1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.582 2009/10/26 13:18:54 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvement to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
78 acl_smtp_notquit, added index entry.
80 NM/09 Bugzilla 787: Potential buffer overflow in string_format
81 Patch provided by Eugene Bujak
83 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
84 Patch provided by Maxim Dounin
86 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
87 Patch provided by Phil Pennock
89 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
91 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
92 Patch provided by Brad "anomie" Jorsch
94 NM/14 Bugzilla 622: Added timeout setting to transport filter.
95 Patch provided by Dean Brooks
97 TK/05 Add native DKIM support (does not depend on external libraries).
99 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
100 Patch provided by Graeme Fowler
102 NM/16 Bugzilla 851: Documentation example syntax fix.
104 NM/17 Changed NOTICE file to remove references to embedded PCRE.
106 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
108 NM/19 Bugzilla 745: TLS version reporting
109 Patch provided by Phil Pennock
111 NM/20 Bugzilla 167: bool: condition support
112 Patch provided by Phil Pennock
114 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
115 Patch provided by Phil Pennock
117 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
118 Patch provided by Brad "anomie" Jorsch
120 NM/23 Bugzilla 687: Fix misparses in eximstats
121 Patch provided by Heiko Schlittermann
123 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
124 Patch provided by Heiko Schlittermann
126 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
128 NM/26 Bugzilla 799: Documentation correction for ratelimit
130 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
131 Patch provided by David Brownlee
133 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
135 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
137 NM/30 Bugzilla 888: TLS documentation bugfixes
139 NM/31 Bugzilla 889: Dovecot buffer overrun fix
145 TK/01 Add preliminary DKIM support. Currently requires a forked version of
146 ALT-N's libdkim that I have put here:
147 http://duncanthrax.net/exim-experimental/
149 Note to Michael Haardt: I had to rename some vars in sieve.c. They
150 were called 'true' and it seems that C99 defines that as a reserved
151 keyword to be used with 'bool' variable types. That means you could
152 not include C99-style headers which use bools without triggering
153 build errors in sieve.c.
155 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
156 as mailq or other aliases. Changed the --help handling significantly
157 to do whats expected. exim_usage() emits usage/help information.
159 SC/01 Added the -bylocaldomain option to eximstats.
161 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
163 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
165 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
171 PH/01 Another patch from the Sieve maintainer.
173 PH/02 When an IPv6 address is converted to a string for single-key lookup
174 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
175 dots are used instead of colons so that keys in lsearch files need not
176 contain colons. This was done some time before quoting was made available
177 in lsearch files. However, iplsearch files do require colons in IPv6 keys
178 (notated using the quote facility) so as to distinguish them from IPv4
179 keys. This meant that lookups for IP addresses in host lists did not work
180 for iplsearch lookups.
182 This has been fixed by arranging for IPv6 addresses to be expressed with
183 colons if the lookup type is iplsearch. This is not incompatible, because
184 previously such lookups could never work.
186 The situation is now rather anomolous, since one *can* have colons in
187 ordinary lsearch keys. However, making the change in all cases is
188 incompatible and would probably break a number of configurations.
190 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
193 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
194 conversion specification without a maximum field width, thereby enabling
195 a rogue spamd server to cause a buffer overflow. While nobody in their
196 right mind would setup Exim to query an untrusted spamd server, an
197 attacker that gains access to a server running spamd could potentially
198 exploit this vulnerability to run arbitrary code as the Exim user.
200 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
201 $primary_hostname instead of what libspf2 thinks the hosts name is.
203 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
204 a directory entry by the name of the lookup key. Previously, if a
205 symlink pointed to a non-existing file or a file in a directory that
206 Exim lacked permissions to read, a lookup for a key matching that
207 symlink would fail. Now it is enough that a matching directory entry
208 exists, symlink or not. (Bugzilla 503.)
210 PH/03 The body_linecount and body_zerocount variables are now exported in the
213 PH/04 Added the $dnslist_matched variable.
215 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
216 This means they are set thereafter only if the connection becomes
219 PH/06 Added the client_condition to authenticators so that some can be skipped
220 by clients under certain conditions.
222 PH/07 The error message for a badly-placed control=no_multiline_responses left
223 "_responses" off the end of the name.
225 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
227 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
228 (without spaces) instead of just copying the configuration text.
230 PH/10 Added the /noupdate option to the ratelimit ACL condition.
232 PH/11 Added $max_received_linelength.
234 PH/12 Added +ignore_defer and +include_defer to host lists.
236 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
237 way in which PCRE > 7.0 is built.
239 PH/14 Implemented queue_only_load_latch.
241 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
242 MAIL command. The effect was to mangle the value on 64-bit systems.
244 PH/16 Another patch from the Sieve maintainer.
246 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
248 PH/18 If a system quota error occurred while trying to create the file for
249 a maildir delivery, the message "Mailbox is full" was not appended to the
250 bounce if the delivery eventually timed out. Change 4.67/27 below applied
251 only to a quota excession during the actual writing of the file.
253 PH/19 It seems that peer DN values may contain newlines (and other non-printing
254 characters?) which causes problems in log lines. The DN values are now
255 passed through string_printing() before being added to log lines.
257 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
258 and InterBase are left for another time.)
260 PH/21 Added message_body_newlines option.
262 PH/22 Guard against possible overflow in moan_check_errorcopy().
264 PH/23 POSIX allows open() to be a macro; guard against that.
266 PH/24 If the recipient of an error message contained an @ in the local part
267 (suitably quoted, of course), incorrect values were put in $domain and
268 $local_part during the evaluation of errors_copy.
274 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
275 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
278 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
279 issue a MAIL command.
281 PH/02 In an ACL statement such as
283 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
285 if a client was not listed at all, or was listed with a value other than
286 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
287 the condition was not true (as it should be), so access was not denied.
288 The bug was that the ! inversion was incorrectly passed on to the second
289 item. This has been fixed.
291 PH/03 Added additional dnslists conditions == and =& which are different from
292 = and & when the dns lookup returns more than one IP address.
294 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
295 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
297 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
298 FSYNC, which compiles an option called disable_fsync that allows for
299 bypassing fsync(). The documentation is heavily laced with warnings.
301 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
303 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
304 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
305 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
306 including adding "make clean"; (3) Added -fPIC when compiling the test
307 dynamically loaded module, to get rid of a warning.
309 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
310 message fails, move_frozen_messages = true and ignore_bounce_errors_after
311 = 0s. The bug is otherwise harmless.
313 PH/07 There was a bug in the dovecot authenticator such that the value of
314 $auth1 could be overwritten, and so not correctly preserved, after a
315 successful authentication. This usually meant that the value preserved by
316 the server_setid option was incorrect.
318 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
320 PH/09 Installed PCRE release 7.0.
322 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
323 run for batched SMTP input. It is now run at the start of every message
324 in the batch. While fixing this I discovered that the process information
325 (output by running exiwhat) was not always getting set for -bs and -bS
326 input. This is fixed, and it now also says "batched" for BSMTP.
328 PH/11 Added control=no_pipelining.
330 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
331 patch, slightly modified), and move the expansion of helo_data till after
332 the connection is made in the smtp transport (so it can use these
335 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
337 PH/14 Added log_selector = +pid.
339 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
341 PH/16 Add ${if forany and ${if forall.
343 PH/17 Added dsn_from option to vary the From: line in DSNs.
345 PH/18 Flush SMTP output before performing a callout, unless control =
346 no_callout_flush is set.
348 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
349 was true (the default) a successful delivery failed to delete the retry
350 item, thus causing premature timeout of the address. The bug is now
353 PH/20 Added hosts_avoid_pipelining to the smtp transport.
355 PH/21 Long custom messages for fakedefer and fakereject are now split up
356 into multiline reponses in the same way that messages for "deny" and
357 other ACL rejections are.
359 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
360 with slight modification.
362 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
363 draft, changing the syntax and factoring some duplicate code".
365 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
366 for deliveries of the second and subsequent messages over the same SMTP
369 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
370 ${reduce, with only minor "tidies".
372 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
374 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
375 expansion side effects.
377 PH/27 When a message times out after an over-quota error from an Exim-imposed
378 quota, the bounce message says "mailbox is full". This message was not
379 being given when it was a system quota that was exceeded. It now should
382 MH/03 Made $recipients available in local_scan(). local_scan() already has
383 better access to the recipient list through recipients_list[], but
384 $recipients can be useful in postmaster-provided expansion strings.
386 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
387 in the case of a MAIL command with additional options following the
388 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
389 were accidentally chopped off.
391 PH/29 SMTP synchronization checks are implemented when a command is read -
392 there is a check that no more input is waiting when there shouldn't be
393 any. However, for some commands, a delay in an ACL can mean that it is
394 some time before the response is written. In this time, more input might
395 arrive, invalidly. So now there are extra checks after an ACL has run for
396 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
397 pipelining has not been advertised.
399 PH/30 MH's patch to allow iscntrl() characters to be list separators.
401 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
402 returned in the SMTP response when smtp_return_error_details was false.
405 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
406 instead of the C I/O that was originally supplied, because problems were
409 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
410 Exim which did not show up earlier: it was assuming that a call to
411 SSL_CTX_set_info_callback() might give an error value. In fact, there is
412 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
413 was a macro that became an assignment, so it seemed to work. This has
414 changed to a proper function call with a void return, hence the compile
415 error. Exim's code has been fixed.
417 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
420 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
422 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
424 PH/37 If a message is not accepted after it has had an id assigned (e.g.
425 because it turns out to be too big or there is a timeout) there is no
426 "Completed" line in the log. When some messages of this type were
427 selected by exigrep, they were listed as "not completed". Others were
428 picked up by some special patterns. I have improved the selection
429 criteria to be more general.
431 PH/38 The host_find_failed option in the manualroute router can now be set
432 to "ignore", to completely ignore a host whose IP address cannot be
433 found. If all hosts are ignored, the behaviour is controlled by the new
434 host_all_ignored option.
436 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
437 homing or because of multiple MX records with /mx) generated more than
438 one IP address, and the following item turned out to be the local host,
439 all the secondary addresses of the first item were incorrectly removed
440 from the list, along with the local host and any following hosts (which
441 is what is supposed to happen).
443 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
444 whoever called Exim into the -H file. In the case of the daemon it was
445 behaving confusingly. When first started, it used values for whoever
446 started the daemon, but after a SIGHUP it used the Exim user (because it
447 calls itself on a restart). I have changed the code so that it now always
450 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
451 message are rejected with the same error (e.g. no authentication or bad
452 sender address), and a DATA command is nevertheless sent (as can happen
453 with PIPELINING or a stupid MUA), the error message that was given to the
454 RCPT commands is included in the rejection of the DATA command. This is
455 intended to be helpful for MUAs that show only the final error to their
458 PH/42 Another patch from the Sieve maintainer.
460 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
461 Eximstats - Fixed some broken HTML links and added missing column headers
463 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
464 columns for Rejects, Temp Rejects, Ham, and Spam rows.
466 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
468 PH/43 Yet another patch from the Sieve maintainer.
470 PH/44 I found a way to check for a TCP/IP connection going away before sending
471 the response to the final '.' that terminates a message, but only in the
472 case where the client has not sent further data following the '.'
473 (unfortunately, this is allowed). However, in many cases there won't be
474 any further data because there won't be any more messages to send. A call
475 to select() can be used: if it shows that the input is "ready", there is
476 either input waiting, or the socket has been closed. An attempt to read
477 the next input character can distinguish the two cases. Previously, Exim
478 would have sent an OK response which the client would never have see.
479 This could lead to message repetition. This fix should cure that, at
480 least in a lot of common cases.
482 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
483 advertised in response to EHLO.
489 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
490 fixed by 4.65/MH/01 (is this a record?) are fixed:
492 (i) An empty string was always treated as zero by the numeric comparison
493 operators. This behaviour has been restored.
495 (ii) It is documented that the numeric comparison operators always treat
496 their arguments as decimal numbers. This was broken in that numbers
497 starting with 0 were being interpreted as octal.
499 While fixing these problems I realized that there was another issue that
500 hadn't been noticed. Values of message_size_limit (both the global option
501 and the transport option) were treated as octal if they started with 0.
502 The documentation was vague. These values are now always treated as
503 decimal, and I will make that clear in the documentation.
509 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
510 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
513 MH/01 Don't check that the operands of numeric comparison operators are
514 integers when their expansion is in "skipping" mode (fixes bug
515 introduced by 4.64-PH/07).
517 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
518 child addresses, Exim now panics and dies. Previously, because the count
519 is held in a short int, deliveries were likely to be lost. As such a
520 large number of recipients for a single message is ridiculous
521 (performance will be very, very poor), I have chosen to impose a limit
522 rather than extend the field.
528 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
529 leftover -K file (the existence of which was triggered by #402).
530 While we were at it, introduced process PID as part of the -K
531 filename. This should rule out race conditions when creating
534 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
535 processing considerably. Previous code took too long for large mails,
536 triggering a timeout which in turn triggers #401.
538 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
539 in the DK code in transports.c. sendfile() is not really portable,
540 hence the _LINUX specificness.
542 TF/01 In the add_headers option to the mail command in an Exim filter,
543 there was a bug that Exim would claim a syntax error in any
544 header after the first one which had an odd number of characters
547 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
548 callout verification, Exim cached a "reject" for the entire domain. This
549 is correct for most verifications, but it is not correct for a recipient
550 verification with use_sender or use_postmaster set, because in that case
551 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
552 case of MAIL FROM:<> rejection from other early rejections (e.g.
553 rejection of HELO). When verifying a recipient using a non-null MAIL
554 address, the cache is ignored if it shows MAIL FROM:<> rejection.
555 Whatever the result of the callout, the value of the domain cache is
556 left unchanged (for any other kind of callout, getting as far as trying
557 RCPT means that the domain itself is ok).
559 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
562 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
563 manifest itself as EPIPE rather than ECONNECT. When tidying away a
564 session, the daemon ignores ECONNECT errors and logs others; it now
565 ignores EPIPE as well.
567 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
568 (quoted-printable decoding).
570 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
571 later the small subsequent patch to fix an introduced bug.
573 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
575 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
577 PH/08 An error is now given if message_size_limit is specified negative.
579 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
580 to be given (somewhat) arbitrary names.
582 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
585 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
586 miscellaneous code fixes
588 PH/10 Added the log_reject_target ACL modifier to specify where to log
591 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
592 hostname. This is wrong, because it relates to the incoming message (and
593 probably the interface on which it is arriving) and not to the outgoing
594 callout (which could be using a different interface). This has been
595 changed to use the value of the helo_data option from the smtp transport
596 instead - this is what is used when a message is actually being sent. If
597 there is no remote transport (possible with a router that sets up host
598 addresses), $smtp_active_hostname is used.
600 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
601 tweaks were necessary in order to get it to work (see also 21 below):
602 (a) The code assumed that strncpy() returns a negative number on buffer
603 overflow, which isn't the case. Replaced with Exim's string_format()
605 (b) There were several signed/unsigned issues. I just did the minimum
606 hacking in of casts. There is scope for a larger refactoring.
607 (c) The code used strcasecmp() which is not a standard C function.
608 Replaced with Exim's strcmpic() function.
609 (d) The code set only $1; it now sets $auth1 as well.
610 (e) A simple test gave the error "authentication client didn't specify
611 service in request". It would seem that Dovecot has changed its
612 interface. Fortunately there's a specification; I followed it and
613 changed what the client sends and it appears to be working now.
615 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
618 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
619 address A is aliased to B and C, where B exists and C does not. Without
620 -v the output is "A verified" because verification stops after a
621 successful redirection if more than one address is generated. However,
622 with -v the child addresses are also verified. Exim was outputting "A
623 failed to verify" and then showing the successful verification for C,
624 with its parentage. It now outputs "B failed to verify", showing B's
625 parentage before showing the successful verification of C.
627 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
628 look up a TXT record in a specific list after matching in a combined
631 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
632 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
633 they consult the DNS. I had assumed they would set it the way they
634 wanted; and indeed my experiments on Linux seem to show that in some
635 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
636 To be on the safe side, however, I have now made the interface to
637 host_find_byname() similar to host_find_bydns(), with an argument
638 containing the DNS resolver options. The host_find_byname() function now
639 sets these options at its start, just as host_find_bydns() does. The smtp
640 transport options dns_qualify_single and dns_search_parents are passed to
641 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
642 of host_find_byname() use the default settings of RES_DEFNAMES
643 (qualify_single) but not RES_DNSRCH (search_parents).
645 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
646 spool_read_header() do less string testing, by means of a preliminary
647 switch on the second character of optional "-foo" lines. (This is
648 overdue, caused by the large number of possibilities that now exist.
649 Originally there were few.) While I was there, I also converted the
650 str(n)cmp tests so they don't re-test the leading "-" and the first
651 character, in the hope this might squeeze out yet more improvement.
653 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
654 flag allowing group syntax was set by the header_syntax check but not
655 turned off, possible causing trouble later; (2) The flag was not being
656 set at all for the header_verify test, causing "group"-style headers to
657 be rejected. I have now set it in this case, and also caused header_
658 verify to ignore an empty address taken from a group. While doing this, I
659 came across some other cases where the code for allowing group syntax
660 while scanning a header line wasn't quite right (mostly, not resetting
661 the flag correctly in the right place). These bugs could have caused
662 trouble for malformed header lines. I hope it is now all correct.
664 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
665 with the "reply" argument non-NULL. The code, however (which originally
666 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
667 but it didn't always do it. This confused somebody who was copying the
668 code for some other use. I have removed all the tests.
670 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
671 feature that was used to support insecure browsers during the U.S. crypto
672 embargo. It requires special client support, and Exim is probably the
673 only MTA that supported it -- and would never use it because real RSA is
674 always available. This code has been removed, because it had the bad
675 effect of slowing Exim down by computing (never used) parameters for the
676 RSA_EXPORT functionality.
678 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
679 authenticator to fail if there's a tab character in the incoming data
680 (there should never be unless someone is messing about, as it's supposed
681 to be base64-encoded). Also added, on Timo's advice, the "secured" option
682 if the connection is using TLS or if the remote IP is the same as the
683 local IP, and the "valid-client-cert option" if a client certificate has
686 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
687 authenticators. This can be used for authorization after authentication
688 succeeds. (In the case of plaintext, it servers for both authentication
691 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
692 if any retry times were supplied.
694 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
695 connection, where there is no client IP address to check. In this
696 situation, the verify now always succeeds.
698 PH/25 Applied John Jetmore's -Mset patch.
700 PH/26 Added -bem to be like -Mset, but loading a message from a file.
702 PH/27 In a string expansion for a processed (not raw) header when multiple
703 headers of the same name were present, leading whitespace was being
704 removed from all of them, but trailing whitespace was being removed only
705 from the last one. Now trailing whitespace is removed from each header
706 before concatenation. Completely empty headers in a concatenation (as
709 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
710 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
712 PH/29 [Removed. This was a change that I later backed out, and forgot to
713 correct the ChangeLog entry (that I had efficiently created) before
714 committing the later change.]
716 PH/30 Exim was sometimes attempting to deliver messages that had suffered
717 address errors (4xx response to RCPT) over the same connection as other
718 messages routed to the same hosts. Such deliveries are always "forced",
719 so retry times are not inspected. This resulted in far too many retries
720 for the affected addresses. The effect occurred only when there were more
721 hosts than the hosts_max_try setting in the smtp transport when it had
722 the 4xx errors. Those hosts that it had tried were not added to the list
723 of hosts for which the message was waiting, so if all were tried, there
724 was no problem. Two fixes have been applied:
726 (i) If there are any address or message errors in an SMTP delivery, none
727 of the hosts (tried or untried) are now added to the list of hosts
728 for which the message is waiting, so the message should not be a
729 candidate for sending over the same connection that was used for a
730 successful delivery of some other message. This seems entirely
731 reasonable: after all the message is NOT "waiting for some host".
732 This is so "obvious" that I'm not sure why it wasn't done
733 previously. Hope I haven't missed anything, but it can't do any
734 harm, as the worst effect is to miss an optimization.
736 (ii) If, despite (i), such a delivery is accidentally attempted, the
737 routing retry time is respected, so at least it doesn't keep
738 hammering the server.
740 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
741 in ${readsocket because some servers need this prod.
743 PH/32 Added some extra debug output when updating a wait-xxx database.
745 PH/33 The hint "could be header name not terminated by colon", which has been
746 given for certain expansion errors for a long time, was not being given
747 for the ${if def:h_colon_omitted{... case.
749 PH/34 The spec says: "With one important exception, whenever a domain list is
750 being scanned, $domain contains the subject domain." There was at least
751 one case where this was not true.
753 PH/35 The error "getsockname() failed: connection reset by peer" was being
754 written to the panic log as well as the main log, but it isn't really
755 panic-worthy as it just means the connection died rather early on. I have
756 removed the panic log writing for the ECONNRESET error when getsockname()
759 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
760 runs only) independently of the message's sender address. This meant
761 that, if the 4xx error was in fact related to the sender, a different
762 message to the same recipient with a different sender could confuse
763 things. In particualar, this can happen when sending to a greylisting
764 server, but other circumstances could also provoke similar problems.
765 I have changed the default so that the retry time for these errors is now
766 based a combination of the sender and recipient addresses. This change
767 can be overridden by setting address_retry_include_sender=false in the
770 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
771 remote server are returned as part of bounce messages. This was not
772 happening for LMTP over a pipe (the lmtp transport), but now it is the
773 same for both kinds of LMTP.
775 PH/38 Despite being documented as not happening, Exim was rewriting addresses
776 in header lines that were in fact CNAMEs. This is no longer the case.
778 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
779 and queue runs started by the daemon processed all messages. This has
780 been fixed so that -R and -S can now usefully be given with -q<time>.
782 PH/40 Import PCRE release 6.7 (fixes some bugs).
784 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
786 PH/42 Give an error if -q is specified more than once.
788 PH/43 Renamed the variables $interface_address and $interface_port as
789 $received_ip_address and $received_port, to make it clear that these
790 values apply to message reception, and not to the outgoing interface when
791 a message is delivered. (The old names remain recognized, of course.)
793 PH/44 There was no timeout on the connect() call when using a Unix domain
794 socket in the ${readsocket expansion. There now is.
796 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
797 be meaningful with "accept".
799 SC/01 Eximstats V1.43
800 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
802 SC/02 Eximstats V1.44
803 Use a glob alias rather than an array ref in the generated
804 parser. This improves both readability and performance.
806 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
807 Collect SpamAssassin and rejection statistics.
808 Don't display local sender or destination tables unless
809 there is data to show.
810 Added average volumes into the top table text output.
812 SC/04 Eximstats V1.46
813 Collect data on the number of addresses (recipients)
814 as well as the number of messages.
816 SC/05 Eximstats V1.47
817 Added 'Message too big' to the list of mail rejection
818 reasons (thanks to Marco Gaiarin).
820 SC/06 Eximstats V1.48
821 Mainlog lines which have GMT offsets and are too short to
822 have a flag are now skipped.
824 SC/07 Eximstats V1.49 (Alain Williams)
825 Added the -emptyok flag.
827 SC/08 Eximstats V1.50
828 Fixes for obtaining the IP address from reject messages.
830 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
831 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
832 whitesspace changes from 4.64-PH/27
834 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
837 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
838 are found, allow negative numbers in numeric criteria)
840 JJ/06 exipick.20061117.2, added new $message_body_missing variable
842 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
843 to match changes made in 4.64-PH/43
845 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
847 PH/47 Put in an explicit test for a DNS lookup of an address record where the
848 "domain" is actually an IP address, and force a failure. This locks out
849 those revolvers/nameservers that support "A-for-A" lookups, in
850 contravention of the specifications.
852 PH/48 When a host name was looked up from an IP address, and the subsequent
853 forward lookup of the name timed out, the host name was left in
854 $sender_host_name, contrary to the specification.
856 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
857 restricted to single-key lookups, Exim was not diagnosing an error if
858 * or *@ was used with a query-style lookup.
860 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
862 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
863 long ago, but noone interested enough thought of it. Let's just say that
864 the "1.1" means that there are some new functions that weren't there at
865 some point in the past.
867 PH/51 Error processing for expansion failure of helo_data from an smtp
868 transport during callout processing was broken.
870 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
871 tested/used via the -bh/-bhc/-bs options.
873 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
874 bug, fixed in subsequent PCRE releases).
876 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
877 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
879 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
885 SC/01 Use a glob alias rather than an array ref in eximstats generated
886 parser. This improves both readability and performance.
888 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
889 Don't display local sender or destination tables in eximstats unless
890 there is data to show.
891 Added average volumes into the eximstats top table text output.
893 SC/03 Collect data on the number of addresses (recipients) as well
894 as the number of messages in eximstats.
896 TF/01 Correct an error in the documentation for the redirect router. Exim
897 does (usually) call initgroups() when daemonizing.
899 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
900 with consistent privilege compared to when running as a daemon.
902 TF/03 Note in the spec that $authenticated_id is not set for local
903 submissions from trusted users.
905 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
906 Thanks to Dean Brooks <dean@iglou.com> for the patch.
908 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
909 by adding some example configuration directives to the default
910 configuration file. A little bit of work is required to uncomment the
911 directives and define how usernames and passwords are checked, but
912 there is now a framework to start from.
914 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
915 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
916 without this. I don't know how relevant this is to other LDAP libraries.
918 PH/02 Add the verb name to the "unknown ACL verb" error.
920 PH/03 Magnus Holmgren's patch for filter_prepend_home.
922 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
924 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
925 directory not expanded when it should be if an expanded home directory
926 was set for the address (which is overridden by the transport).
928 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
931 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
932 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
933 because it is too late at that time, and has no effect.
935 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
936 security issue with \' (bugzilla #107). I could not use the
937 PQescapeStringConn() function, because it needs a PGconn value as one of
940 PH/08 When testing addresses using -bt, indicate those final addresses that
941 are duplicates that would not cause an additional delivery. At least one
942 person was confused, thinking that -bt output corresponded to deliveries.
943 (Suppressing duplicates isn't a good idea as you lose the information
944 about possibly different redirections that led to the duplicates.)
946 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
947 systems where poll() doesn't work, in particular OS X.
949 PH/10 Added more information to debugging output for retry time not reached.
951 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
952 operations in malware.c.
954 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
957 PH/13 If write_rejectlog was set false when logging was sent to syslog with
958 syslog_duplication set false, log lines that would normally be written
959 both the the main log and to the reject log were not written to syslog at
962 PH/14 In the default configuration, change the use of "message" in ACL warn
963 statements to "add_header".
965 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
966 not followed by a command (e.g. "seen endif").
968 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
969 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
972 PH/17 Added extra conditions to the default value of delay_warning_condition
976 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
977 { match{$h_precedence:}{(?i)bulk|list|junk} } \
978 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
981 The Auto-Submitted: and various List- headers are standardised, whereas I
982 don't think Precedence: ever was.
984 PH/18 Refactored debugging code in route_finduser() to show more information,
985 in particular, the error code if getpwnam() issues one.
987 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
988 This is apparently needed in addition to the PH/07 change above to avoid
989 any possible encoding problems.
991 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
992 but not after initializing Perl.
994 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
995 output them only if debugging. By default they are written stderr,
996 apparently, which is not desirable.
998 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1001 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1004 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1006 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1007 authenticated or an ident call has been made. Suppress the default
1008 values for $authenticated_id and $authenticated_sender (but permit -oMai
1009 and -oMas) when testing with -bh.
1011 PH/24 Re-jigged the order of the tests in the default configuration so that the
1012 tests for valid domains and recipients precede the DNS black list and CSA
1013 tests, on the grounds that those ones are more expensive.
1015 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1016 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1017 This bug exists in every version of Exim that I still have, right back to
1020 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1021 However, an attempt to turn on case-sensitivity in a regex key by
1022 including (?-i) didn't work because the subject string was already
1023 lowercased, and the effects were non-intuitive. It turns out that a
1024 one-line patch can be used to allow (?-i) to work as expected.
1030 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1031 other effects) broke the use of negated acl sub-conditions.
1033 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1036 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1037 "Deny" causes Exim to reject the incoming connection with a 554 error.
1038 Unfortunately, if there is a major crisis, such as a disk failure,
1039 tcp-wrappers gives "deny", whereas what one would like would be some
1040 kind of temporary error. A kludge has been added to help with this.
1041 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1042 554 error is used if errno is still zero or contains ENOENT (which occurs
1043 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1046 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1048 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1049 errors. Otherwise a message that provokes a temporary error (when other
1050 messages do not) can cause a whole host to time out.
1052 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1053 addresses were routed directly to files or pipes from a redirect router.
1054 File deliveries just didn't batch; pipe deliveries might have suffered
1057 PH/06 A failure to get a lock for a hints database would erroneously always say
1058 "Failed to get write lock", even when it was really a read lock.
1060 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1061 of 0600. This has been changed to use the value of the lockfile_mode
1062 option (which defaults to 0600).
1064 PH/08 Applied small patch from the Sieve maintainer.
1066 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1067 folder from quota calculations, a direct delivery into this folder messed
1068 up the contents of the maildirsize file. This was because the regex was
1069 used only to exclude .Trash (or whatever) when the size of the mailbox
1070 was calculated. There was no check that a delivery was happening into an
1071 excluded directory. This bug has been fixed by ignoring all quota
1072 processing for deliveries into excluded directories.
1074 PH/10 Added the maildirfolder_create_regex option to appendfile.
1080 PH/01 The code for finding all the local interface addresses on a FreeBSD
1081 system running IPv6 was broken. This may well have applied to all BSD
1082 systems, as well as to others that have similar system calls. The broken
1083 code found IPv4 interfaces correctly, but gave incorrect values for the
1084 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1085 that it would not match correctly against @[] and not recognize the IPv6
1088 PH/02 The ipliteral router was not recognizing addresses of the form user@
1089 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1091 PH/03 Added disable_ipv6.
1093 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1094 decoded form, because it is most often used to construct To: headers
1095 lines in autoreplies, and the decoded form may well be syntactically
1096 invalid. However, $reply_address has leading white space removed, and all
1097 newlines turned into spaces so that the autoreply transport does not
1100 PH/05 If group was specified without a user on a router, and no group or user
1101 was specified on a transport, the group from the router was ignored.
1103 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1104 for visible compile-time settings that can be used to change these
1105 numbers, for those that want even more. Backwards compatibility with old
1106 spool files has been maintained. However, going back to a previous Exim
1107 release will lost any variables that are in spool files.
1109 PH/07 Two small changes when running in the test harness: increase delay when
1110 passing a TCP/IP connection to a new process, in case the original
1111 process has to generate a bounce, and remove special handling of
1112 127.0.0.2 (sic), which is no longer necessary.
1114 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1115 be the same on different OS.
1117 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1120 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1121 whether --show-vars was specified or not
1123 JJ/02 exipick: Added support for new ACL variable spool format introduced
1126 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1127 syntactically invalid From: or Reply-to: line, and a filter used this to
1128 generate an autoreply, and therefore failed to obtain an address for the
1129 autoreply, Exim could try to deliver to a non-existent relative file
1130 name, causing unrelated and misleading errors. What now happens is that
1131 it logs this as a hard delivery error, but does not attempt to create a
1134 PH/11 The exinext utility has a -C option for testing purposes, but although
1135 the given file was scanned by exinext itself; it wasn't being passed on
1136 when Exim was called.
1138 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1139 an end-of-file indication when reading a command response.
1141 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1142 compiled. In many other places in Exim, IPv6 addresses are always
1143 recognized, so I have changed this. It also means that IPv4 domain
1144 literals of the form [IPV4:n.n.n.n] are now always recognized.
1146 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1147 used if the router is not running as root, for example, when verifying at
1148 ACL time, or when using -bh. The debugging output from this situation was
1149 non-existent - all you got was a failure to exec. I have made two
1152 (a) Failures to set uid/gid, the current directory, or a process leader
1153 in a subprocess such as that created by queryprogram now generate
1154 suitable debugging ouput when -d is set.
1156 (b) The queryprogram router detects when it is not running as root,
1157 outputs suitable debugging information if -d is set, and then runs
1158 the subprocess without attempting to change uid/gid.
1160 PH/15 Minor change to Makefile for building test_host (undocumented testing
1163 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1164 additional section of a DNS packet that returns MX or SRV records.
1165 Instead, it always explicitly searches for A/AAAA records. This avoids
1166 major problems that occur when a DNS server includes only records of one
1167 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1168 fixed another bug: if SRV records were looked up and the corresponding
1169 address records were *not* found in the additional section, the port
1170 values from the SRV records were lost.
1172 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1173 using the correct key (the original address) when searching the retry
1174 rules in order to find which one to use for generating the retry hint.
1176 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1177 adding the default one. Similarly, if it contains a Reply-To: header, the
1178 errors_reply_to option, if set, is not used.
1180 PH/19 When calculating a retry time, Exim used to measure the "time since
1181 failure" by looking at the "first failed" field in the retry record. Now
1182 it does not use this if it is later than than the arrival time of the
1183 message. Instead it uses the arrival time. This makes for better
1184 behaviour in cases where some deliveries succeed, thus re-setting the
1185 "first failed" field. An example is a quota failure for a huge message
1186 when small messages continue to be delivered. Without this change, the
1187 "time since failure" will always be short, possible causing more frequent
1188 delivery attempts for the huge message than are intended.
1189 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1191 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1192 $1, $2, $3) because the numerical variables can be reset during some
1193 expansion items (e.g. "match"), thereby losing the authentication data.
1195 PH/21 Make -bV show the size of off_t variables so that the test suite can
1196 decide whether to run tests for quotas > 2G.
1198 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1199 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1200 filecount value is greater than 2G or if a quota value is greater than 2G
1201 on a system where the size of off_t is not greater than 4, a panic error
1204 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1205 never match. The debug and -bh output now contains an explicit error
1206 message indicating a malformed IPv4 address or mask.
1208 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1209 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1210 PH/23 above applies.
1212 PH/25 Do not write to syslog when running in the test harness. The only
1213 occasion when this arises is a failure to open the main or panic logs
1214 (for which there is an explicit test).
1216 PH/26 Added the /no_tell option to "control=freeze".
1218 PH/27 If a host name lookup failed very early in a connection, for example, if
1219 the IP address matched host_lookup and the reverse lookup yielded a name
1220 that did not have a forward lookup, an error message of the form "no IP
1221 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1222 could be logged. Now it outputs the IP address instead of "NULL".
1224 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1225 allows the sender and the authenticated sender to be set when
1226 submitting a message from within Exim. Since child_open_exim() is
1227 documented for local_scan(), the new function should be too.
1229 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1230 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1231 results in an empty string is now treated as unset.
1233 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1235 PH/31 Added sender_verify_fail logging option.
1237 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1238 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1239 tidied the source and removed it altogether.
1241 PH/33 When a queue run was abandoned because the load average was too high, a
1242 log line was always written; now it is written only if the queue_run log
1243 selector is set. In addition, the log line for abandonment now contains
1244 information about the queue run such as the pid. This is always present
1245 in "start" and "stop" lines but was omitted from the "abandon" line.
1247 PH/34 Omit spaces between a header name and the colon in the error message that
1248 is given when verify = headers_syntax fails (if there are lots of them,
1249 the message gets confusing).
1251 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1252 names, as there are now some PTR records that contain slashes. This check
1253 is only to protect against broken name servers that fall over on strange
1254 characters, so the fact that it applies to all lookups doesn't matter.
1256 PH/36 Now that the new test suite is complete, we can remove some of the
1257 special code in Exim that was needed for the old test suite. For example,
1258 sorting DNS records because real resolvers return them in an arbitrary
1259 order. The new test suite's fake resolver always returns records in the
1262 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1263 bounces) except when queue_only is set, to avoid logging races between
1264 the different processes.
1266 PH/38 Panic-die if .include specifies a non-absolute path.
1268 PH/39 A tweak to the "H" retry rule from its user.
1270 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1271 a label. They prevented compilation on older perls.
1273 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1274 a warning to be raised on newish perls.
1276 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1277 on queue. Changes to match documented behaviour of showing count of
1278 messages matching specified criteria.
1280 PH/40 Changed the default ident timeout from 30s to 5s.
1282 PH/41 Added support for the use of login_cap features, on those BSD systems
1283 that have them, for controlling the resources used by pipe deliveries.
1285 PH/42 The content-scanning code uses fopen() to create files in which to put
1286 message data. Previously it was not paying any attention to the mode of
1287 the files. Exim runs with umask(0) because the rest of the code creates
1288 files with open(), and sets the required mode explicitly. Thus, these
1289 files were ending up world-writeable. This was not a big issue, because,
1290 being within the spool directory, they were not world-accessible. I have
1291 created a function called modefopen, which takes an additional mode
1292 argument. It sets umask(777), creates the file, chmods it to the required
1293 mode, then resets the umask. All the relevant calls to fopen() in the
1294 content scanning code have been changed to use this function.
1296 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1297 to 24 hours. This avoids potential overflow problems when processing G
1298 and H retry rules. I suspect nobody ever tinkers with this value.
1300 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1302 PH/45 When the plaintext authenticator is running as a client, the server's
1303 challenges are checked to ensure they are valid base64 strings. By
1304 default, the authentication attempt is cancelled if an invalid string is
1305 received. Setting client_ignore_invalid_base64 true ignores these errors.
1306 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1307 they are received. Thus, the responses can be made to depend on the
1308 challenges. If an invalid string is ignored, an empty string is placed in
1311 PH/46 Messages that are created by the autoreply transport now contains a
1312 References: header, in accordance with RFCs 2822 and 3834.
1314 PH/47 Added authenticated_sender_force to the smtp transport.
1316 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1318 PH/49 Installed latest patch from the Sieve maintainer.
1320 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1321 was also set, the appendfile transport was unnecessarily scanning a
1322 directory of message files (e.g. for maildir delivery) to find the count
1323 of files (along with the size), even though it did not need this
1324 information. It now does the scan only if it needs to find either the
1325 size of the count of files.
1327 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1329 PH/52 Two bugs concerned with error handling when the smtp transport is
1332 (i) Exim was not creating retry information for temporary errors given
1333 for individual recipients after the DATA command when the smtp transport
1334 was used in LMTP mode. This meant that they could be retried too
1335 frequently, and not timed out correctly.
1337 (ii) Exim was setting the flag that allows error details to be returned
1338 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1339 recipients that were returned after the DATA command.
1341 PH/53 This is related to PH/52, but is more general: for any failing address,
1342 when detailed error information was permitted to be returned to the
1343 sender, but the error was temporary, then after the final timeout, only
1344 "retry timeout exceeded" was returned. Now it returns the full error as
1345 well as "retry timeout exceeded".
1347 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1348 do this, and (what is worse) MTAs that accept it.
1350 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1351 will now be deprecated.
1353 PH/56 New os.c-cygwin from the Cygwin maintainer.
1355 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1356 formats (previously only available in exim formats via -bpr, -bpru,
1357 and -bpra. Now also available in native and exiqgrep formats)
1359 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1360 with very large, slow to parse queues
1362 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1364 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1366 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1367 responses to authentication challenges, though it was showing the
1368 challenges; (ii) I've removed the CR characters from the debug output for
1371 PH/58 Allow for the insertion of a newline as well as a space when a string
1372 is turned into more than one encoded-word during RFC 2047 encoding. The
1373 Sieve code now uses this.
1375 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1376 data_4xx, lost_connection, tls_required.
1378 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1379 message was being sent as an SMTP response.
1381 PH/61 Add -l and -k options to exicyclog.
1383 PH/62 When verifying, if an address was redirected to one new address, so that
1384 verification continued, and the new address failed or deferred after
1385 having set something in $address_data, the value of $address_data was not
1386 passed back to the ACL. This was different to the case when no
1387 redirection occurred. The value is now passed back in both cases.
1389 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1390 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1391 use login_cap.h, so on its own it isn't the distinguishing feature. The
1392 new name refers directly to the setclassresources() function.
1394 PH/65 Added configuration files for NetBSD3.
1396 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1398 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1399 is preferred over IPv4.
1401 PH/68 The bounce_return_message and bounce_return_body options were not being
1402 honoured for bounces generated during the reception of non-SMTP messages.
1403 In particular, this applied to messages rejected by the ACL. This bug has
1404 been fixed. However, if bounce_return_message is true and bounce_return_
1405 body is false, the headers that are returned for a non-SMTP message
1406 include only those that have been read before the error was detected.
1407 (In the case of an ACL rejection, they have all been read.)
1409 PH/69 The HTML version of the specification is now built in a directory called
1410 spec_html instead of spec.html, because the latter looks like a path with
1411 a MIME-type, and this confuses some software.
1413 PH/70 Catch two compiler warnings in sieve.c.
1415 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1416 function verify_get_ident() calls ip_connect() to connect a socket, but
1417 if the "connect()" function timed out, ip_connect() used to close the
1418 socket. However, verify_get_ident() also closes the socket later, and in
1419 between Exim writes to the log, which may get opened at this point. When
1420 the socket was closed in ip_connect(), the log could get the same file
1421 descriptor number as the socket. This naturally causes chaos. The fix is
1422 not to close the socket in ip_connect(); the socket should be closed by
1423 the function that creates it. There was only one place in the code where
1424 this was missing, in the iplookup router, which I don't think anybody now
1425 uses, but I've fixed it anyway.
1427 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1428 well as to direct DNS lookups. Otherwise the handling of names in host
1429 lists is inconsistent and therefore confusing.
1435 PH/01 Two changes to the default runtime configuration:
1437 (1) Move the checks for relay_from_hosts and authenticated clients from
1438 after to before the (commented out) DNS black list checks.
1440 (2) Add control=submission to the relay_from_hosts and authenticated
1441 clients checks, on the grounds that messages accepted by these
1442 statements are most likely to be submissions.
1444 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1446 (1) Generate an error if the third argument for the ${prvs expansion is
1449 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1452 (3) Reset the variables that are obtained from the first argument of
1453 ${prvscheck and used in the second argument before leaving the code,
1454 because their memory is reclaimed, so using them afterwards may do
1457 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1458 one (it's much easier than Tom thought :-).
1460 (5) Because of (4), we can now allow for the use of $prvscheck_result
1461 inside the third argument.
1463 PH/03 For some reason, the default setting of PATH when running a command from
1464 a pipe transport was just "/usr/bin". I have changed it to
1467 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1468 anything to be listed in the output from -bV.
1470 PH/05 When a filter generated an autoreply, the entire To: header line was
1471 quoted in the delivery log line, like this:
1473 => >A.N.Other <ano@some.domain> <original@ddress> ...
1475 This has been changed so that it extracts the operative address. There
1476 may be more than one such address. If so, they are comma-separated, like
1479 => >ano@some.domain,ona@other.domain <original@ddress> ...
1481 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1482 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1483 not being looked up in the rDNS to get a host name, Exim was showing the
1484 IP address twice in Received: lines, even though the IP addresses were
1485 identical. For example:
1487 Received: from [1.2.3.4] (helo=[1.2.3.4])
1489 However, if the real host name was known, it was omitting the HELO data
1490 if it matched the actual IP address. This has been tidied up so that it
1491 doesn't show the same IP address twice.
1493 PH/07 When both +timestamp and +memory debugging was on, the value given by
1494 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1495 called by the debug printing, thereby overwriting the timestamp buffer.
1496 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1498 PH/08 When the original message was included in an autoreply transport, it
1499 always said "this is a copy of the message, including all the headers",
1500 even if body_only or headers_only was set. It now gives an appropriate
1503 PH/09 Applied a patch from the Sieve maintainer which:
1505 o fixes some comments
1506 o adds the (disabled) notify extension core
1507 o adds some debug output for the result of if/elsif tests
1508 o points to the current vacation draft in the documentation
1509 and documents the missing references header update
1513 o fixes a bug in processing the envelope test (when testing
1514 multiple envelope elements, the last element determinted the
1517 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1518 Electronic Mail") by including:
1520 Auto-submitted: auto-generated
1522 in the messages that it generates (bounce messages and others, such as
1523 warnings). In the case of bounce messages for non-SMTP mesages, there was
1524 also a typo: it was using "Auto_submitted" (underscore instead of
1525 hyphen). Since every message generated by Exim is necessarily in response
1526 to another message, thes have all been changed to:
1528 Auto-Submitted: auto-replied
1530 in accordance with these statements in the RFC:
1532 The auto-replied keyword:
1534 - SHOULD be used on messages sent in direct response to another
1535 message by an automatic process,
1537 - MUST NOT be used on manually-generated messages,
1539 - MAY be used on Delivery Status Notifications (DSNs) and Message
1540 Disposition Notifications (MDNs),
1542 - MUST NOT be used on messages generated by automatic or periodic
1543 processes, except for messages which are automatic responses to
1546 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1547 to the default Received: header definition.
1549 PH/12 Added log selector acl_warn_skipped (default on).
1551 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1552 variables because (a) they are in the wrong storage pool and (b) even if
1553 they were copied, it wouldn't work properly because of the caching.
1555 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1556 checking when decoding. Apparently there are clients that generate
1557 overlong encoded strings. Why am I not surprised?
1559 PH/15 If the first argument of "${if match_address" was not empty, but did not
1560 contain an "@" character, Exim crashed. Now it writes a panic log message
1561 and treats the condition as false.
1563 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1565 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1566 extension "envelope-auth". The code is finished and in agreement with
1567 other implementations, but there is no documentation so far and in fact,
1568 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1569 not changing the active code.
1571 Print executed "if" and "elsif" statements when debugging is used. This
1572 helps a great deal to understand what a filter does.
1574 Document more things not specified clearly in RFC3028. I had all this
1575 sorted out, when out of a sudden new issues came to my mind. Oops."
1577 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1580 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1581 canonical form (as documented). However, after a host name lookup from
1582 the IP address, check_host() was doing a simple string comparison with
1583 addresses acquired from the DNS when checking that the found name did
1584 have the original IP as one of its addresses. Since any found IPv6
1585 addresses are likely to be in abbreviated form, the comparison could
1586 fail. Luckily, there already exists a function for doing the comparison
1587 by converting both addresses to binary, so now that is used instead of
1588 the text comparison.
1590 PH/20 There was another similar case to PH/19, when a complete host name was
1591 given in a host list; looking up its IP address could give an abbreviated
1592 form, whereas the current host's name might or might not be abbreviated.
1593 The same fix has been applied.
1599 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1600 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1603 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1604 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1606 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1608 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1609 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1610 Also, for "Auto-Submitted", treat anything other than "no" as
1611 non-personal, in accordance with RFC 3834. (Previously it treated
1612 anything starting "auto-" as non-personal.)
1614 TF/01 The control=submission/name=... option had a problem with syntax
1615 errors if the name included a slash character. The /name= option
1616 now slurps the rest of the string, so it can include any characters
1617 but it must come last in the list of options (after /sender_retain
1620 PH/05 Some modifications to the interface to the fake nameserver for the new
1628 TK/01 Added the "success_on_redirect" address verification option. See
1629 NewStuff for rationale and an example.
1631 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1633 PH/02 Patch to exigrep to allow it to work on syslog lines.
1635 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1636 fread() to skip over the body file's header line, because in Cygwin the
1637 header line is locked and is inaccessible.
1639 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1640 co-exist for some time) to make it clear that it is the Exim ID that is
1641 referenced, not the Message-ID: header line.
1643 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1644 string_format() function, because snprintf() does not exist on all
1647 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1648 ${stat: expansion item.
1650 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1651 protocol synchronization error", to keep the pedants happy.
1653 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1654 well as for IRIX systems, when gcc is being used. See the host.c source
1657 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1659 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1662 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1663 transport and to the smtp transport in LMTP mode.
1665 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1667 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1668 run a filter in a subprocess. This could lead to confusion in subsequent
1669 lookups in the parent process. There should also be a search_tidyup() at
1670 the end of the subprocess.
1672 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1673 only if the host matched helo_try_verify_hosts, which caused the
1674 verification to occur when the EHLO/HELO command was issued. The ACL just
1675 tested the remembered result. Now, if a previous verification attempt has
1676 not happened, "verify = helo" does it there and then.
1678 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1680 TK/03 Fix log output including CR from clamd.
1682 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1683 exist provoked a memory error which could cause a segfault.
1685 PH/15 Installed PCRE 6.2
1687 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1689 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1690 of the problem. Specifically, suggested +O2 rather than +O1 for the
1693 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1695 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1696 record, the port was not correctly logged when the outgoing_port log
1697 selector was set (it logged the transort's default port).
1699 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1700 fallback_hosts, and "hosts" in the smtp transport.
1702 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1703 host errors such as "Connection refused".
1705 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1706 authentication with radiusclient 0.4.9:
1708 - Error returned from rc_read_config was caught wrongly
1709 - Username/password not passed on to radius server due to wrong length.
1711 The presumption is that some radiusclient API changes for 4.51/PH/17
1712 were not taken care of correctly. The code is still untested by me (my
1713 Linux distribution still has 0.3.2 of radiusclient), but it was
1714 contributed by a Radius user.
1716 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1717 expanding the "port" option of the smtp transport.
1719 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1720 while reading a MIME header. Thanks to Tom Hughes for a patch.
1722 PH/24 Include config.h inside local_scan.h so that configuration settings are
1725 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1726 that in an ACL for RCPT (for example), you can examine exactly what was
1729 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1730 commands, but it was not correctly comparing the address with the actual
1731 client host address. Thus, it would show the EHLO address in Received:
1732 header lines when this was not necessary.
1734 PH/27 Added the % operator to ${eval:}.
1736 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1737 it should be ignoring failures (because with -C, for example, it has lost
1738 privilege). It wasn't ignoring creation failures other than "already
1741 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1744 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1745 because an input line was too long, either on its own, or by virtue of
1746 too many continuations, the temporary file was not being removed, and the
1747 return code was incorrect.
1749 PH/31 Missing "BOOL" in function definition in filtertest.c.
1751 PH/32 Applied Sieve patches from the maintainer.
1753 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1755 PH/33 Added "verify = not_blind".
1757 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1758 Local/Makefile (with some defaults set). These are used in built scripts
1759 such as exicyclog, but they have never been used in the exim_install
1760 script (though there are many overriding facilities there). I have
1761 arranged that the exim_install script now takes note of these two
1764 PH/35 Installed configuration files for Dragonfly.
1766 PH/36 When a locally submitted message by a trusted user did not contain a
1767 From: header, and the sender address was obtained from -f or from an SMTP
1768 MAIL command, and the trusted user did not use -F to supply a sender
1769 name, $originator_name was incorrectly used when constructing a From:
1770 header. Furthermore, $originator_name was used for submission mode
1771 messages from external hosts without From: headers in a similar way,
1772 which is clearly wrong.
1774 PH/37 Added control=suppress_local_fixups.
1776 PH/38 When log_selector = +received_sender was set, and the addition of the
1777 sender made the log line's construction buffer exactly full, or one byte
1778 less than full, an overflow happened when the terminating "\n" was
1781 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1782 when the result of a list match is failure because a DNS lookup failed.
1784 PH/40 RM_COMMAND is now used in the building process.
1786 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1787 the "build-* directories that it finds.
1789 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1790 address in a domain literal was a prefix of an interface address.
1792 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1793 when verifying a sender address, unless rewrite_headers is false.
1795 PH/44 Wrote a long comment about why errors_to addresses are verified as
1796 recipients, not senders.
1798 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1799 the ratelimit ACL was added.
1801 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1803 PH/46 Added extra information about PostgreSQL errors to the error string.
1805 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1806 avoiding the need to install special zones in a real server. This is
1807 backwards compatible; if it can't find the fake resolver, it drops back.
1808 Thus, both old and new test suites can be run.
1810 TF/02 Added util/ratelimit.pl
1812 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1813 clock is set back in time.
1815 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1816 Candler <B.Candler@pobox.com>.
1818 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1819 OK for addresses that are the result of redirections.
1821 PH/48 A number of further additions for the benefit of the new test suite,
1822 including a fake gethostbyname() that interfaces to the fake DNS resolver
1825 TF/06 The fix for widen_domains has also been applied to qualify_single and
1826 search_parents which are the other dnslookup options that can cause
1829 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1832 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1834 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1835 service" if it encountered an MX record with an empty target hostname.
1836 The message is now "an MX or SRV record indicated no SMTP service".
1838 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1839 if the DNS of the sending site is misconfigured. This is quite a
1840 common situation. This change restores the behaviour of treating a
1841 helo verification defer as a failure.
1843 PH/51 If self=fail was set on a router, the bounce message did not include the
1844 actual error message.
1850 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1852 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1853 command itself ended in error, the underlying message about the transport
1854 filter timeout was being overwritten with the pipe command error. Now the
1855 underlying error message should be appended to the second error message.
1857 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1859 PH/02 Reduce the amount of output that "make" produces by default. Full output
1860 can still be requested.
1862 PH/03 The warning log line about a condition test deferring for a "warn" verb
1863 was being output only once per connection, rather than after each
1864 occurrence (because it was using the same function as for successful
1865 "warn" verbs). This seems wrong, so I have changed it.
1867 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1868 it should not have, which might have caused a crash in the right
1869 circumstances, but probably never did.
1871 PH/04 Installed a modified version of Tony Finch's patch to make submission
1872 mode fix the return path as well as the Sender: header line, and to
1873 add a /name= option so that you can make the user's friendly name appear
1876 TF/03 Added the control = fakedefer ACL modifier.
1878 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1879 Mark Lowes for thorough testing.
1881 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1883 TK/03 Merged latest SRS patch from Miles Wilton.
1885 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1886 with the definition in sysexits.h (which is #included earlier).
1887 Fortunately, Exim does not actually use EX_OK. The code used to try to
1888 preserve the sysexits.h value, by assumimg that macro definitions were
1889 scanned for macro replacements. I have been disabused of this notion,
1890 so now the code just undefines EX_OK before #including unistd.h.
1892 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1893 in the smtp transport. When a block could not be written in a single
1894 write() function, the timeout was being re-applied to each part-write.
1895 This seems wrong - if the receiver was accepting one byte at a time it
1896 would take for ever. The timeout is now adjusted when this happens. It
1897 doesn't have to be particularly precise.
1899 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1900 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1902 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1903 without a domain if the check to <postmaster@domain> fails.
1905 SC/01 Eximstats: added -xls and the ability to specify output files
1906 (patch written by Frank Heydlauf).
1908 SC/02 Eximstats: use FileHandles for outputing results.
1910 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1912 SC/04 Eximstats: fixed display of large numbers with -nvr option
1914 SC/05 Eximstats: fixed merging of reports with empty tables.
1916 SC/06 Eximstats: added the -include_original_destination flag
1918 SC/07 Eximstats: removed tabs and trailing whitespace.
1920 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1922 TK/06 MBOX spool code: Add real "From " MBOX separator line
1923 so the .eml file is really in mbox format (even though
1924 most programs do not really care). Patch from Alex Miller.
1926 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1927 The latter is generated from $received_to and is only set if the
1928 message has one envelope recipient. SA can use these headers,
1929 obviously out-of-the-box. Patch from Alex Miller.
1931 PH/08 The ${def test on a variable was returning false if the variable's
1932 value was "0", contrary to what the specification has always said!
1933 The result should be true unless the variable is empty.
1935 PH/09 The syntax error of a character other than { following "${if
1936 def:variable_name" (after optional whitespace) was not being diagnosed.
1937 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1938 accidental colon was present, for example, could give incorrect results.
1940 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1941 result is used (not including appendfile, where other changes are about
1944 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1945 This involved changing a lot of size variables from int to off_t. It
1946 should work with maildirs and everything.
1948 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1949 spamd dying while we are connected to it.
1951 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1954 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1955 function for BDB 4.3.
1957 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1959 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1960 the sources for PCRE have been re-arranged and more files are now
1963 PH/15 The code I had for printing potentially long long variables in PH/11
1964 above was not the best (it lost precision). The length of off_t variables
1965 is now inspected at build time, and an appropriate printing format (%ld
1966 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1967 to be "long long int" or "long int". This is needed for the internal
1968 formatting function string_vformat().
1970 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1971 the configuration file to be ":syslog", then the script "guesses" where
1972 the logs files are, rather than using the compiled in default. In our
1973 case the guess is not the same as the compiled default, so the script
1974 suddenly stopped working when I started to use syslog. The patch checks
1975 to see if log_file_path is "". If so, it attempts to read it from exim
1976 with no configuration file to get the compiled in version, before it
1977 falls back to the previous guessing code."
1979 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1980 implementing BATV in an Exim configuration. See NewStuff for the gory
1983 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1984 Makefile that are specific to HP-UX.
1986 PH/18 If the "use_postmaster" option was set for a recipient callout together
1987 with the "random" option, the postmaster address was used as the MAIL
1988 FROM address for the random test, but not for the subsequent recipient
1989 test. It is now used for both.
1991 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1992 patch removes a few documentation additions to RFC 3028, because the
1993 latest draft now contains them. It adds the new en;ascii-case comparator
1994 and a new error check for 8bit text in MIME parts. Comparator and
1995 require names are now matched exactly. I enabled the subaddress
1996 extension, but it is not well tested yet (read: it works for me)."
1998 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1999 rework some of the code of TK/09 above to avoid the hardwired use of
2000 "%lld" and "long long". Replaced the call to snprintf() with a call to
2003 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2004 records point to non-existent hosts", "retry timeout exceeded", and
2005 "retry time not reached for any host after a long failure period".
2007 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2008 experimental DomainKeys support:
2010 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2011 (2) On an error such as an illegally used "control", the wrong name for
2012 the control was given.
2014 These problems did NOT occur unless DomainKeys support was compiled.
2016 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2018 PH/24 Added ${if match_ip condition.
2020 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2021 databases so that it will be absolutely obvious if a crash occurs in the
2022 DB library. This is a regular occurrence (often caused by mis-matched
2025 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2026 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2027 picked up on a user's system that detects such things. There doesn't seem
2028 to be a gcc warning option for this - only an attribute that has to be
2029 put on the function's prototype. It seems that in Fedora Core 4 they have
2030 set this on a number of new functions. No doubt there will be more in due
2033 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2034 specify a transport. However, if an address that was verified by such a
2035 router was the subject of a callout, Exim crashed because it tried to
2036 read the rcpt_include_affixes from the non-existent transport. Now it
2037 just assumes that the setting of that option is false. This bug was
2038 introduced by 4.51/PH/31.
2040 PH/28 Changed -d+all to exclude +memory, because that information is very
2041 rarely of interest, but it makes the output a lot bigger. People tend to
2042 do -d+all out of habit.
2044 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2045 code in os-type was giving problems when libc.so lives in lib64, like on
2048 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2049 aren't the modern standard, and it seems that some systems' include files
2050 don't always have them. Exim was already checking for some of the newer
2051 ones like T_AAAA, and defining it itself. I've added checks for all the
2052 record types that Exim uses.
2054 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2055 not automatically generating a new one, as it is supposed to. This
2056 prevented TLS from working. If the file did exist, but contained invalid
2057 data, a new version was generated, as expected. It was only the case of a
2058 non-existent file that was broken.
2060 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2061 with a change in libdomainkeys > 0.64.
2063 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2064 from DNS. If the selector record carries the flag, it now has
2065 precedence over the domain-wide flag.
2067 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2069 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2070 the use of a port name that isn't defined in /etc/services) occurred, the
2071 message was deferred as in a normal delivery, and thus remained on the
2072 spool, instead of being failed because of the mua_wrapper setting. This
2073 is now fixed, and I tidied up some of the mua_wrapper messages at the
2076 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2077 the messages in a hash of arrays rather than using individual hashes.
2078 This is a bit cleaner and results in dramatic memory savings, albeit
2079 at a slight CPU cost.
2081 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2082 as requested by Marc Sherman.
2084 SC/10 Eximstats: added histograms for user specified patterns as requested
2087 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2089 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2090 fopen() in the content-scanning modules that did not already have it.
2096 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2097 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2099 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2101 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2103 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2104 to test Sieve filters that use "vacation".
2106 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2107 that changes the way the GnuTLS parameters are stored in the cache file.
2108 The new format can be generated externally. For backward compatibility,
2109 if the data in the cache doesn't make sense, Exim assumes it has read an
2110 old-format file, and it generates new data and writes a new file. This
2111 means that you can't go back to an older release without removing the
2114 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2115 work if there are any delivery delays because "one_time" forces the
2116 parent to be marked "delivered", so its unseen clone is never tried
2117 again. For this reason, Exim now forbids the simultaneous setting of
2120 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2121 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2122 entry if you want to know the details. The fix, however, neglected to
2123 consider the case when local delivery batching is involved. The test for
2124 "previously delivered" was not happening when checking to see if an
2125 address could be batched with a previous (undelivered) one; under
2126 certain circumstances this could lead to multiple deliveries to the same
2129 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2130 in its include files, and this causes problems building Exim.
2132 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2133 header_syntax) but Exim was just ignoring anything given after a slash.
2134 In particular, this caused confusion with an attempt to use "verify =
2135 reverse_host_lookup/defer_ok". An error is now given when options are
2136 supplied for verify items that do not have them. (Maybe reverse_host_
2137 lookup should have a defer_ok option, but that's a different point.)
2139 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2140 defined by RFC 821) to 2048, because there were problems with some AUTH
2141 commands, and RFC 1869 says the size should be increased for extended
2142 SMTP commands that take arguments.
2144 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2147 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2148 "unknown" error; now it says that the functionality isn't in the binary.
2150 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2151 an address' error message when a string expansion fails (syntax or
2152 whatever). Otherwise the password may appear in the log. Following change
2153 PH/42 below, there is no longer a chance of it appearing in a bounce
2156 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2158 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2159 of its addresses was ever tried. (Bugzilla bug #2.)
2161 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2162 the result incorrectly in the debug output. (It correctly added a newline
2163 to what was transported.)
2165 TF/01 Added $received_time.
2167 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2168 commented out examples of how to interface to a virus scanner and to
2169 SpamAssassin. Also added commented examples of av_scanner and
2170 spamd_address settings.
2172 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2173 and controls are allowed in which ACLs. There were a couple of minor
2174 errors. Some of the entries in the conditions table (which is a table of
2175 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2176 negation of where the condition IS allowed.
2178 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2180 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2181 header file does not have a version number, so I've had to invent a new
2182 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2183 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2184 radiusclient), but it was contributed by a Radius user.
2186 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2187 files or directories, for OpenSSL.
2189 PH/19 When an Exim process that is running as root has to create an Exim log
2190 file, it does so in a subprocess that runs as exim:exim so as to get the
2191 ownership right at creation (otherwise, other Exim processes might see
2192 the file with the wrong ownership). There was no test for failure of this
2193 fork() call, which would lead to the process getting stuck as it waited
2194 for a non-existent subprocess. Forks do occasionally fail when resources
2195 run out. I reviewed all the other calls to fork(); they all seem to check
2198 PH/20 When checking for unexpected SMTP input at connect time (before writing
2199 the banner), Exim was not dealing correctly with a non-positive return
2200 from the read() function. If the client had disconnected by this time,
2201 the result was a log entry for a synchronization error with an empty
2202 string after "input=" when read() returned zero. If read() returned -1
2203 (an event I could not check), uninitialized data bytes were printed.
2204 There were reports of junk text (parts of files, etc) appearing after
2207 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2209 PH/22 Added support for macro redefinition, and (re)definition in between
2210 driver and ACL definitions.
2212 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2213 forgetting to use the resulting value; it was using the unexpanded value.
2215 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2216 hadn't been configured. The fix is from Juergen Kreileder, who
2217 understands it better than I do:
2219 "Here's what I see happening with three configured cyrus_sasl
2220 authenticators configured (plain, login, cram-md5):
2222 On startup auth_cyrus_sasl_init() gets called for each of these.
2223 This means three calls to sasl_listmech() without a specified mech_list.
2224 => SASL tests which mechs of all available mechs actually work
2225 => three warnings about OTP not working
2226 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2228 With the patch, sasl_listmech() also gets called three times. But now
2229 SASL's mech_list option is set to the server_mech specified in the the
2230 authenticator. Or in other words, the answer from sasl_listmech()
2231 gets limited to just the mech you're testing for (which is different
2233 => the return list contains just 'plain' or 'login', 'cram-md5' or
2234 nothing depending on the value of ob->server_mech.
2236 I've just tested the patch: Authentication still works fine,
2237 unavailable mechs specified in the exim configuration are still
2238 caught, and the auth.log warnings about OTP are gone."
2240 PH/25 When debugging is enabled, the contents of the command line are added
2241 to the debugging output, even when log_selector=+arguments is not
2244 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2245 answer is "GNU", and only if the return is "GNU/something" is the answer
2248 PH/27 $acl_verify_message is now set immediately after the failure of a
2249 verification in an ACL, and so is available in subsequent modifiers. In
2250 particular, the message can be preserved by coding like this:
2252 warn !verify = sender
2253 set acl_m0 = $acl_verify_message
2255 Previously, $acl_verify_message was set only while expanding "message"
2256 and "log_message" when a very denied access.
2258 PH/28 Modified OS/os.c-Linux with
2260 -#ifndef OS_LOAD_AVERAGE
2261 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2263 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2264 nomenclature these days.)
2266 PH/29 Installed patch from the Sieve maintainer that adds the options
2267 sieve_useraddress and sieve_subaddress to the redirect router.
2269 PH/30 In these circumstances:
2270 . Two addresses routed to the same list of hosts;
2271 . First host does not offer TLS;
2272 . First host accepts first address;
2273 . First host gives temporary error to second address;
2274 . Second host offers TLS and a TLS session is established;
2275 . Second host accepts second address.
2276 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2277 and peerdn, if requested) that were in fact used only for the second
2280 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2281 attention to any local part prefix or suffix that was matched by the
2282 router that accepted the address. It now behaves in the same way as it
2283 does for delivery: the affixes are removed from the local part unless
2284 rcpt_include_affixes is set on the transport.
2286 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2287 timeout during the DATA phase of an incoming message.
2289 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2290 applied a patch sanctioned by the Sieve maintainer.
2292 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2293 the uid or gid is negative. A case of a negative gid caused this to be
2294 noticed. The fix allows for either to be negative.
2296 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2297 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2300 PH/36 The change PH/12 above was broken. Fixed it.
2302 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2303 the grounds that routing the same address twice would always produce the
2304 same answer. This might have been true once, but it is certainly no
2305 longer true now. Routing a child address may depend on the previous
2306 routing that produced that child. Some complicated redirection strategies
2307 went wrong when messages had multiple recipients, and made Exim's
2308 behaviour dependent on the order in which the addresses were given.
2310 I have moved the duplicate checking until after the routing is complete.
2311 Exim scans the addresses that are assigned to local and remote
2312 transports, and removes any duplicates. This means that more work will be
2313 done, as duplicates will always all be routed, but duplicates are
2314 presumably rare, so I don't expect this is of any significance.
2316 For deliveries to pipes, files, and autoreplies, the duplicate checking
2317 still happens during the routing process, since they are not going to be
2320 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2321 It corrects a timeout issue with spamd. This is Ian's comment: "The
2322 background is that sometimes spamd either never reads data from a
2323 connection it has accepted, or it never writes response data. The exiscan
2324 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2325 blindly assumes that writes won't block so it may never time out."
2327 PH/39 Allow G after quota size as well as K and M.
2329 PH/40 The value set for $authenticated_id in an authenticator may not contain
2330 binary zeroes or newlines because the value is written to log lines and
2331 to spool files. There was no check on this. Now the value is run through
2332 the string_printing() function so that such characters are converted to
2333 printable escape sequences.
2335 PH/41 $message_linecount is a new variable that contains the total number of
2336 lines in the message. Compare $body_linecount, which is the count for the
2339 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2340 bounce and delay warning messages, except in certain special cases, which
2343 (a) An SMTP error message from a remote host;
2344 (b) A message specified in a :fail: redirection;
2345 (c) A message specified in a "fail" command in a system filter;
2346 (d) A message specified in a FAIL return from the queryprogram router;
2347 (e) A message specified by the cannot_route_message router option.
2349 In these cases only, Exim does include the error details in bounce and
2350 warning messages. There are also a few cases where bland messages such
2351 as "unrouteable address" or "local delivery error" are given.
2353 PH/43 $value is now also set for the "else" part of a ${run expansion.
2355 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2356 being worked on, but at least Exim now implements the latest version to
2359 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2360 process to complete was treated as a delivery failure, a timeout while
2361 writing the message to the pipe was logged, but erroneously treated as a
2362 successful delivery. Such timeouts include transport filter timeouts. For
2363 consistency with the overall process timeout, these timeouts are now
2364 treated as errors, giving rise to delivery failures by default. However,
2365 there is now a new Boolean option for the pipe transport called
2366 timeout_defer, which, if set TRUE, converts the failures into defers for
2367 both kinds of timeout. A transport filter timeout is now identified in
2370 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2371 systems where "make" and "gmake" are different, calling "gmake" at top
2372 level broke things. I've arranged for the value of $(MAKE) to be passed
2373 from the Makefile to this script so that it can call the same version of
2377 A note about Exim versions 4.44 and 4.50
2378 ----------------------------------------
2380 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2381 changes of various kinds. As a consequence, a big documentation update was
2382 needed. This delayed the release for rather longer than seemed good, especially
2383 in the light of a couple of (minor) security issues. Therefore, the changes
2384 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2385 release. So 4.44 and 4.50 are in effect two different branches that both start
2388 I have left the 4.50 change log unchanged; it contains all the changes since
2389 4.43. The change log for 4.44 is below; many of its items are identical to
2390 those for 4.50. This seems to be the most sensible way to preserve the
2391 historical information.
2397 1. Minor wording change to the doc/README.SIEVE file.
2399 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2400 computation of the current number of files was incorrect.
2402 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2403 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2406 4. Give more explanation in the error message when the command for a transport
2407 filter fails to execute.
2409 5. There are several places where Exim runs a non-Exim command in a
2410 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2411 was being done only for the command run by the queryprogram router. It is
2412 now done for all such subprocesses. The other cases are: ${run, transport
2413 filters, and the commands run by the lmtp and pipe transports.
2415 6. Added CONFIGURE_GROUP build-time option.
2417 7. Some older OS have a limit of 256 on the maximum number of file
2418 descriptors. Exim was using setrlimit() to set 1000 as a large value
2419 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2420 systems. I've change it so that if it can't get 1000, it tries for 256.
2422 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2423 was an oversight, and furthermore, ever since the addition of extra
2424 controls (e.g. 4.43/32), the checks on when to allow different forms of
2425 "control" were broken. There should now be diagnostics for all cases when a
2426 control that does not make sense is encountered.
2428 9. Added the /retain_sender option to "control=submission".
2430 10. $recipients is now available in the predata ACL (oversight).
2432 11. Tidy the search cache before the fork to do a delivery from a message
2433 received from the command line. Otherwise the child will trigger a lookup
2434 failure and thereby defer the delivery if it tries to use (for example) a
2435 cached ldap connection that the parent has called unbind on.
2437 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2438 of $address_data from the recipient verification was clobbered by the
2439 sender verification.
2441 13. The value of address_data from a sender verification is now available in
2442 $sender_address_data in subsequent conditions in the ACL statement.
2444 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2446 15. Added a new option "connect=<time>" to callout options, to set a different
2449 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2450 was its contents. (It was OK if the option was not defined at all.)
2452 17. A "Completed" log line is now written for messages that are removed from
2453 the spool by the -Mrm option.
2455 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2456 information about exactly what failed.
2458 19. Added -dd to debug only the daemon process.
2460 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2461 handles timeouts, both on the server side and network timeouts. Renamed the
2462 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2464 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2465 It is now set to "smtps".
2467 22. $host_address is now set to the target address during the checking of
2468 ignore_target_hosts.
2470 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2471 being passed; this would have caused $sender_host_name to have been used if
2472 matching the list had actually called for a host name (not very likely,
2473 since this list is usually IP addresses). A host name is now passed as
2476 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2477 code that specifies a non-restarting handler (typically sigaction() in
2478 modern systems) in an attempt to fix a rare and obscure crash bug.
2480 25. Narrowed the window for a race in the daemon that could cause it to ignore
2481 SIGCHLD signals. This is not a major problem, because they are used only to
2482 wake it up if nothing else does.
2484 26. A malformed maildirsize file could cause Exim to calculate negative values
2485 for the mailbox size or file count. Odd effects could occur as a result.
2486 The maildirsize information is now recalculated if the size or filecount
2489 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2490 support for a long time. Removed HAVE_SYS_VFS_H.
2492 28. Installed the latest version of exipick from John Jetmore.
2494 29. In an address list, if the pattern was not a regular expression, an empty
2495 subject address (from a bounce message) matched only if the pattern was an
2496 empty string. Non-empty patterns were not even tested. This was the wrong
2497 because it is perfectly reasonable to use an empty address as part of a
2498 database query. An empty address is now tested by patterns that are
2499 lookups. However, all the other forms of pattern expect the subject to
2500 contain a local part and a domain, and therefore, for them, an empty
2501 address still always fails if the pattern is not itself empty.
2503 30. Exim went into a mad DNS loop when attempting to do a callout where the
2504 host was specified on an smtp transport, and looking it up yielded more
2505 than one IP address.
2507 31. Re-factored the code for checking spool and log partition space into a
2508 function that finds that data and another that does the check. The former
2509 is then used to implement four new variables: $spool_space, $log_space,
2510 $spool_inodes, and $log_inodes.
2512 32. The RFC2047 encoding function was originally intended for short strings
2513 such as real names; it was not keeping to the 75-character limit for
2514 encoded words that the RFC imposes. It now respects the limit, and
2515 generates multiple encoded words if necessary. To be on the safe side, I
2516 have increased the buffer size for the ${rfc2047: expansion operator from
2519 33. It is now permitted to omit both strings after an "if" condition; if the
2520 condition is true, the result is "true". As before, when the second string
2521 is omitted, a false condition yields an empty string. This makes it less
2522 cumbersome to write custom ACL and router conditions.
2524 34. Failure to deliver a bounce message always caused it to be frozen, even if
2525 there was an errors_to setting on the router. The errors_to setting is now
2528 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2529 canonical form (fully expanded) before being placed in
2530 $sender_host_address.
2532 36. The table in the code that translates DNS record types into text (T_A to
2533 "A" for instance) was missing entries for NS and CNAME. It is just possible
2534 that this could have caused confusion if both these types were looked up
2535 for the same domain, because the text type is used as part of Exim's
2536 per-process caching. But the chance of anyone hitting this buglet seems
2539 37. The dnsdb lookup has been extended in a number of ways.
2541 (1) There is a new type, "zns", which walks up the domain tree until it
2542 finds some nameserver records. It should be used with care.
2544 (2) There is a new type, "mxh", which is like "mx" except that it returns
2545 just the host names, not the priorities.
2547 (3) It is now possible to give a list of domains (or IP addresses) to be
2548 looked up. The behaviour when one of the lookups defers can be
2549 controlled by a keyword.
2551 (4) It is now possible to specify the separator character for use when
2552 multiple records are returned.
2554 38. The dnslists ACL condition has been extended: it is now possible to supply
2555 a list of IP addresses and/or domains to be looked up in a particular DNS
2558 39. Added log_selector=+queue_time_overall.
2560 40. When running the queue in the test harness, wait just a tad after forking a
2561 delivery process, to get repeatability of debugging output.
2563 41. Include certificate and key file names in error message when GnuTLS fails
2564 to set them up, because the GnuTLS error message doesn't include the name
2565 of the failing file when there is a problem reading it.
2567 42. Allow both -bf and -bF in the same test run.
2569 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2571 44. The "Exiscan patch" is now merged into the mainline Exim source.
2573 45. Sometimes the final signoff response after QUIT could fail to get
2574 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2575 < 0 before doing a fflush(). This bug looks as though it goes back to the
2576 introduction of TLS in release 3.20, but "sometimes" must have been rare
2577 because the tests only now provoked it.
2579 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2580 (this can affect the format of dates).
2582 47. exim_tidydb, when checking for the continued existence of a message for
2583 which it has found a message-specific retry record, was not finding
2584 messages that were in split spool directories. Consequently, it was
2585 deleting retry records that should have stayed in existence.
2587 48. Steve fixed some bugs in eximstats.
2589 49. The SPA authentication driver was not abandoning authentication and moving
2590 on to the next authenticator when an expansion was forced to fail,
2591 contradicting the general specification for all authenticators. Instead it
2592 was generating a temporary error. It now behaves as specified.
2594 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2595 (the order specifies the preference for clients). The order is now AES256,
2596 AES128, 3DES, ARCFOUR128.
2598 51. Small patch to Sieve code - explicitly set From: when generating an
2601 52. Exim crashed if a remote delivery caused a very long error message to be
2602 recorded - for instance if somebody sent an entire SpamAssassin report back
2603 as a large number of 550 error lines. This bug was coincidentally fixed by
2604 increasing the size of one of Exim's internal buffers (big_buffer) that
2605 happened as part of the Exiscan merge. However, to be on the safe side, I
2606 have made the code more robust (and fixed the comments that describe what
2609 53. Now that there can be additional text after "Completed" in log lines (if
2610 the queue_time_overall log selector is set), a one-byte patch to exigrep
2611 was needed to allow it to recognize "Completed" as not the last thing in
2614 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2615 patch that reportedly fixes this has been added. I am not expert enough to
2616 create a test for it. This is what the patch creator wrote:
2618 "I found a little strange behaviour of ldap code when working with
2619 Windows 2003 AD Domain, where users was placed in more than one
2620 Organization Units. When I tried to give exim partial DN, the exit code
2621 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2622 But simultaneously result of request was absolutely normal ldap result,
2623 so I produce this patch..."
2625 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2626 REFERENCE, so I have modified the code to exclude the patch when that macro
2629 55. Some experimental protocols are using DNS PTR records for new purposes. The
2630 keys for these records are domain names, not reversed IP addresses. The
2631 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2632 leaves it alone. Component reversal etc. now happens only for IP addresses.
2635 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2637 57. Double the size of the debug message buffer (to 2048) so that more of very
2638 long debug lines gets shown.
2640 58. The exicyclog utility now does better if the number of log files to keep
2641 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2643 59. Two changes related to the smtp_active_hostname option:
2645 (1) $smtp_active_hostname is now available as a variable.
2646 (2) The default for smtp_banner uses $smtp_active_hostname instead
2647 of $primary_hostname.
2649 60. The host_aton() function is supposed to be passed a string that is known
2650 to be a valid IP address. However, in the case of IPv6 addresses, it was
2651 not checking this. This is a hostage to fortune. Exim now panics and dies
2652 if the condition is not met. A case was found where this could be provoked
2653 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2654 components; fortuitously, this particular loophole had already been fixed
2655 by change 4.50/55 above.
2657 If there are any other similar loopholes, the new check in host_aton()
2658 itself should stop them being exploited. The report I received stated that
2659 data on the command line could provoke the exploit when Exim was running as
2660 exim, but did not say which command line option was involved. All I could
2661 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2662 running as the user.
2665 61. There was a buffer overflow vulnerability in the SPA authentication code
2666 (which came originally from the Samba project). I have added a test to the
2667 spa_base64_to_bits() function which I hope fixes it.
2670 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2671 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2673 63. The daemon start-up calls getloadavg() while still root for those OS that
2674 need the first call to be done as root, but it missed one case: when
2675 deliver_queue_load_max is set with deliver_drop_privilege. This is
2676 necessary for the benefit of the queue runner, because there is no re-exec
2677 when deliver_drop_privilege is set.
2679 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2680 This has been fixed.
2682 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2683 was in use, was not putting the data itself into the right store pool;
2684 consequently, it could be overwritten for a subsequent message in the same
2685 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2688 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2690 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2691 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2692 the function were treating the return as a boolean value, which happened to
2693 work because 0=false and not-0=true, but is not correct code.
2695 68. The host_aton() function was not handling scoped IPv6 addresses (those
2696 with, for example, "%eth0" on the end) correctly.
2698 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2699 negated items (that is, ~something) in unsigned ints. Some compilers
2700 apparently mutter when there is no cast.
2702 70. If an address verification called from an ACL failed, and did not produce a
2703 user-specific message (i.e. there was only a "system" message), nothing was
2704 put in $acl_verify_message. In this situation, it now puts the system
2707 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2708 session; change 4.31/43 added the unwanted input to the log line - except
2709 that it did not do this in the start of session case. It now does.
2711 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2712 This is wrong and can cause the other end to generate a synchronization
2713 error if it is another Exim or anything else that does the synchronization
2714 check. A QUIT command is no longer sent after a timeout.
2716 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2717 during host lookups.
2719 74. The defer_ok option of callout verification was not working if it was used
2720 when verifying addresses in header lines, that is, for this case:
2722 verify = header_sender/callout=defer_ok
2724 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2725 those file descriptors could be used for SMTP connections. If anything
2726 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2727 could be sent to the SMTP client, causing chaos. The daemon now opens
2728 stdin, stdout, and stderr to /dev/null when it puts itself into the
2731 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2732 log by default. The user can override this with suitable Perl magic.
2734 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2735 the log message when discard triggers, was not working for the DATA ACL or
2736 for the non-SMTP ACL.
2738 78. Error message wording change in sieve.c.
2740 79. If smtp_accept_max_per_host was set, the number of connections could be
2741 restricted to fewer than expected, because the daemon was trying to set up
2742 a new connection before checking whether the processes handling previous
2743 connections had finished. The check for completed processes is now done
2744 earlier. On busy systems, this bug wouldn't be noticed because something
2745 else would have woken the daemon, and it would have reaped the completed
2748 80. If a message was submitted locally by a user whose login name contained one
2749 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2750 It caused a spool format error. I have fixed the spool reading code. A
2751 related problem was that the "from" clause in the Received: line became
2752 illegal because of the space(s). It is now covered by ${quote_local_part.
2754 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2757 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2759 83. Patch from Sieve maintainer to fix unterminated string problem in
2760 "vacation" handling.
2762 84. Some minor changes to the Linux configuration files to help with other
2763 OS variants using glibc.
2765 85. One more patch for Sieve to update vacation handling to latest spec.
2768 ----------------------------------------------------
2769 See the note above about the 4.44 and 4.50 releases.
2770 ----------------------------------------------------
2776 1. Change 4.43/35 introduced a bug that caused file counts to be
2777 incorrectly computed when quota_filecount was set in an appendfile
2780 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2781 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2784 3. Give more explanation in the error message when the command for a transport
2785 filter fails to execute.
2787 4. There are several places where Exim runs a non-Exim command in a
2788 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2789 was being done only for the command run by the queryprogram router. It is
2790 now done for all such subprocesses. The other cases are: ${run, transport
2791 filters, and the commands run by the lmtp and pipe transports.
2793 5. Some older OS have a limit of 256 on the maximum number of file
2794 descriptors. Exim was using setrlimit() to set 1000 as a large value
2795 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2796 systems. I've change it so that if it can't get 1000, it tries for 256.
2798 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2799 was an oversight, and furthermore, ever since the addition of extra
2800 controls (e.g. 4.43/32), the checks on when to allow different forms of
2801 "control" were broken. There should now be diagnostics for all cases when a
2802 control that does not make sense is encountered.
2804 7. $recipients is now available in the predata ACL (oversight).
2806 8. Tidy the search cache before the fork to do a delivery from a message
2807 received from the command line. Otherwise the child will trigger a lookup
2808 failure and thereby defer the delivery if it tries to use (for example) a
2809 cached ldap connection that the parent has called unbind on.
2811 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2812 of $address_data from the recipient verification was clobbered by the
2813 sender verification.
2815 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2816 was its contents. (It was OK if the option was not defined at all.)
2818 11. A "Completed" log line is now written for messages that are removed from
2819 the spool by the -Mrm option.
2821 12. $host_address is now set to the target address during the checking of
2822 ignore_target_hosts.
2824 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2825 being passed; this would have caused $sender_host_name to have been used if
2826 matching the list had actually called for a host name (not very likely,
2827 since this list is usually IP addresses). A host name is now passed as
2830 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2831 code that specifies a non-restarting handler (typically sigaction() in
2832 modern systems) in an attempt to fix a rare and obscure crash bug.
2834 15. Narrowed the window for a race in the daemon that could cause it to ignore
2835 SIGCHLD signals. This is not a major problem, because they are used only to
2836 wake it up if nothing else does.
2838 16. A malformed maildirsize file could cause Exim to calculate negative values
2839 for the mailbox size or file count. Odd effects could occur as a result.
2840 The maildirsize information is now recalculated if the size or filecount
2843 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2844 support for a long time. Removed HAVE_SYS_VFS_H.
2846 18. Updated exipick to current release from John Jetmore.
2848 19. Allow an empty sender to be matched against a lookup in an address list.
2849 Previously the only cases considered were a regular expression, or an
2852 20. Exim went into a mad DNS lookup loop when doing a callout where the
2853 host was specified on the transport, if the DNS lookup yielded more than
2856 21. The RFC2047 encoding function was originally intended for short strings
2857 such as real names; it was not keeping to the 75-character limit for
2858 encoded words that the RFC imposes. It now respects the limit, and
2859 generates multiple encoded words if necessary. To be on the safe side, I
2860 have increased the buffer size for the ${rfc2047: expansion operator from
2863 22. Failure to deliver a bounce message always caused it to be frozen, even if
2864 there was an errors_to setting on the router. The errors_to setting is now
2867 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2868 canonical form (fully expanded) before being placed in
2869 $sender_host_address.
2871 24. Updated eximstats to version 1.33
2873 25. Include certificate and key file names in error message when GnuTLS fails
2874 to set them up, because the GnuTLS error message doesn't include the name
2875 of the failing file when there is a problem reading it.
2877 26. Expand error message when OpenSSL has problems setting up cert/key files.
2880 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2881 (this can affect the format of dates).
2883 28. exim_tidydb, when checking for the continued existence of a message for
2884 which it has found a message-specific retry record, was not finding
2885 messages that were in split spool directories. Consequently, it was
2886 deleting retry records that should have stayed in existence.
2888 29. eximstats updated to version 1.35
2889 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2890 1.35 - bugfix such that pie charts by volume are generated correctly
2892 30. The SPA authentication driver was not abandoning authentication and moving
2893 on to the next authenticator when an expansion was forced to fail,
2894 contradicting the general specification for all authenticators. Instead it
2895 was generating a temporary error. It now behaves as specified.
2897 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2898 (the order specifies the preference for clients). The order is now AES256,
2899 AES128, 3DES, ARCFOUR128.
2901 31. Small patch to Sieve code - explicitly set From: when generating an
2904 32. Exim crashed if a remote delivery caused a very long error message to be
2905 recorded - for instance if somebody sent an entire SpamAssassin report back
2906 as a large number of 550 error lines. This bug was coincidentally fixed by
2907 increasing the size of one of Exim's internal buffers (big_buffer) that
2908 happened as part of the Exiscan merge. However, to be on the safe side, I
2909 have made the code more robust (and fixed the comments that describe what
2912 33. Some experimental protocols are using DNS PTR records for new purposes. The
2913 keys for these records are domain names, not reversed IP addresses. The
2914 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2915 leaves it alone. Component reversal etc. now happens only for IP addresses.
2918 34. The host_aton() function is supposed to be passed a string that is known
2919 to be a valid IP address. However, in the case of IPv6 addresses, it was
2920 not checking this. This is a hostage to fortune. Exim now panics and dies
2921 if the condition is not met. A case was found where this could be provoked
2922 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2923 components; fortuitously, this particular loophole had already been fixed
2924 by change 4.50/55 or 4.44/33 above.
2926 If there are any other similar loopholes, the new check in host_aton()
2927 itself should stop them being exploited. The report I received stated that
2928 data on the command line could provoke the exploit when Exim was running as
2929 exim, but did not say which command line option was involved. All I could
2930 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2931 running as the user.
2934 35. There was a buffer overflow vulnerability in the SPA authentication code
2935 (which came originally from the Samba project). I have added a test to the
2936 spa_base64_to_bits() function which I hope fixes it.
2939 36. The daemon start-up calls getloadavg() while still root for those OS that
2940 need the first call to be done as root, but it missed one case: when
2941 deliver_queue_load_max is set with deliver_drop_privilege. This is
2942 necessary for the benefit of the queue runner, because there is no re-exec
2943 when deliver_drop_privilege is set.
2945 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2946 was in use, was not putting the data itself into the right store pool;
2947 consequently, it could be overwritten for a subsequent message in the same
2948 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2951 38. Sometimes the final signoff response after QUIT could fail to get
2952 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2953 < 0 before doing a fflush(). This bug looks as though it goes back to the
2954 introduction of TLS in release 3.20, but "sometimes" must have been rare
2955 because the tests only now provoked it.
2961 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2962 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2963 Now it returns an integer. A number of calls were still expecting a T/F
2964 return. Fortuitously, in all cases, the tests worked in OK situations,
2965 which is the norm. However, things would have gone wrong on any write
2966 failures on the smtp file descriptor. This function is used when sending
2967 messages over SMTP and also when doing verify callouts.
2969 2. When Exim is called to do synchronous delivery of a locally submitted
2970 message (the -odf or -odi options), it no longer closes stderr before doing
2973 3. Implemented the mua_wrapper option.
2975 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2977 5. Implemented the functions header_remove(), header_testname(),
2978 header_add_at_position(), and receive_remove_recipient(), and exported them
2981 6. If an ACL "warn" statement specified the addition of headers, Exim already
2982 inserted X-ACL-Warn: at the start if there was no header name. However, it
2983 was not making this test for the second and subsequent header lines if
2984 there were newlines in the string. This meant that an invalid header could
2985 be inserted if Exim was badly configured.
2987 7. Allow an ACL "warn" statement to add header lines at the start or after all
2988 the Received: headers, as well as at the end.
2990 8. Added the rcpt_4xx retry error code.
2992 9. Added postmaster_mailfrom=xxx to callout verification option.
2994 10. Added mailfrom=xxxx to the callout verification option, for verify=
2997 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2998 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3000 12. Inserted some casts to stop certain compilers warning when using pointer
3001 differences as field lengths or precisions in printf-type calls (mostly
3002 affecting debugging statements).
3004 13. Added optional readline() support for -be (dynamically loaded).
3006 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3007 same clock tick as a message's arrival, so that its received time was the
3008 same as the "first fail" time on the retry record, and that message
3009 remained on the queue past the ultimate address timeout, every queue runner
3010 would try a delivery (because it was past the ultimate address timeout) but
3011 after another failure, the ultimate address timeout, which should have then
3012 bounced the address, did not kick in. This was a "< instead of <=" error;
3013 in most cases the first failure would have been in the next clock tick
3014 after the received time, and all would be well.
3016 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3017 being recognized when the domain list was tested by the match_domain
3018 condition in an expansion string.
3020 16. Added the ${str2b64: operator.
3022 17. Exim was always calling setrlimit() to set a large limit for the number of
3023 processes, without checking whether the existing limit was already
3024 adequate. (It did check for the limit on file descriptors.) Furthermore,
3025 errors from getrlimit() and setrlimit() were being ignored. Now they are
3026 logged to the main and panic logs, but Exim does carry on, to try to do its
3027 job under whatever limits there are.
3029 18. Imported PCRE 5.0.
3031 19. Trivial typo in log message " temporarily refused connection" (the leading
3034 20. If the log selector return_path_on_delivery was set and an address was
3035 redirected to /dev/null, the delivery process crashed because it assumed
3036 that a return path would always be set for a "successful" delivery. In this
3037 case, the whole delivery is bypassed as an optimization, and therefore no
3040 21. Internal re-arrangement: the function for sending a challenge and reading
3041 a response while authentication was assuming a zero-terminated challenge
3042 string. It's now changed to take a pointer and a length, to allow for
3043 binary data in such strings.
3045 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3047 23. Exim was not respecting finduser_retries when seeking the login of the
3048 uid under which it was called; it was always trying 10 times. (The default
3049 setting of finduser_retries is zero.) Also, it was sleeping after the final
3050 failure, which is pointless.
3052 24. Implemented tls_on_connect_ports.
3054 25. Implemented acl_smtp_predata.
3056 26. If the domain in control=submission is set empty, Exim assumes that the
3057 authenticated id is a complete email address when it generates From: or
3058 Sender: header lines.
3060 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3061 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3062 chown and chgrp in /bin and hostname in /usr/bin.
3064 28. Exim was keeping the "process log" file open after each use, just as it
3065 does for the main log. This opens the possibility of it remaining open for
3066 long periods when the USR1 signal hits a daemon. Occasional processlog
3067 errors were reported, that could have been caused by this. Anyway, it seems
3068 much more sensible not to leave this file open at all, so that is what now
3071 29. The long-running daemon process does not normally write to the log once it
3072 has entered its main loop, and it closes the log before doing so. This is
3073 so that log files can straightforwardly be renamed and moved. However,
3074 there are a couple of unusual error situations where the daemon does write
3075 log entries, and I had neglected to close the log afterwards.
3077 30. The text of an SMTP error response that was received during a remote
3078 delivery was being truncated at 512 bytes. This is too short for some of
3079 the long messages that one sometimes sees. I've increased the limit to
3082 31. It is now possible to make retry rules that apply only when a message has a
3083 specific sender, in particular, an empty sender.
3085 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3086 it possible to be selective about when SMTP synchronization is enforced.
3088 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3090 32. Implemented hosts_connection_nolog.
3092 33. Added an ACL for QUIT.
3094 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3097 35. Added mailbox_size and mailbox_filecount to appendfile.
3099 36. Added control = no_multiline_responses to ACLs.
3101 37. There was a bug in the logic of the code that waits for the clock to tick
3102 in the case where the clock went backwards by a substantial amount such
3103 that the microsecond fraction of "now" was more than the microsecond
3104 fraction of "then" (but the whole seconds number was less).
3106 38. Added support for the libradius Radius client library this is found on
3107 FreeBSD (previously only the radiusclient library was supported).
3113 1. When certain lookups returned multiple values in the form name=value, the
3114 quoting of the values was not always being done properly. Specifically:
3115 (a) If the value started with a double quote, but contained no whitespace,
3117 (b) If the value contained whitespace other than a space character (i.e.
3118 tabs or newlines or carriage returns) it was not quoted.
3119 This fix has been applied to the mysql and pgsql lookups by writing a
3120 separate quoting function and calling it from the lookup code. The fix
3121 should probably also be applied to nisplus, ibase and oracle lookups, but
3122 since I cannot test any of those, I have not disturbed their existing code.
3124 2. A hit in the callout cache for a specific address caused a log line with no
3125 reason for rejecting RCPT. Now it says "Previous (cached) callout
3126 verification failure".
3128 3. There was an off-by-one bug in the queryprogram router. An over-long
3129 return line was truncated at 256 instead of 255 characters, thereby
3130 overflowing its buffer with the terminating zero. As well as fixing this, I
3131 have increased the buffer size to 1024 (and made a note to document this).
3133 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3134 when Exim is waiting for an SMTP response from a remote server, Exim
3135 restarts its select() call on the socket, thereby resetting its timeout.
3136 This is not a problem when such interrupts are rare. Somebody set up a cron
3137 job to run exiwhat every 2 minutes, which is less than the normal select()
3138 timeout (5 or 10 minutes). This meant that the select() timeout never
3139 kicked in because it was always reset. I have fixed this by comparing the
3140 time when an interrupt arrives with the time at the start of the first call
3141 to select(). If more time than the timeout has elapsed, the interrupt is
3142 treated as a timeout.
3144 5. Some internal re-factoring in preparation for the addition of Sieve
3145 extensions (by MH). In particular, the "personal" test is moved to a
3146 separate function, and given an option for scanning Cc: and Bcc: (which is
3147 not set for Exim filters).
3149 6. When Exim created an email address using the login of the caller as the
3150 local part (e.g. when creating a From: or Sender: header line), it was not
3151 quoting the local part when it contained special characters such as @.
3153 7. Installed new OpenBSD configuration files.
3155 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3156 try to make them clearer.
3158 9. Callout options, other than the timeout value, were being ignored when
3159 verifying sender addresses in header lines. For example, when using
3161 verify = header_sender/callout=no_cache
3163 the cache was (incorrectly) being used.
3165 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3166 only the Cygwin environment.
3168 11. When return_path_on_delivery was set as a log selector, if different remote
3169 addresses in the same message used different return paths and parallel
3170 remote delivery occurred, the wrong values would sometimes be logged.
3171 (Whenever a remote delivery process finished, the return path value from
3172 the most recently started remote delivery process was logged.)
3174 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3175 lines when AUTH and/or TLS are in use. This is the "received protocol"
3176 field. Exim used to use "asmtp" for authenticated SMTP, without any
3177 indication (in the protocol name) for TLS use. Now it follows the RFC and
3178 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3179 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3180 names appear in log lines as well as in Received: header lines.
3182 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3183 extensions, and comparison tests, and to fix some bugs.
3185 14. Changes to the "personal" filter test:
3187 (1) The test was buggy in that it was just doing the equivalent of
3188 "contains" tests on header lines. For example, if a user's address was
3189 anne@some.where, the "personal" test would incorrectly be true for
3191 To: susanne@some.where
3193 This test is now done by extracting each address from the header in turn,
3194 and checking the entire address. Other tests that are part of "personal"
3195 are now done using regular expressions (for example, to check local parts
3196 of addresses in From: header lines).
3198 (2) The list of non-personal local parts in From: addresses has been
3199 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3200 taken from the Sieve specification recommendations.
3202 (3) If the message contains any header line starting with "List-" it is
3203 treated as non-personal.
3205 (4) The test for "circular" in the Subject: header line has been removed
3206 because it now seems ill-conceived.
3208 15. Minor typos in src/EDITME comments corrected.
3210 16. Installed latest exipick from John Jetmore.
3212 17. If headers_add on a router specified a text string that was too long for
3213 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3214 of string_sprintf() is now avoided.
3216 18. $message_body_size was not set (it was always zero) when running the DATA
3217 ACL and the local_scan() function.
3219 19. For the "mail" command in an Exim filter, no default was being set for
3220 the once_repeat time, causing a random time value to be used if "once" was
3221 specified. (If the value happened to be <= 0, no repeat happened.) The
3222 default is now 0s, meaning "never repeat". The "vacation" command was OK
3223 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3224 (I found it when inspecting the code).
3226 20. There is now an overall timeout for performing a callout verification. It
3227 defaults to 4 times the callout timeout, which applies to individual SMTP
3228 commands during the callout. The overall timeout applies when there is more
3229 than one host that can be tried. The timeout is checked before trying the
3230 next host. This prevents very long delays if there are a large number of
3231 hosts and all are timing out (e.g. when the network connections are timing
3232 out). The value of the overall timeout can be changed by specifying an
3233 additional sub-option for "callout", called "maxwait". For example:
3235 verify = sender/callout=5s,maxwait=20s
3237 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3238 to the end before writing, but this should make it even safer).
3240 22. Exim was forgetting that it had advertised PIPELINING for the second and
3241 subsequent messages on an SMTP connection. It was also not resetting its
3242 memory on STARTTLS and an internal HELO.
3244 23. When Exim logs an SMTP synchronization error within a session, it now
3245 records whether PIPELINING has been advertised or not.
3247 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3248 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3249 rather than long int.
3251 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3253 26. Added the never_mail option to autoreply.
3259 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3260 crash if the getsockname() call failed; this can happen if a connection is
3261 closed very soon after it is established. The problem was simply in the
3262 order in which certain operations were done, causing Exim to try to write
3263 to the SMTP stream before it had set up the file descriptor. The bug has
3264 been fixed by making things happen in the correct order.
3270 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3271 before the connection was closed, thus losing the rejection response.
3273 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3274 some early Solaris releases, but causes trouble in current releases where
3275 socklen_t is defined.
3277 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3280 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3283 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3284 that releasing the top of it at the end releases what was used for sub-
3285 expansions (unless the block got too big). However, discard this block if
3286 the first thing is a variable or header, so that we can use its block when
3287 it is dynamic (useful for very large $message_headers, for example).
3289 6. Lookups now cache *every* query, not just the most recent. A new, separate
3290 store pool is used for this. It can be recovered when all lookup caches are
3291 flushed. Lookups now release memory at the end of their result strings.
3292 This has involved some general refactoring of the lookup sources.
3294 7. Some code has been added to the store_xxx() functions to reduce the amount
3295 of flapping under certain conditions.
3297 8. log_incoming_interface used to affect only the <= reception log lines. Now
3298 it causes the local interface and port to be added to several more SMTP log
3299 lines, for example "SMTP connection from", and rejection lines.
3301 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3303 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3305 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3306 could be overwritten at the end of the current message (or the start of a
3307 new message if it was set in a HELO ACL). The value is now preserved for
3308 the duration of the SMTP connection.
3310 12. If a transport had a headers_rewrite setting, and a matching header line
3311 contained an unqualified address, that address was qualified, even if it
3312 did not match any rewriting rules. The underlying bug was that the values
3313 of the flags that permit the existence of unqualified sender and recipient
3314 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3315 non-local messages, and by -bnq for local messages) were not being
3316 preserved with the message after it was received.
3318 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3319 "next input=" as part of the text comprising the host identity instead of
3320 the correct text. The code was using the same buffer for two different
3321 strings. However, depending on which order the printing function evaluated
3322 its arguments, the bug did not always show up. Under Linux, for example, my
3323 test suite worked just fine.
3325 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3326 doesn't work with some older versions of Perl. It has been changed to "my",
3327 which in any case is probably the better facility to use.
3329 15. A really picky compiler found some instances of statements for creating
3330 error messages that either had too many or two few arguments for the format
3333 16. The size of the buffer for calls to the DNS resolver has been increased
3334 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3335 for addresses that have a lot of PTR records. This alleviates a problem; it
3336 does not fully solve it.
3338 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3339 buffer now truncates the list and logs the incident, which is the same
3340 action as happens when Exim is looking up a host name and its aliases.
3341 Previously in this situation something unpredictable would happen;
3342 sometimes it was "internal error: store_reset failed".
3344 18. If a server dropped the connection unexpectedly when an Exim client was
3345 using GnuTLS and trying to read a response, the client delivery process
3346 crashed while trying to generate an error log message.
3348 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3349 string, for example:
3351 warn message = H1: something\nH2: something
3353 the text was added as a single header line from Exim's point of view
3354 though it ended up OK in the delivered message. However, searching for the
3355 second and subsequent header lines using $h_h2: did not work. This has been
3356 fixed. Similarly, if a system filter added multiple headers in this way,
3357 the routers could not see them.
3359 20. Expanded the error message when iplsearch is called with an invalid key to
3360 suggest using net-iplsearch in a host list.
3362 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3363 ACLs are no longer actually imposed (and a message to that effect is
3366 22. If a "gecos" field in a passwd entry contained escaped characters, in
3367 particular, if it contained a \" sequence, Exim got it wrong when building
3368 a From: or a Sender: header from that name. A second bug also caused
3369 incorrect handling when an unquoted " was present following a character
3370 that needed quoting.
3372 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3373 was not being matched caselessly.
3375 24. Arranged for all hyphens in the exim.8 source to be escaped with
3378 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3379 properly. Recipient callout cache records were still being keyed to include
3380 the sender, even when use_sender was set false. This led to far more
3381 callouts that were necessary. The sender is no longer included in the key
3382 when use_sender is false.
3384 26. Added "control = submission" modifier to ACLs.
3386 27. Added the ${base62d: operator to decode base 62 numbers.
3388 28. dnsdb lookups can now access SRV records.
3390 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3391 the configuration file.
3393 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3394 (-v) mode. This makes the output for a verbose queue run more intelligible.
3396 31. Added a use_postmaster feature to recipient callouts.
3398 32. Added the $body_zerocount variable, containing the number of binary zero
3399 bytes in the message body.
3401 33. The time of last modification of the "new" subdirectory is now used as the
3402 "mailbox time last read" when there is a quota error for a maildir
3405 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3407 35. Added +ignore_unknown as a special item in host lists.
3409 36. Code for decoding IPv6 addresses in host lists is now included, even if
3410 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3411 address was recognized as an IP address, but was then not correctly decoded
3412 into binary, causing unexpected and incorrect effects when compared with
3419 1. Very minor rewording of debugging text in manualroute to say "list of
3420 hosts" instead of "hostlist".
3422 2. If verify=header_syntax was set, and a header line with an unqualified
3423 address (no domain) and a large number of spaces between the end of the
3424 name and the colon was received, the reception process suffered a buffer
3425 overflow, and (when I tested it) crashed. This was caused by some obsolete
3426 code that should have been removed. The fix is to remove it!
3428 3. When running in the test harness, delay a bit after writing a bounce
3429 message to get a bit more predictability in the log output.
3431 4. Added a call to search_tidyup() just before forking a reception process. In
3432 theory, someone could use a lookup in the expansion of smtp_accept_max_
3433 per_host which, without the tidyup, could leave open a database connection.
3435 5. Added the variables $recipient_data and $sender_data which get set from a
3436 lookup success in an ACL "recipients" or "senders" condition, or a router
3437 "senders" option, similar to $domain_data and $local_part_data.
3439 6. Moved the writing of debug_print from before to after the "senders" test
3442 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3443 problems for message scanning, either using a data ACL, or using
3444 local_scan() because the Received: header was not generated till after they
3445 were called (in order to set the time as the time of reception completion).
3446 I have revised the way this works. The header is now generated after the
3447 body is received, but before the ACL or local_scan() are called. After they
3448 are run, the timestamp in the header is updated.
3454 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3455 before starting a queue runner without re-exec. This happened only when
3456 deliver_drop_privilege was set or when the Exim user was set to root. The
3457 effect of the bug was that timeouts during subsequent deliveries caused
3458 crashes instead of being properly handled. The handler is now left at its
3459 default (and expected) setting.
3461 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3462 message, again when deliver_drop_privilege is set or Exim is run as root.
3463 The bug described in (1) was not present in this case, but the tidying up
3464 of the other signals was missing. I have made the two cases consistent.
3466 3. The ignore_target_hosts setting on a manualroute router was being ignored
3467 for hosts that were looked up using the /MX notation.
3469 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3472 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3473 operated on the sender address. After changing the $sender_address to <>
3474 for the sender address verify, Exim was re-instated it as the original
3475 (before rewriting) address, but remembering that it had rewritten it, so it
3476 wasn't rewriting it again. This bug also had the effect of breaking the
3477 sender address verification caching when the sender address was rewritten.
3479 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3480 This has been changed so that if the ip literal address matches
3481 ignore_target_hosts, the router declines.
3483 7. Added expansion conditions match_domain, match_address, and match_local_
3484 part (NOT match_host).
3486 8. The placeholder for the Received: header didn't have a length field set.
3488 9. Added code to Exim itself and to exim_lock to test for a specific race
3489 condition that could lead to file corruption when using MBX delivery. The
3490 issue is with the lockfile that is created in /tmp. If this file is removed
3491 after a process has opened it but before that process has acquired a lock,
3492 there is the potential for a second process to recreate the file and also
3493 acquire a lock. This could lead to two Exim processes writing to the file
3494 at the same time. The added code performs the same test as UW imapd; it
3495 checks after acquiring the lock that its file descriptor still refers to
3496 the same named file.
3498 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3499 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3502 11. Added the smtp_active_hostname option. If used, this will typically be made
3503 to depend on the incoming interface address. Because $interface_address is
3504 not set up until the daemon has forked a reception process, error responses
3505 that can happen earlier (such as "too many connections") no longer contain
3508 12. If an expansion in a condition on a "warn" statement fails because a lookup
3509 defers, the "warn" statement is abandoned, and the next ACL statement is
3510 processed. Previously this caused the whole ACL to be aborted.
3512 13. Added the iplsearch lookup type.
3514 14. Added ident_timeout as a log selector.
3516 15. Added tls_certificate_verified as a log selector.
3518 16. Added a global option tls_require_ciphers (compare the smtp transport
3519 option of the same name). This controls incoming TLS connections.
3521 17. I finally figured out how to make tls_require_ciphers do a similar thing
3522 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3523 before starting the TLS session.
3525 18. Tabs are now shown as \t in -bP output.
3527 19. If the log selector return_path_on_delivery was set, Exim crashed when
3528 bouncing a message because it had too many Received: header lines.
3530 20. If two routers both had headers_remove settings, and the first one included
3531 a superfluous trailing colon, the final name in the first list and the
3532 first name in the second list were incorrectly joined into one item (with a
3533 colon in the middle).
3539 1. Added -C and -D options to the exinext utility, mainly to make it easier
3540 to include in the automated testing, but these could be helpful when
3541 multiple configurations are in use.
3543 2. The exinext utility was not formatting the output nicely when there was
3544 an alternate port involved in the retry record key, nor when there was a
3545 message id as well (for retries that were specific to a specific message
3546 and a specific host). It was also confused by IPv6 addresses, because of
3547 the additional colons they contain. I have fixed the IPv4 problem, and
3548 patched it up to do a reasonable job for IPv6.
3550 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3551 delivery, the log line now contains "pipelined" if PIPELINING was used.
3553 4. An SMTP transport process used to panic and die if the bind() call to set
3554 an explicit outgoing interface failed. This has been changed; it is now
3555 treated in the same way as a connect() failure.
3557 5. A reference to $sender_host_name in the part of a conditional expansion
3558 that was being skipped was still causing a DNS lookup. This no longer
3561 6. The def: expansion condition was not recognizing references to header lines
3562 that used bh_ and bheader_.
3564 7. Added the _cache feature to named lists.
3566 8. The code for checking quota_filecount in the appendfile transport was
3567 allowing one more file than it should have been.
3569 9. For compatibility with Sendmail, the command line option
3577 and sets the incoming protocol and host name (for trusted callers). The
3578 host name and its colon can be omitted when only the protocol is to be set.
3579 Note the Exim already has two private options, -pd and -ps, that refer to
3580 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3581 "s", but I don't think that's a major issue.
3583 10. A number of refactoring changes to the code, none of which should affect
3586 (a) The number of logging options was getting close to filling up the
3587 32-bit word that was used as a bit map. I have split them into two classes:
3588 those that are passed in the argument to log_write(), and those that are
3589 only ever tested independently outside of that function. These are now in
3590 separate 32-bit words, so there is plenty of room for expansion again.
3591 There is no change in the user interface or the logging behaviour.
3593 (b) When building, for example, log lines, the code previously used a
3594 macro that called string_cat() twice, in order to add two strings. This is
3595 not really sufficiently general. Furthermore, there was one instance where
3596 it was actually wrong because one of the argument was used twice, and in
3597 one call a function was used. (As it happened, calling the function twice
3598 did not affect the overall behaviour.) The macro has been replaced by a
3599 function that can join an arbitrary number of extra strings onto a growing
3602 (c) The code for expansion conditions now uses a table and a binary chop
3603 instead of a serial search (which was left over from when there were very
3604 few conditions). Also, it now recognizes conditions like "pam" even when
3605 the relevant support is not compiled in: a suitably worded error message is
3606 given if an attempt is made to use such a condition.
3608 11. Added ${time_interval:xxxxx}.
3610 12. A bug was causing one of the ddress fields not to be passed back correctly
3611 from remote delivery subprocesses. The field in question was not being
3612 subsequently used, so this caused to problems in practice.
3614 13. Added new log selectors queue_time and deliver_time.
3616 14. Might have fixed a bug in maildirsizefile handling that threw up
3617 "unexpected character" debug warnings, and recalculated the data
3618 unnecessarily. In any case, I expanded the warning message to give more
3621 15. Added the message "Restricted characters in address" to the statements in
3622 the default ACL that block characters like @ and % in local parts.
3624 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3625 Three changes have been made:
3627 (a) There was a serious bug; a negative response to MAIL caused the whole
3628 recipient domain to be cached as invalid, thereby blocking all messages
3629 to all local parts at the same domain, from all senders. This bug has
3630 been fixed. The domain is no longer cached after a negative response to
3631 MAIL if the sender used is not empty.
3633 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3636 (c) A new callout option, "use_sender" has been added for people who want
3637 the modified behaviour.
3643 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3646 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3647 indeed breaks things for older releases.
3649 3. Added additional logging to the case where there is a problem reading data
3650 from a filter that is running in a subprocess using a pipe, in order to
3651 try to track down a specific problem.
3653 4. Testing facility fudge: when running in the test harness and attempting
3654 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3655 getting "No route to host". Convert this to a timeout.
3657 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3660 6. Some OS don't have socklen_t but use size_t instead. This affects the
3661 fifth argument of getsockopt() amongst other things. This is now
3662 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3663 can be set for individual OS. I have set it for SunOS5, OSF1, and
3664 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3665 some earlier ones do not.
3667 7. Change 4.30/15 was not doing the test caselessly.
3669 8. The standard form for an IPv6 address literal was being rejected by address
3670 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3671 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3672 this, as well as the form without the "IPv6" on the front (but only when
3673 address literals are enabled, of course).
3675 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3677 10. Exim crashed if a message with an empty sender address specified by -f
3678 encountered a router with an errors_to setting. This could be provoked only
3679 by a command such as
3683 where an empty string was supplied; "<>" did not hit this bug.
3685 11. Installed PCRE release 4.5.
3687 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3688 remained set. It is now erased.
3690 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3691 times from message ids (which are base 36 rather than the normal 62).
3693 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3694 were being counted as actual protocol errors, and logged if the log
3695 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3696 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3697 and DATA following a set of rejected RCPTs do not count as protocol errors.
3698 In other words, Exim assumes they were pipelined, though this may not
3699 actually be the case. Of course, in all cases the client gets an
3700 appropriate error code.
3702 15. If a lookup fails in an ACL condition, a message about the failure may
3703 be available; it is used if testing the ACL cannot continue, because most
3704 such messages specify what the cause of the deferral is. However, some
3705 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3706 that caused an old message to be retained and used if a later statement
3707 caused a defer, replacing the real cause of the deferral.
3709 16. If an IP address had so many PTR records that the DNS lookup buffer
3710 was not large enough to hold them, Exim could crash while trying to process
3711 the truncated data. It now detects and logs this case.
3713 17. Further to 4.21/58, another change has been made: if (and only if) the
3714 first line of a message (the first header line) ends with CRLF, a bare LF
3715 in a subsequent header line has a space inserted after it, so as not to
3716 terminate the header.
3718 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3719 unnecessarily, used atoi() instead of strtol(), and didn't check the
3720 termination when getting file sizes from file names by regex.
3722 19. Completely re-implemented the support for maildirsize files, in the light
3723 of a number of problems with the previous contributed implementation
3724 (4.30/29). In particular:
3726 . If the quota is zero, the maildirsize file is maintained, but no quota is
3729 . If the maildir directory does not exist, it is created before any attempt
3730 to write a maildirsize file.
3732 . The quota value in the file is just a cache; if the quota is changed in
3733 the transport, the new value overrides.
3735 . A regular expression is available for excluding directories from the
3738 20. The autoreply transport checks the characters in options that define the
3739 message's headers; it allows continued headers, but it was checking with
3740 isspace() after an embedded newline instead of explicitly looking for a
3743 21. If all the "regular" hosts to which an address was routed had passed their
3744 expiry times, and had not reached their retry times, the address was
3745 bounced, even if fallback hosts were defined. Now Exim should go on to try
3748 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3749 equivalent code in the SMTP transport. Some hosts send humungous responses
3750 to HELO/EHLO, more than 1024 it seems.
3752 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3753 gives compiler warnings in some environments. I've now done it "properly",
3756 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3757 (because of problems with the built-in one) was declared to return uschar *
3758 instead of char *, causing compiler failure.
3760 25. Fixed a file descriptor leak when processing alias/forward files.
3762 26. Fixed a minor format string issue in dbfn.c.
3764 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3766 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3767 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3770 29. When a sender address is verified, it is cached, to save repeating the test
3771 when there is more than one recipient in a message. However, when the
3772 verification involves a callout, it is possible for different callout
3773 options to be set for different recipients. It is too complicated to keep
3774 track of this in the cache, so now Exim always runs a verification when a
3775 callout is required, relying on the callout cache for the optimization.
3776 The overhead is duplication of the address routing, but this should not be
3779 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3780 to be verified with callout=postmaster, and the main callout worked but the
3781 postmaster check failed, the verification correctly failed. However, if a
3782 subsequent RCPT command asked for sender verification *without* the
3783 postmaster check, incorrect caching caused this verification also to fail,
3786 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3787 it was not caching the DNS options (qualify_single, search_parents) that
3788 were used when the lookup failed. A subsequent lookup with different
3789 options therefore always gave the same answer, though there were cases
3790 where it should not have. (Example: a "domains = !$mx_any" option on a
3791 dnslookup router: the "domains" option is always processed without any
3792 widening, but the router might have qualify_single set.) Now Exim uses the
3793 cached value only when the same options are set.
3795 32. Added John Jetmore's "exipick" utility to the distribution.
3797 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3798 than a timeout (e.g. a certificate is required, and is not provided), an
3799 Exim server now closes the connection immediately. Previously it waited for
3800 the client to close - but if the client is SSL, it seems that they each
3801 wait for each other, leading to a delay before one of them times out.
3803 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3804 maintained 0.8.x compatibility because I don't think many are using it, and
3805 it is clearly obsolete.
3807 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3810 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3811 client certificate was expired. A simple patch fixes this, though I don't
3812 understand the full logic of why the verify callback is called multiple
3815 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3816 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3817 which causes problems with some clients (such as the Certicom SSL Plus
3818 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3819 disables the coutermeasure allowing Eudora to connect."
3821 38. Exim was not checking that a write() to a log file succeeded. This could
3822 lead to Bad Things if a log got too big, in particular if it hit a file
3823 size limit. Exim now panics and dies if it cannot write to a log file, just
3824 as it does if it cannot open a log file.
3826 39. Modified OS/Makefile-Linux so that it now contains
3828 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3830 The two -D definitions ensure that Exim is compiled with large file
3831 support, which makes it possible to handle log files that are bigger than
3834 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3835 instance) a domain was checked against a named list that involved a lookup,
3836 causing $domain_data to be set, then another domain was checked against the
3837 same list, then the first domain was re-checked, the value of $domain_data
3838 after the final check could be wrong. In particular, if the second check
3839 failed, it could be set empty. This bug probably also applied to
3842 41. The strip_trailing_dot option was not being applied to the address given
3843 with the -f command-line option.
3845 42. The code for reading a message's header from the spool was incrementing
3846 $received_count, but never initializing it. This meant that the value was
3847 incorrect (doubled) while delivering a message in the same process in which
3848 it was received. In the most common configuration of Exim, this never
3849 happens - a fresh exec is done - but it can happen when
3850 deliver_drop_privilege is set.
3852 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3853 it now includes up to 150 characters of the unexpected data in the log
3856 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3857 and building data strings. The size of both of these buffers was 10 000
3858 bytes - far larger than anybody would *ever* want, thought I. Needless to
3859 say, somebody hit the limit. I have increased the maximum line length to
3860 20 000 and the maximum data length of concatenated lines to 100 000. I have
3861 also fixed two bugs, because there was no checking on these buffers. Tsk,
3862 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3863 buffer is too small.
3865 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3866 lsearch lookups. Now it does.
3868 46. When parsing a route_list item in a manualroute router, a fixed-length
3869 buffer was used for the list of hosts. I made this 1024 bytes long,
3870 thinking that nobody would ever have a list of hosts that long. Wrong.
3871 Somebody had a whole pile of complicated expansion conditions, and the
3872 string was silently truncated, leading to an expansion error. It turns out
3873 that it is easier to change to an unlimited length (owing to other changes
3874 that have happened since this code was originally written) than to build
3875 structure for giving a limitation error. The length of the item that
3876 expands into the list of hosts is now unlimited.
3878 47. The lsearch lookup could not handle data where the length of text line was
3879 more than 4095 characters. Such lines were truncated, leading to shortened
3880 data being returned. It should now handle lines of any length.
3882 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3883 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3886 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3887 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3888 no longer happens in comment lines. A list of replacements is now placed
3889 at the head of all of the source files, except those whose only change is
3890 to replace PERL_COMMAND in the very first #! line.
3892 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3893 messages on the queue, with a bottom-up merge sort, using code contributed
3894 by Michael Haardt. This should make operations like -bp somewhat faster on
3895 large queues. It won't affect queue runners, except when queue_run_in_order
3898 51. Installed eximstats 1.31 in the distribution.
3900 52. Added support for SRV lookups to the dnslookup router.
3902 53. If an ACL referred to $message_body or $message_body_end, the value was not
3903 reset for any messages that followed in the same SMTP session.
3905 54. The store-handling optimization for building very long strings was not
3906 differentiating between the different store pools. I don't think this
3907 actually made any difference in practice, but I've tidied it.
3909 55. While running the routers to verify a sender address, $sender_address
3910 was still set to the sender address. This is wrong, because when routing to
3911 send a bounce to the sender, it would be empty. Therefore, I have changed
3912 it so that, while verifying a sender address, $sender_address is set to <>.
3913 (There is no change to what happens when verifying a recipient address.)
3915 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3916 target A or AAAA records (if not already returned) without resetting the
3917 qualify_single or search_parents options of the DNS resolver. These are
3918 inappropriate in this case because the targets of MX and SRV records must
3919 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3920 target that, when qualified, matched something in the local domain. These
3921 two options are now turned off when doing these lookups.
3923 57. It seems that at least some releases of Reiserfs (which does not have the
3924 concept of a fixed number of inodes) returns zero and not -1 for the
3925 number of available inodes. This interacted badly with check_spool_inodes,
3926 which assumed that -1 was the "no such thing" setting. What I have done is
3927 to check that the total number of inodes is greater than zero before doing
3928 the test of how many are available.
3930 58. When a "warn" ACL statement has a log_message modifier, the message is
3931 remembered, and not repeated. This is to avoid a lot of repetition when a
3932 message has many recipients that cause the same warning to be written.
3933 Howewer, Exim was preserving the list of already written lines for an
3934 entire SMTP session, which doesn't seem right. The memory is now reset if a
3935 new message is started.
3937 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3938 debugging output unless log_rewrite was also set.
3940 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3941 of (int)(handle) when we know that handle contains (void *)(-1).
3943 61. The Exim daemon panic-logs an error return when it closes the incoming
3944 connection. However "connection reset by peer" seems to be common, and
3945 isn't really an error worthy of noting specially, so that particular error
3948 62. When Exim is trying to find all the local interfaces, it used to panic and
3949 die if the ioctl to get the interface flags failed. However, it seems that
3950 on at least one OS (Solaris 9) it is possible to have an interface that is
3951 included in the list of interfaces, but for which you get a failure error
3952 for this call. This happens when the interface is not "plumbed" into a
3953 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3954 failure of the "get flags" call assumes that the interface is down.
3956 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3957 makes life easier for people who are doing arithmetic on fields extracted
3958 from dates, where you often get leading zeros that should not be
3959 interpreted as octal.
3961 64. Added qualify_domain to the redirect router, to override the global
3964 65. If a pathologically long header line contained very many addresses (the
3965 report of this problem mentioned 10 000) and each of them was rewritten,
3966 Exim could use up a very large amount of memory. (It kept on making new
3967 copies of the header line as it rewrote, and never released the old ones.)
3968 At the expense of a bit more processing, the header rewriting function has
3969 been changed so that it no longer eats memory in this way.
3971 66. The generation of the Received: header has been moved from the time that a
3972 message starts to be received, to the time that it finishes. The timestamp
3973 in the Received: header should now be very close to that of the <= log
3974 line. There are two side-effects of this change:
3976 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3977 logged header lines no longer include the local Received: line, because
3978 it has not yet been created. The same applies to a copy of the message
3979 that is returned to a non-SMTP sender when a message is rejected.
3981 (b) When a filter file is tested using -bf, no additional Received: header
3982 is added to the test message. After some thought, I decided that this
3985 This change does not affect the value of $received_for. It is still set
3986 after address rewriting, but before local_scan() is called.
3988 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3990 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3991 gave an unhelpful panic error message, and a defer error. I have managed to
3992 change this behaviour so that it now rejects any supplied certificate,
3993 which seems right, as the list of acceptable certificates is empty.
3995 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3996 gave an unhelpful defer error. I have not managed to make this reject any
3997 supplied certificates, but the error message it gives is "no certificate
3998 supplied", which is not helpful.
4000 70. exigrep's output now also includes lines that are not associated with any
4001 message, but which match the given pattern. Implemented by a patch from
4002 Martin Sluka, which also tidied up the Perl a bit.
4004 71. Recipient callout verification, like sender verification, was using <> in
4005 the MAIL FROM command. This isn't really the right thing, since the actual
4006 sender may affect whether the remote host accepts the recipient or not. I
4007 have changed it to use the actual sender in the callout; this means that
4008 the cache record is now keyed on a recipient/sender pair, not just the
4009 recipient address. There doesn't seem to be a real danger of callout loops,
4010 since a callout by the remote host to check the sender would use <>.
4011 [SEE ABOVE: changed after hitting problems.]
4013 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4014 temporary errors. However, in the case of such a code being given after
4015 the end of a data transmission (i.e. after ".") Exim was failing to write
4016 a retry record for the message. (Yes, there was some broken host that was
4017 actually sending 8xx at this point.)
4019 73. An unknown lookup type in a host list could cause Exim to panic-die when
4020 the list was checked. (An example that provoked this was putting <; in the
4021 middle of a list instead of at the start.) If this happened during a DATA
4022 ACL check, a -D file could be left lying around. This kind of configuration
4023 error no longer causes Exim to die; instead it causes a defer errror. The
4024 incident is still logged to the main and panic logs.
4026 74. Buglet left over from Exim 3 conversion. The message "too many messages
4027 in one connection" was written to the rejectlog but not the mainlog, except
4028 when address rewriting (yes!) was being logged.
4030 75. Added write_rejectlog option.
4032 76. When a system filter was run not as root (that is, when system_filter_user
4033 was set), the values of the $n variables were not being returned to the
4034 main process; thus, they were not subsequently available in the $sn
4037 77. Added +return_path_on_delivery log selector.
4039 78. A connection timeout was being treated differently from recipients deferred
4040 when testing hosts_max_try with a message that was older than the host's
4041 retry timeout. (The host should not be counted, thus allowing all hosts to
4042 be tried at least once before bouncing.) This may have been the cause of an
4043 occasionally reported bug whereby a message would remain on the queue
4044 longer than the retry timeout, but would be bounced if a delivery was
4045 forced. I say "may" because I never totally pinned down the problem;
4046 setting up timeout/retry tests is difficult. See also the next item.
4048 79. The ultimate address timeout was not being applied to errors that involved
4049 a combination of host plus message (for example, a timeout on a MAIL
4050 command). When an address resolved to a number of possible hosts, and they
4051 were not all tried for each delivery (e.g. because of hosts_max_try), a
4052 message could remain on the queue longer than the retry timeout.
4054 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4057 81. Fixed an obscure SMTP outgoing bug which required at least the following
4058 conditions: (a) there was another message waiting for the same server;
4059 (b) the server returned 5xx to all RCPT commands in the first message so
4060 that the message was not completed; (c) the server dropped the connection
4061 or gave a negative response to the RSET that Exim sends to abort the
4062 transaction. The observed case was a dropped connection after DATA that had
4063 been sent in pipelining mode. That is, the server had advertised PIPELINING
4064 but was not implementing it correctly. The effect of the bug was incorrect
4065 behaviour, such as trying another host, and this could lead to a crash.
4071 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4072 and daemon.c were passed as pointers to ints; they should have been
4073 pointers to socklen_t variables (which are typically unsigned ints).
4075 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4078 3. Fixed a really odd bug that affected only the testing scheme; patching a
4079 certain fixed string in the binary changed the value of another string that
4080 happened to be identical to the end of the original first string.
4082 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4083 name", it returns that address as the IP address. On some operating
4084 systems (e.g. Solaris), it also passes back the IP address string as the
4085 "host name". However, on others (e.g. Linux), it passes back an empty
4086 string. Exim wasn't checking for this, and was changing the host name to an
4087 empty string, assuming it had been canonicized.
4089 5. Although rare, it is permitted to have more than one PTR record for a given
4090 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4091 all the names associated with an address, because they do in Solaris.
4092 However, it seems that they do not in Linux for data that comes from the
4093 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4094 I found this out when I moved to a new Linux workstation and tried to run
4095 the Exim test suite.
4097 To get round this problem I have changed the code so that it now does its
4098 own call to the DNS to look up PTR records when searching for a host name.
4099 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4100 addresses that are only in /etc/hosts are still found.
4102 This behaviour is, however, controlled by an option called host_lookup_
4103 order, which defaults to "bydns:byaddr". If people want to use the other
4104 order, or indeed, just use one or the other means of lookup, they can
4105 specify it in this variable.
4107 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4108 some operating systems, this comes back from gethostbyaddr() as an empty
4109 string, and this is what Exim used to test for. However, it seems that in
4110 other systems, "." is yielded. Exim now tests for this case too.
4112 7. The values of check_spool_space and check_log_space are now held internally
4113 as a number of kilobytes instead of an absolute number of bytes. If a
4114 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4115 kilobyte. This means that much larger values can be stored.
4117 8. Exim monitor: an attempt to get the action menu when not actually pointing
4118 at a message produces an empty menu entitled "No message selected". This
4119 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4120 no entries in it ("Shell widget menu has zero width and/or height"). So I
4121 have added a single, blank menu entry in this case.
4123 9. Added ${quote_local_part.
4125 10. MIME decoding is now applied to the contents of Subject: header lines when
4128 11. Now that a reference to $sender_host_address automatically causes a reverse
4129 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4130 host lookup before query-style lookups in lists that might use this
4131 variable. This has therefore been abolished, and the "net-" prefix is no
4132 longer necessary for query-style lookups.
4134 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4135 have been a typo for LFLAGS, so it has been changed.
4137 13. The install script calls Exim with "-C /dev/null" in order to find the
4138 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4139 to be output. Howeve, since Exim outputs its version number before the
4140 error, it didn't break the script. It just looked ugly. I fixed this by
4141 always allowing "-C /dev/null" if the caller is root.
4143 14. Ignore overlarge ACL variable number when reading spool file - insurance
4144 against a later release with more variables having written the file.
4146 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4147 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4148 this, as well as the form without the "IPv6" on the front.
4150 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4151 OS/Makefile-Darwin file.
4153 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4154 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4156 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4157 variable. However, some OS use ints for inodes. I've added cast to long int
4158 to get rid of the compiler warning.
4160 19. I had forgotten to lock out "/../" in configuration file names when
4161 ALT_CONFIG_PREFIX was set.
4163 20. Routers used for verification do not need to specify transports. However,
4164 if such a router generated a host list, and callout was configured, Exim
4165 crashed, because it could not find a port number from the (non-existent)
4166 transport. It now assumes port 25 in this circumstance.
4168 21. Added the -t option to exigrep.
4170 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4171 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4172 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4173 them is set, LOOKUP_LSEARCH is forced.
4175 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4177 24. Added sender and host information to the "rejected by local_scan()" log
4178 line; previously there was no indication of these.
4180 25. Added .include_if_exists.
4182 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4183 Linux. It turns out that not all file systems support this. Apparently some
4184 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4185 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4186 not supported on the file descriptor, is now ignored when Exim is trying to
4187 sync a directory. This applies only to Linux.
4189 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4191 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4192 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4195 29. Added support for maildirsize files from supplied patch (modified a bit).
4197 30. The use of :fail: followed by an empty string could lead Exim to respond to
4198 sender verification failures with (e.g.):
4200 550 Verification failed for <xxx>
4201 550 Sender verify failed
4203 where the first response line was missing the '-' that indicates it is not
4204 the final line of the response.
4206 31. The loop for finding the name of the user that called Exim had a hardwired
4207 limit of 10; it now uses the value of finduser_retries, which is used for
4208 all other user lookups.
4210 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4213 33. Exim was neglecting to zero errno before one call of strtol() when
4214 expanding a string and expecting an integer value. On some systems this
4215 resulted in spurious "integer overflow" errors. Also, it was casting the
4216 result into an int without checking.
4218 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4219 did not work. The code looks as if it has *never* worked, though it appears
4220 to have been documented since at least releast 1.62. I have made it work.
4222 35. The "timeout_DNS" error in retry rules, also documented since at least
4223 1.62, also never worked. As it isn't clear exactly what this means, and
4224 clearly it isn't a major issue, I have abolished the feature by treating it
4225 as "timeout", and writing a warning to the main and panic logs.
4227 36. The display of retry rules for -brt wasn't always showing the error code
4230 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4231 timeout_connect_A, timeout_connect_MX.
4233 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4234 to the empty sender.
4236 39. The daemon was not analysing the content of -oX till after it had closed
4237 stderr and disconnected from the controlling terminal. This meant that any
4238 syntax errors were only noted on the panic log, and the return code from
4239 the command was 0. By re-arranging the code a little, I've made the
4240 decoding happen first, so such errors now appear on stderr, and the return
4241 code is 1. However, the actual setting up of the sockets still happens in
4242 the disconnected process, so errors there are still only recorded on the
4245 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4246 connections (as happens on some IP stacks) was logged at start up time as
4247 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4248 it from "IPv6 and IPv4", which means that two separate sockets are being
4251 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4252 says whether AF_INET or AF_INET6 was passed as an argument.
4254 42. Exiwhat output was messed up when time zones were included in log
4257 43. Exiwhat now gives more information about the daemon's listening ports,
4258 and whether -tls-on-connect was used.
4260 44. The "port" option of the smtp transport is now expanded.
4262 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4263 silently ignored. Now an error message is written to the main and panic
4266 46. There's a new ACL modifier called "logwrite" which writes to a log file
4267 as soon as it is encountered.
4269 47. Added $local_user_uid and $local_user_gid at routing time.
4271 48. Exim crashed when trying to verify a sender address that was being
4274 49. Exim was recognizing only a space character after ".include". It now also
4275 recognizes a tab character.
4277 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4278 extracting the relevant information from the specification. The man page no
4279 longer contains scrambled data for the -d option, and I've added a section
4280 at the front about calling Exim under different names.
4282 51. Added "extra_headers" argument to the "mail" command in filter files.
4284 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4287 53. Installed eximstats 1.29.
4289 54. Added transport_filter_timeout as a generic transport option.
4291 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4292 Cc: header lines. This was required by RFC 822, but it not required by RFC
4295 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4296 incoming messages that did not have them. Now it does so only if the
4297 message originates locally, that is, if there is no associated remote host
4298 address. When Resent- header lines are present, this applies to the Resent-
4299 lines rather than the non-Resent- lines.
4301 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4302 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4304 58. Messages for configuration errors now include the name of the main
4305 configuration file - useful now that there may be more than one file in a
4306 list (.included file names were always shown).
4308 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4309 for those rare installations that do not start the daemon as root or run it
4310 setuid root. I've cut out the call to initgroups() if the daemon is not
4313 60. The Exim user and group can now be bound into the binary as text strings
4314 that are looked up at the start of Exim's processing.
4316 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4318 62. Added $mailstore_basename variable.
4320 63. Installed patch to sieve.c from Michael Haardt.
4322 64. When Exim failed to open the panic log after failing to open the main log,
4323 the original message it was trying to log was written to stderr and debug
4324 output, but if they were not available (the usual case in production), it
4325 was lost. Now it is written to syslog before the two lines that record the
4326 failures to open the logs.
4328 65. Users' Exim filters run in subprocesses under the user's uid. It is
4329 possible for a "deliver" command or an alias in a "personal" command to
4330 provoke an address rewrite. If logging of address rewriting is configured,
4331 this fails because the process is not running as root or exim. There may be
4332 a better way of dealing with this, but for the moment (because 4.30 needs
4333 to be released), I have disabled address rewrite logging when running a
4334 filter in a non-root, non-exim process.
4340 1. The buildconfig auxiliary program wasn't quoting the value set for
4341 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4342 not defined. This bug was masked in 4.22 by the effect that was fixed in
4345 2. Some messages that were rejected after a message id was allocated were
4346 shown as "incomplete" by exigrep. It no longer does this for messages that
4347 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4349 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4350 have allow_domain_literals set, the ID did not get logged in the <= line.
4351 Domain literals are now always recognized in Message-ID: header lines.
4353 4. The first argument for a ${extract expansion item is the key name or field
4354 number. Leading and trailing spaces in this item were not being ignored,
4355 causing some misleading effects.
4357 5. When deliver_drop_privilege was set, single queue runner processes started
4358 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4359 same command in the process it spins off) were not dropping privilege.
4361 6. When the daemon running as "exim" started a queue runner, it always
4362 re-executed Exim in the spun-off process. This is a waste of effort when
4363 deliver_drop_privilege is set. The new process now just calls the
4364 queue-runner function directly.
4370 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4373 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4374 ignored. Though the use of -oP was forcing the writing of a pid file, it
4375 was always written to the default place.
4377 3. If the message "no IP address found for host xxxx" is generated during
4378 incoming verification, it is now followed by identification of the incoming
4379 connection (so you can more easily find what provoked it).
4381 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4383 5. Added some features to "harden" Exim a bit more against certain attacks:
4385 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4386 be put in Local/Makefile. This is like the never_users runtime option,
4387 but it cannot be overridden. The default setting is "root".
4389 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4390 prefix string with which any file named in a -C command line option
4393 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4394 is retained for -C and -D only if the caller of Exim is root. Without
4395 it, the exim user may also use -C and -D and retain privilege.
4397 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4398 command line option is disabled.
4400 6. Macro names set by the -D option must start with an upper case letter, just
4401 like macro names defined in the configuration file.
4403 7. Added "dereference=" facility to LDAP.
4405 8. Two instances of the typo "uknown" in the source files are fixed.
4407 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4408 the Configure-Makefile script screwed up while processing it.
4410 10. Incorporated PCRE 4.4.
4412 11. The SMTP synchronization check was not operating right at the start of an
4413 SMTP session. For example, it could not catch a HELO sent before the client
4414 waited for the greeting. There is now a check for outstanding input at the
4415 point when the greeting is written. Because of the duplex, asynchronous
4416 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4417 way, but not yet received, when the check is performed.
4419 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4420 on TCP/IP sockets, because this apparently causes some broken clients to
4423 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4424 unchanged) from the Cygwin maintainer.
4426 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4427 was supported instead of testing for mbx. Effectively a typo.
4429 15. The spa authenticator server code was not checking that the input it
4430 received was valid base64.
4432 16. The debug output line for the "set" modifier in ACLs was not showing the
4433 name of the variable that was being set.
4435 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4437 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4438 lookup on its own. Something else was needed to trigger the lookup. For
4439 example, a match in host_lookup or the need for a host name in a host list.
4440 Now, if $sender_host_name is referenced and the host name has not yet been
4441 looked up, a lookup is performed. If the lookup fails, the variable remains
4442 empty, and $host_lookup_failed is set to "1".
4444 19. Added "eqi" as a case-independent comparison operator.
4446 20. The saslauthd authentication condition could segfault if neither service
4447 nor realm was specified.
4449 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4450 error message that was logged was misleading, and incoming SMTP
4451 connections were dropped. The message is now more accurate, and temporary
4452 errors are given to SMTP connections.
4454 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4455 (except RSET) with a 421 error, until QUIT is received. However, it was
4456 failing to send a response to QUIT.
4458 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4459 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4460 ACL is now run after the helo_try_verify_hosts code.
4462 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4463 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4464 case-independent, so other case variants are also recognized). Apparently
4465 some systems use these upper case variants.
4467 25. If more than two messages were waiting for the same host, and a transport
4468 filter was specified for the transport, Exim sent two messages over the
4469 same TCP/IP connection, and then failed with "socket operation on non-
4470 socket" when it tried to send the third.
4472 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4474 27. The extern definition of crypt16() in expand.c was not being excluded when
4475 the OS had its own crypt16() function.
4477 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4478 as a preferred synonym for return_size_limit, both as an option and as an
4481 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4483 30. Changed the default configuration ACL to relax the local part checking rule
4484 for addresses that are not in any local domains. For these addresses,
4485 slashes and pipe symbols are allowed within local parts, but the sequence
4486 /../ is explicitly forbidden.
4488 31. SPA server authentication was not clearing the challenge buffer before
4491 32. log_message in a "warn" ACL statement was writing to the reject log as
4492 well as to the main log, which contradicts the documentation and doesn't
4493 seem right (because no rejection is happening). So I have stopped it.
4495 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4496 However, I am unable to do any testing of this.
4498 34. Fixed an infelicity in the appendfile transport. When checking directories
4499 for a mailbox, to see if any needed to be created, it was accidentally
4500 using path names with one or more superfluous leading slashes; tracing
4501 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4503 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4504 contents are added to the log line that is written for every discarded
4505 recipient. (Previously a log_message setting was ignored.)
4507 36. The ${quote: operator now quotes the string if it is empty.
4509 37. The install script runs exim in order to find its version number. If for
4510 some reason other than non-existence or emptiness, which it checks, it
4511 could not run './exim', it was installing it with an empty version number,
4512 i.e. as "exim-". This error state is now caught, and the installation is
4515 38. An argument was missing from the function that creates an error message
4516 when Exim fails to connect to the socket for saslauthd authentication.
4517 This could cause Exim to crash, or give a corrupted message.
4519 39. Added isip, isip4, and isip6 to ${if conditions.
4521 40. The ACL variables $acl_xx are now saved with the message, and can be
4522 accessed later in routers, transports, and filters.
4524 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4525 strings in the file are not string-expanded.
4527 42. If a MAIL command specified a SIZE value that was too large to fit into an
4528 int variable, the check against message_size_limit failed. Such values are
4529 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4530 day this will have to be increased, but I don't think I want to be around
4531 when emails are that large.
4538 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4539 iconv() is not standard in FreeBSD.
4541 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4542 IPv6 enabled. The observed symptom was a segmentation fault on return from
4543 the function os_common_find_running_interfaces() in src/os.c.
4545 3. In the check_special_case() function in daemon.c I had used "errno" as an
4546 argument name, which causes warnings on some systems. This was basically a
4547 typo, since it was named "eno" in the comments!
4549 4. The code that waits for the clock to tick (at a resolution of some fraction
4550 of a second) so as to ensure message-id uniqueness was always waiting for
4551 at least one whole tick, when it could have waited for less. [This is
4552 almost certainly not relevant at current processor speeds, where it is
4553 unlikely to ever wait at all. But we try to future-proof.]
4555 5. The function that sleeps for a time interval that includes fractions of a
4556 second contained a race. It did not block SIGALRM between setting the
4557 timer, and suspending (a couple of lines later). If the interval was short
4558 and the sigsuspend() was delayed until after it had expired, the suspension
4559 never ended. On busy systems this could lead to processes getting stuck for
4562 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4563 process, before it forks any delivery processes. The open lookup caching
4564 mechanism meant that the open file or database connection was passed into
4565 the delivery process. The problem was that delivery processes always tidy
4566 up cached lookup data. This could cause a problem for the next delivery
4567 process started by the queue runner, because the external queue runner
4568 process does not know about the closure. So the next delivery process
4569 still has data in the lookup cache. In the case of a file lookup, there was
4570 no problem because closing a file descriptor in a subprocess doesn't affect
4571 the parent. However, if the lookup was caching a connection to a database,
4572 the connection was closed, and the second delivery process was likely to
4573 see errors such as "PGSQL: query failed: server closed the connection
4574 unexpectedly". The problem has been fixed by closing all cached lookups
4575 in a queue runner before running a delivery process.
4577 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4578 seem to have the "const" qualifier which it has on other OS. I've
4581 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4582 *of the same type* (client or server) with the same public name that an
4583 error should be diagnosed.
4585 9. When Exim looked up a host name for an IP address, but failed to find the
4586 original IP address when looking up the host name (a safety check), it
4587 output the message "<ip address> does not match any IP for NULL", which was
4588 confusing, to say the least. The bug was that the host name should have
4589 appeared instead of "NULL".
4591 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4592 user that is built into the binary, and the -C or -D options is used, root
4593 privilege is dropped before the configuration file is read. In addition,
4594 logging is switched to stderr instead of the normal log files. If the
4595 configuration then re-defines the Exim user, the unprivileged environment
4596 is probably not what is expected, so Exim logs a panic warning message (but
4599 However, if deliver_drop_privilege is set, the unprivileged state may well
4600 be exactly what is intended, so the warning has been cut out in that case,
4601 and Exim is allowed to try to write to its normal log files.
4607 1. smtp_return_error_details was not giving details for temporary sender
4608 or receiver verification errors.
4610 2. Diagnose a configuration error if two authenticators have the same public
4613 3. Exim used not to create the message log file for a message until the first
4614 delivery attempt. This could be confusing when incoming messages were held
4615 for policy or load reasons. The message log file is now created at the time
4616 the message is received, and an initial "Received" line is written to it.
4618 4. The automatically generated man page for command line options had a minor
4619 bug that caused no ill effects; however, a more serious problem was that
4620 the procedure for building the man page automatically didn't always
4621 operate. Consequently, release 4.20 contains an out-of-date version. This
4622 shouldn't happen again.
4624 5. When building Exim with embedded Perl support, the script that builds the
4625 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4626 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4628 6. The freeze_tell option was not being used for messages that were frozen on
4629 arrival, either by an ACL or by local_scan().
4631 7. Added the smtp_incomplete_transaction log selector.
4633 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4634 was accepting AUTH without a new EHLO.
4636 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4637 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4640 10. Logging of TCP/IP connections (when configured) now happens in the main
4641 daemon process instead of the child process, so that the TCP/IP connection
4642 count is more accurate (but it can never be perfect).
4644 11. The use of "drop" in a nested ACL was not being handled correctly in the
4645 outer ACL. Now, if condition failure induced by the nested "drop" causes
4646 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4647 or "require"), the connection is dropped.
4649 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4650 that yield "discard" can now be used with an "accept" or a "discard" verb,
4651 but an error is generated for any others (because I can't see a useful way
4652 to define what should happen).
4654 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4655 are now processed in the same way as lines in the Exim configuration file.
4656 In particular, continuation lines are supported.
4658 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4660 15. Added -ti meaning -t -i.
4662 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4663 domains, and warn by logging if others are found.
4665 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4666 returned from GIFCONF when Exim is trying to find the list of interfaces on
4667 a host. The code in os.c has been modified to copy each ifreq to an aligned
4668 structure in all cases.
4670 Also, in some cases, the returned ifreq's were being copied to a 'struct
4671 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4672 means the last couple of bytes of an IPv6 address could be chopped if the
4673 ifreq contained only a normal sockaddr (14 bytes storage).
4675 18. Named domain lists were not supported in the hosts_treat_as_local option.
4676 An entry such as +xxxx was not recognized, and was treated as a literal
4679 19. Ensure that header lines added by a DATA ACL are included in the reject log
4680 if the ACL subsequently rejects the message.
4682 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4683 MD5 (which is deprecated).
4685 21. When testing a filter file using -bf, Exim was writing a message when it
4686 took the sender from a "From " line in the message, but it was not doing so
4687 when it took $return_path from a Return-Path: header line. It now does.
4689 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4690 with a valid header line field name (a series of printing characters
4691 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4693 23. Changed "disc" in the source to "disk" to conform to the documentation and
4694 the book and for uniformity.
4696 24. Ignore Sendmail's -Ooption=value command line item.
4698 25. When execve() failed while trying to run a command in a pipe transport,
4699 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4700 could be confused with a return value of 69 from the command itself. This
4701 has been changed to 127, the value the shell returns if it is asked to run
4702 a non-existent command. The wording for the related log line suggests a
4703 non-existent command as the problem.
4705 26. If received_header_text expands to an empty string, do not add a Received:
4706 header line to the message. (Well, it adds a token one on the spool, but
4707 marks it "old" so that it doesn't get used or transmitted.)
4709 27. Installed eximstats 1.28 (addition of -nt option).
4711 28. There was no check for failure on the call to getsockname() in the daemon
4712 code. This can fail if there is a shortage of resources on the system, with
4713 ENOMEM, for example. A temporary error is now given on failure.
4715 29. Contrary to the C standard, it seems that in some environments, the
4716 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4717 program. Exim now does this explicitly; it affects the formatting of
4718 timestamps using strftime().
4720 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4721 complaints. I've now initialized all the variables, to avoid this.
4723 32. Header lines added by a system filter were not being "seen" during
4724 transport-time rewrites.
4726 33. The info_callback() function passed to OpenSSL is set up with type void
4727 (*)(SSL *, int, int), as described somewhere. However, when calling the
4728 function (actually a macro) that sets it up, the type void(*)() is
4729 expected. I've put in a cast to prevent warnings from picky compilers.
4731 34. If a DNS black list lookup found a CNAME record, but there were no A
4732 records associated with the domain it pointed at, Exim crashed.
4734 35. If a DNS black list lookup returned more than one A record, Exim ignored
4735 all but the first. It now scans all returned addresses if a particular IP
4736 value is being sought. In this situation, the contents of the
4737 $dnslist_value variable are a list of all the addresses, separated by a
4740 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4741 to accept a host name and all its aliases if the forward lookup for any of
4742 them yielded the IP address of the incoming connection. Now it accepts only
4743 those names whose forward lookup yields the correct IP address. Any other
4744 names are discarded. This closes a loophole whereby a rogue DNS
4745 administrator could create reverse DNS records to break through a
4746 wildcarded host restriction in an ACL.
4748 37. If a user filter or a system filter that ran in a subprocess used any of
4749 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4750 the wrong values were passed to the pipe command ($thisaddress had the
4751 value of $0, $0 had the value of $1, etc). This bug was introduced by
4752 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4754 38. Improved the line breaking for long SMTP error messages from ACLs.
4755 Previously, if there was no break point between 40 and 75 characters, Exim
4756 left the rest of the message alone. Two changes have been made: (a) I've
4757 reduced the minimum length to 35 characters; (b) if it can't find a break
4758 point between 35 and 75 characters, it looks ahead and uses the first one
4759 that it finds. This may give the occasional overlong line, but at least the
4760 remaining text gets split now.
4762 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4763 file descriptors might be low, and that setting 1000 would always raise it.
4764 It turns out that in some environments, the limit is already over 1000 and
4765 that lowering it causes trouble. So now Exim takes care not to decrease it.
4767 40. When delivering a message, the value of $return_path is set to $sender_
4768 address at the start of routing (routers may change the value). By an
4769 oversight, this default was not being set up when an address was tested by
4770 -bt or -bv, which affected the outcome if any router or filter referred to
4773 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4774 log only when "message" or "log_message" are set. However, if one of the
4775 conditions was an address verification, or a call to a nested ACL, the
4776 messages generated by the underlying test were being passed through. This
4777 no longer happens. The underlying message is available in $acl_verify_
4778 message for both "message" and "log_message" expansions, so it can be
4779 passed through if needed.
4781 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4782 new expansion $bh_ to give the encoded byte string without charset
4783 translation. Translation happens only if iconv() is available; HAVE_ICONV
4784 indicates this at build time. HEADERS_CHARSET gives the charset to
4785 translate to; headers_charset can change it in the configuration, and
4786 "headers charset" can change it in an individual filter file.
4788 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4789 that creates RFC 2047 encoded "words" labels them as that charset instead
4790 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4791 expansion operator; (ii) when Exim creates a From: line for a local
4792 message; (iii) when a header line is rewritten to include a "phrase" part.
4794 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4795 buggy, causing it to skip the first lines of messages whose message ID
4796 ended in 'D'. This would not have bitten before Exim release 4.14, because
4797 message IDs were unlikely to end in 'D' before then. The effect was to have
4798 incorrect size information for certain domains.
4800 45. #include "config.h" was missing at the start of the crypt16.c module. This
4801 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4804 46. If there was a timeout during a "random" callout check, Exim treated it as
4805 a failure of the random address, and carried on sending RSET and the real
4806 address. If the delay was just some slowness somewhere, the response to the
4807 original RCPT would be taken as a response to RSET and so on, causing
4808 mayhem of various kinds.
4810 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4811 when I implemented it. It didn't allow for the fact that some option values
4812 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4813 the right test for positive values.
4815 48. Domain names in DNS records are case-independent. Exim always looks them up
4816 in lower case. Some resolvers return domain names in exactly the case they
4817 appear in the zone file, that is, they may contain uppercase letters. Not
4818 all resolvers do this - some return always lower case. Exim was treating a
4819 change of case by a resolver as a change of domain, similar to a widening
4820 of a domain abbreviation. This triggered its re-routing code and so it was
4821 trying to route what was effectively the same domain again. This normally
4822 caused routing to fail (because the router wouldn't handle the domain
4823 twice). Now Exim checks for this case specially, and just changes the
4824 casing of the domain that it ultimately uses when it transmits the message
4827 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4830 50. If a filter generated a file delivery with a non-absolute name (possible if
4831 no home directory exists for the router), the forbid_file option was not
4834 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4835 the existing equality matching.
4837 52. Exim was using ints instead of ino_t variables in some places where it was
4838 dealing with inode numbers.
4840 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4841 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4842 called TMPDIR, and if it finds it is different, it changes its value.
4844 54. The smtp_printf() function is now made available to local_scan() so
4845 additional output lines can be written before returning. There is also an
4846 smtp_fflush() function to enable the detection of a dropped connection.
4847 The variables smtp_input and smtp_batched_input are exported to
4850 55. Changed the default runtime configuration: the message "Unknown user"
4851 has been removed from the ACL, and instead placed on the localuser router,
4852 using the cannot_route_message feature. This means that any verification
4853 failures that generate their own messages won't get overridden. Similarly,
4854 the "Unrouteable address" message that was in the ACL for unverifiable
4855 relay addresses has also been removed.
4857 56. Added hosts_avoid_esmtp to the smtp transport.
4859 57. The exicyclog script was not checking for the esoteric option
4860 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4861 will work only if exicyclog is run under the appropriate euid.
4863 58. Following a discussion on the list, the rules by which Exim recognises line
4864 endings on incoming messages have been changed. The -dropcr and drop_cr
4865 options are now no-ops, retained only for backwards compatibility. The
4866 following line terminators are recognized: LF CRLF CR. However, special
4867 processing applies to CR:
4869 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4870 nor a local message in the state where . is a terminator.
4872 (ii) If a bare CR is encountered in a header line, an extra space is added
4873 after the line terminator so as not to end the header. The reasoning
4874 behind this is that bare CRs in header lines are most likely either
4875 to be mistakes, or people trying to play silly games.
4877 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4878 was being incorrectly given as 18 bytes larger than it should have been.
4879 This is a VOB (very old bug).
4881 60. This may never have affected anything current, but just in case it has:
4882 When the local host is found other than at the start of a list of hosts,
4883 the local host, those with the same MX, and any that follow, are discarded.
4884 When the list in question was part of a longer list of hosts, the following
4885 hosts (not currently being processed) were also being discarded. This no
4886 longer happens. I'm not sure if this situation could ever has previously
4889 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4892 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4893 header. This is something that is recommended in a new Internet Draft, and
4894 is something that is documented as being done by Sendmail. There are two
4895 possible values. For messages generated by the autoreply transport, Exim
4898 Auto-Submitted: auto-replied
4900 whereas for all other generated messages (e.g. bounces) it adds
4902 Auto-Submitted: auto-generated
4904 63. The "personal" condition in filters now includes a test for the
4905 Auto-Submitted: header. If it contains the string "auto-" the message it
4906 not considered personal.
4908 64. Added rcpt_include_affixes as a generic transport option.
4910 65. Added queue_only_override (default true).
4912 66. Added the syslog_duplication option.
4914 67. If what should have been the first header line of a message consisted of
4915 a space followed by a colon, Exim was mis-interpreting it as a header line.
4916 It isn't of course - it is syntactically invalid and should therefore be
4917 treated as the start of the message body. The misbehaviour could have
4918 caused a number of strange effects, including loss of data in subsequent
4919 header lines, and spool format errors.
4921 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4922 client host had authenticated. This control can now be exercised by an ACL
4923 for more flexibility.
4925 69. By default, callouts do not happen when testing with -bh. There is now a
4926 variant, -bhc, which does actually run the callout code, including
4927 consulting and updating the callout cache.
4929 70. Added support for saslauthd authentication, courtesy of Alexander
4932 71. If statvfs() failed on the spool or log directories while checking their
4933 size for availability, Exim confusingly gave the error "space shortage".
4934 Furthermore, in debugging mode it crashed with a floating point exception.
4935 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4936 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4937 serious problem, Exim now writes to the main and panic logs when this
4938 happens, with details of the failure. It then refuses to accept the
4939 incoming message, giving the message "spool directory problem" or "log
4940 directory problem" with a 421 code for SMTP messages.
4942 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4943 0, 1, and 2 exist, because some OS complain for execs without them (see
4944 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4945 descriptors. However, the code omitted to check that the open succeeded,
4946 causing mysterious errors if for some reason the permissions on /dev/null
4947 got screwed. Now Exim writes a message to the main and panic logs, and
4948 bombs out if it can't open /dev/null.
4950 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4951 interact so that it is all more flexible. It is supposed to remain
4952 backwards compatible. Also added extra_local_interfaces.
4954 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4955 to bomb out with an assertion failure - to the client this appears as a
4956 connection drop. This problem occurs in the part of the code that was taken
4957 from the Samba project. Fortunately, the assertion is in a very simple
4958 function, so I have fixed this by reproducing the function inline in the
4959 one place where it is called, and arranging for authentication to fail
4960 instead of killing the process with assert().
4962 75. The SPA client code was not working when the server requested OEM rather
4963 than Unicode encoding.
4965 76. Added code to make require_files with a specific uid setting more usable in
4966 the case where statting the file as root fails - usually a non-root-mounted
4967 NFS file system. When this happens and the failure is EACCES, Exim now
4968 forks a subprocess and does the per-uid checking as the relevant uid.
4970 77. Added process_log_path.
4972 78. If log_file_path was not explicitly set, a setting of check_log_space or
4973 check_log_inodes was ignored.
4975 79. If a space check for the spool or log partitions fails, the incident is now
4976 logged. Of course, in the latter case the data may get lost...
4978 80. Added the %p formatting code to string_format() so that it can be used to
4979 print addresses in debug_print(). Adjusted all the address printing in the
4980 debugging in store.c to use %p rather than %d.
4982 81. There was a concern that a line of code in smtp_in.c could overflow a
4983 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4984 initially expressed, the concern was not well-founded, because trailing
4985 spaces are removed early. However, if the trailing spaces were followed by
4986 a NULL, they did not get removed, so the overflow was possible. Two fixes
4989 (a) I re-wrote the offending code in a cleaner fashion.
4990 (b) If an incoming SMTP command contains a NULL character, it is rejected
4993 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4994 runs initgroups(), so that if the Exim user is in any additional groups,
4995 they will be used during message reception.
5001 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob