If what should have been the first header line of a message consisted of
4954 a space followed by a colon, Exim was mis-interpreting it as a header line.
4955 It isn't of course - it is syntactically invalid and should therefore be
4956 treated as the start of the message body. The misbehaviour could have
4957 caused a number of strange effects, including loss of data in subsequent
4958 header lines, and spool format errors.
4960 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4961 client host had authenticated. This control can now be exercised by an ACL
4962 for more flexibility.
4964 69. By default, callouts do not happen when testing with -bh. There is now a
4965 variant, -bhc, which does actually run the callout code, including
4966 consulting and updating the callout cache.
4968 70. Added support for saslauthd authentication, courtesy of Alexander
4971 71. If statvfs() failed on the spool or log directories while checking their
4972 size for availability, Exim confusingly gave the error "space shortage".
4973 Furthermore, in debugging mode it crashed with a floating point exception.
4974 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4975 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4976 serious problem, Exim now writes to the main and panic logs when this
4977 happens, with details of the failure. It then refuses to accept the
4978 incoming message, giving the message "spool directory problem" or "log
4979 directory problem" with a 421 code for SMTP messages.
4981 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4982 0, 1, and 2 exist, because some OS complain for execs without them (see
4983 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4984 descriptors. However, the code omitted to check that the open succeeded,
4985 causing mysterious errors if for some reason the permissions on /dev/null
4986 got screwed. Now Exim writes a message to the main and panic logs, and
4987 bombs out if it can't open /dev/null.
4989 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4990 interact so that it is all more flexible. It is supposed to remain
4991 backwards compatible. Also added extra_local_interfaces.
4993 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4994 to bomb out with an assertion failure - to the client this appears as a
4995 connection drop. This problem occurs in the part of the code that was taken
4996 from the Samba project. Fortunately, the assertion is in a very simple
4997 function, so I have fixed this by reproducing the function inline in the
4998 one place where it is called, and arranging for authentication to fail
4999 instead of killing the process with assert().
5001 75. The SPA client code was not working when the server requested OEM rather
5002 than Unicode encoding.
5004 76. Added code to make require_files with a specific uid setting more usable in
5005 the case where statting the file as root fails - usually a non-root-mounted
5006 NFS file system. When this happens and the failure is EACCES, Exim now
5007 forks a subprocess and does the per-uid checking as the relevant uid.
5009 77. Added process_log_path.
5011 78. If log_file_path was not explicitly set, a setting of check_log_space or
5012 check_log_inodes was ignored.
5014 79. If a space check for the spool or log partitions fails, the incident is now
5015 logged. Of course, in the latter case the data may get lost...
5017 80. Added the %p formatting code to string_format() so that it can be used to
5018 print addresses in debug_print(). Adjusted all the address printing in the
5019 debugging in store.c to use %p rather than %d.
5021 81. There was a concern that a line of code in smtp_in.c could overflow a
5022 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5023 initially expressed, the concern was not well-founded, because trailing
5024 spaces are removed early. However, if the trailing spaces were followed by
5025 a NULL, they did not get removed, so the overflow was possible. Two fixes
5028 (a) I re-wrote the offending code in a cleaner fashion.
5029 (b) If an incoming SMTP command contains a NULL character, it is rejected
5032 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5033 runs initgroups(), so that if the Exim user is in any additional groups,
5034 they will be used during message reception.
5040 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob