1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.588 2009/11/06 13:29:47 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
127 plus update to original patch.
129 NM/26 Bugzilla 799: Documentation correction for ratelimit
131 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
132 Patch provided by David Brownlee
134 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
136 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
138 NM/30 Bugzilla 888: TLS documentation bugfixes
140 NM/31 Bugzilla 896: Dovecot buffer overrun fix
142 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
143 Unlike the original bugzilla I have changed all shell scripts in src tree
145 NM/33 Bugzilla 898: Transport filter timeout fix
146 Patch by Todd Rinaldo
148 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
149 Patch by Serge Demonchaux
151 NM/35 Bugzilla 39: Base64 decode bug fixes
152 Patch by Jakob Hirsch
158 TK/01 Add preliminary DKIM support. Currently requires a forked version of
159 ALT-N's libdkim that I have put here:
160 http://duncanthrax.net/exim-experimental/
162 Note to Michael Haardt: I had to rename some vars in sieve.c. They
163 were called 'true' and it seems that C99 defines that as a reserved
164 keyword to be used with 'bool' variable types. That means you could
165 not include C99-style headers which use bools without triggering
166 build errors in sieve.c.
168 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
169 as mailq or other aliases. Changed the --help handling significantly
170 to do whats expected. exim_usage() emits usage/help information.
172 SC/01 Added the -bylocaldomain option to eximstats.
174 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
176 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
178 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
184 PH/01 Another patch from the Sieve maintainer.
186 PH/02 When an IPv6 address is converted to a string for single-key lookup
187 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
188 dots are used instead of colons so that keys in lsearch files need not
189 contain colons. This was done some time before quoting was made available
190 in lsearch files. However, iplsearch files do require colons in IPv6 keys
191 (notated using the quote facility) so as to distinguish them from IPv4
192 keys. This meant that lookups for IP addresses in host lists did not work
193 for iplsearch lookups.
195 This has been fixed by arranging for IPv6 addresses to be expressed with
196 colons if the lookup type is iplsearch. This is not incompatible, because
197 previously such lookups could never work.
199 The situation is now rather anomolous, since one *can* have colons in
200 ordinary lsearch keys. However, making the change in all cases is
201 incompatible and would probably break a number of configurations.
203 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
206 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
207 conversion specification without a maximum field width, thereby enabling
208 a rogue spamd server to cause a buffer overflow. While nobody in their
209 right mind would setup Exim to query an untrusted spamd server, an
210 attacker that gains access to a server running spamd could potentially
211 exploit this vulnerability to run arbitrary code as the Exim user.
213 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
214 $primary_hostname instead of what libspf2 thinks the hosts name is.
216 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
217 a directory entry by the name of the lookup key. Previously, if a
218 symlink pointed to a non-existing file or a file in a directory that
219 Exim lacked permissions to read, a lookup for a key matching that
220 symlink would fail. Now it is enough that a matching directory entry
221 exists, symlink or not. (Bugzilla 503.)
223 PH/03 The body_linecount and body_zerocount variables are now exported in the
226 PH/04 Added the $dnslist_matched variable.
228 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
229 This means they are set thereafter only if the connection becomes
232 PH/06 Added the client_condition to authenticators so that some can be skipped
233 by clients under certain conditions.
235 PH/07 The error message for a badly-placed control=no_multiline_responses left
236 "_responses" off the end of the name.
238 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
240 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
241 (without spaces) instead of just copying the configuration text.
243 PH/10 Added the /noupdate option to the ratelimit ACL condition.
245 PH/11 Added $max_received_linelength.
247 PH/12 Added +ignore_defer and +include_defer to host lists.
249 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
250 way in which PCRE > 7.0 is built.
252 PH/14 Implemented queue_only_load_latch.
254 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
255 MAIL command. The effect was to mangle the value on 64-bit systems.
257 PH/16 Another patch from the Sieve maintainer.
259 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
261 PH/18 If a system quota error occurred while trying to create the file for
262 a maildir delivery, the message "Mailbox is full" was not appended to the
263 bounce if the delivery eventually timed out. Change 4.67/27 below applied
264 only to a quota excession during the actual writing of the file.
266 PH/19 It seems that peer DN values may contain newlines (and other non-printing
267 characters?) which causes problems in log lines. The DN values are now
268 passed through string_printing() before being added to log lines.
270 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
271 and InterBase are left for another time.)
273 PH/21 Added message_body_newlines option.
275 PH/22 Guard against possible overflow in moan_check_errorcopy().
277 PH/23 POSIX allows open() to be a macro; guard against that.
279 PH/24 If the recipient of an error message contained an @ in the local part
280 (suitably quoted, of course), incorrect values were put in $domain and
281 $local_part during the evaluation of errors_copy.
287 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
288 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
291 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
292 issue a MAIL command.
294 PH/02 In an ACL statement such as
296 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
298 if a client was not listed at all, or was listed with a value other than
299 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
300 the condition was not true (as it should be), so access was not denied.
301 The bug was that the ! inversion was incorrectly passed on to the second
302 item. This has been fixed.
304 PH/03 Added additional dnslists conditions == and =& which are different from
305 = and & when the dns lookup returns more than one IP address.
307 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
308 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
310 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
311 FSYNC, which compiles an option called disable_fsync that allows for
312 bypassing fsync(). The documentation is heavily laced with warnings.
314 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
316 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
317 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
318 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
319 including adding "make clean"; (3) Added -fPIC when compiling the test
320 dynamically loaded module, to get rid of a warning.
322 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
323 message fails, move_frozen_messages = true and ignore_bounce_errors_after
324 = 0s. The bug is otherwise harmless.
326 PH/07 There was a bug in the dovecot authenticator such that the value of
327 $auth1 could be overwritten, and so not correctly preserved, after a
328 successful authentication. This usually meant that the value preserved by
329 the server_setid option was incorrect.
331 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
333 PH/09 Installed PCRE release 7.0.
335 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
336 run for batched SMTP input. It is now run at the start of every message
337 in the batch. While fixing this I discovered that the process information
338 (output by running exiwhat) was not always getting set for -bs and -bS
339 input. This is fixed, and it now also says "batched" for BSMTP.
341 PH/11 Added control=no_pipelining.
343 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
344 patch, slightly modified), and move the expansion of helo_data till after
345 the connection is made in the smtp transport (so it can use these
348 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
350 PH/14 Added log_selector = +pid.
352 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
354 PH/16 Add ${if forany and ${if forall.
356 PH/17 Added dsn_from option to vary the From: line in DSNs.
358 PH/18 Flush SMTP output before performing a callout, unless control =
359 no_callout_flush is set.
361 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
362 was true (the default) a successful delivery failed to delete the retry
363 item, thus causing premature timeout of the address. The bug is now
366 PH/20 Added hosts_avoid_pipelining to the smtp transport.
368 PH/21 Long custom messages for fakedefer and fakereject are now split up
369 into multiline reponses in the same way that messages for "deny" and
370 other ACL rejections are.
372 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
373 with slight modification.
375 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
376 draft, changing the syntax and factoring some duplicate code".
378 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
379 for deliveries of the second and subsequent messages over the same SMTP
382 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
383 ${reduce, with only minor "tidies".
385 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
387 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
388 expansion side effects.
390 PH/27 When a message times out after an over-quota error from an Exim-imposed
391 quota, the bounce message says "mailbox is full". This message was not
392 being given when it was a system quota that was exceeded. It now should
395 MH/03 Made $recipients available in local_scan(). local_scan() already has
396 better access to the recipient list through recipients_list[], but
397 $recipients can be useful in postmaster-provided expansion strings.
399 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
400 in the case of a MAIL command with additional options following the
401 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
402 were accidentally chopped off.
404 PH/29 SMTP synchronization checks are implemented when a command is read -
405 there is a check that no more input is waiting when there shouldn't be
406 any. However, for some commands, a delay in an ACL can mean that it is
407 some time before the response is written. In this time, more input might
408 arrive, invalidly. So now there are extra checks after an ACL has run for
409 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
410 pipelining has not been advertised.
412 PH/30 MH's patch to allow iscntrl() characters to be list separators.
414 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
415 returned in the SMTP response when smtp_return_error_details was false.
418 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
419 instead of the C I/O that was originally supplied, because problems were
422 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
423 Exim which did not show up earlier: it was assuming that a call to
424 SSL_CTX_set_info_callback() might give an error value. In fact, there is
425 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
426 was a macro that became an assignment, so it seemed to work. This has
427 changed to a proper function call with a void return, hence the compile
428 error. Exim's code has been fixed.
430 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
433 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
435 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
437 PH/37 If a message is not accepted after it has had an id assigned (e.g.
438 because it turns out to be too big or there is a timeout) there is no
439 "Completed" line in the log. When some messages of this type were
440 selected by exigrep, they were listed as "not completed". Others were
441 picked up by some special patterns. I have improved the selection
442 criteria to be more general.
444 PH/38 The host_find_failed option in the manualroute router can now be set
445 to "ignore", to completely ignore a host whose IP address cannot be
446 found. If all hosts are ignored, the behaviour is controlled by the new
447 host_all_ignored option.
449 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
450 homing or because of multiple MX records with /mx) generated more than
451 one IP address, and the following item turned out to be the local host,
452 all the secondary addresses of the first item were incorrectly removed
453 from the list, along with the local host and any following hosts (which
454 is what is supposed to happen).
456 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
457 whoever called Exim into the -H file. In the case of the daemon it was
458 behaving confusingly. When first started, it used values for whoever
459 started the daemon, but after a SIGHUP it used the Exim user (because it
460 calls itself on a restart). I have changed the code so that it now always
463 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
464 message are rejected with the same error (e.g. no authentication or bad
465 sender address), and a DATA command is nevertheless sent (as can happen
466 with PIPELINING or a stupid MUA), the error message that was given to the
467 RCPT commands is included in the rejection of the DATA command. This is
468 intended to be helpful for MUAs that show only the final error to their
471 PH/42 Another patch from the Sieve maintainer.
473 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
474 Eximstats - Fixed some broken HTML links and added missing column headers
476 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
477 columns for Rejects, Temp Rejects, Ham, and Spam rows.
479 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
481 PH/43 Yet another patch from the Sieve maintainer.
483 PH/44 I found a way to check for a TCP/IP connection going away before sending
484 the response to the final '.' that terminates a message, but only in the
485 case where the client has not sent further data following the '.'
486 (unfortunately, this is allowed). However, in many cases there won't be
487 any further data because there won't be any more messages to send. A call
488 to select() can be used: if it shows that the input is "ready", there is
489 either input waiting, or the socket has been closed. An attempt to read
490 the next input character can distinguish the two cases. Previously, Exim
491 would have sent an OK response which the client would never have see.
492 This could lead to message repetition. This fix should cure that, at
493 least in a lot of common cases.
495 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
496 advertised in response to EHLO.
502 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
503 fixed by 4.65/MH/01 (is this a record?) are fixed:
505 (i) An empty string was always treated as zero by the numeric comparison
506 operators. This behaviour has been restored.
508 (ii) It is documented that the numeric comparison operators always treat
509 their arguments as decimal numbers. This was broken in that numbers
510 starting with 0 were being interpreted as octal.
512 While fixing these problems I realized that there was another issue that
513 hadn't been noticed. Values of message_size_limit (both the global option
514 and the transport option) were treated as octal if they started with 0.
515 The documentation was vague. These values are now always treated as
516 decimal, and I will make that clear in the documentation.
522 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
523 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
526 MH/01 Don't check that the operands of numeric comparison operators are
527 integers when their expansion is in "skipping" mode (fixes bug
528 introduced by 4.64-PH/07).
530 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
531 child addresses, Exim now panics and dies. Previously, because the count
532 is held in a short int, deliveries were likely to be lost. As such a
533 large number of recipients for a single message is ridiculous
534 (performance will be very, very poor), I have chosen to impose a limit
535 rather than extend the field.
541 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
542 leftover -K file (the existence of which was triggered by #402).
543 While we were at it, introduced process PID as part of the -K
544 filename. This should rule out race conditions when creating
547 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
548 processing considerably. Previous code took too long for large mails,
549 triggering a timeout which in turn triggers #401.
551 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
552 in the DK code in transports.c. sendfile() is not really portable,
553 hence the _LINUX specificness.
555 TF/01 In the add_headers option to the mail command in an Exim filter,
556 there was a bug that Exim would claim a syntax error in any
557 header after the first one which had an odd number of characters
560 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
561 callout verification, Exim cached a "reject" for the entire domain. This
562 is correct for most verifications, but it is not correct for a recipient
563 verification with use_sender or use_postmaster set, because in that case
564 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
565 case of MAIL FROM:<> rejection from other early rejections (e.g.
566 rejection of HELO). When verifying a recipient using a non-null MAIL
567 address, the cache is ignored if it shows MAIL FROM:<> rejection.
568 Whatever the result of the callout, the value of the domain cache is
569 left unchanged (for any other kind of callout, getting as far as trying
570 RCPT means that the domain itself is ok).
572 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
575 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
576 manifest itself as EPIPE rather than ECONNECT. When tidying away a
577 session, the daemon ignores ECONNECT errors and logs others; it now
578 ignores EPIPE as well.
580 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
581 (quoted-printable decoding).
583 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
584 later the small subsequent patch to fix an introduced bug.
586 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
588 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
590 PH/08 An error is now given if message_size_limit is specified negative.
592 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
593 to be given (somewhat) arbitrary names.
595 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
598 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
599 miscellaneous code fixes
601 PH/10 Added the log_reject_target ACL modifier to specify where to log
604 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
605 hostname. This is wrong, because it relates to the incoming message (and
606 probably the interface on which it is arriving) and not to the outgoing
607 callout (which could be using a different interface). This has been
608 changed to use the value of the helo_data option from the smtp transport
609 instead - this is what is used when a message is actually being sent. If
610 there is no remote transport (possible with a router that sets up host
611 addresses), $smtp_active_hostname is used.
613 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
614 tweaks were necessary in order to get it to work (see also 21 below):
615 (a) The code assumed that strncpy() returns a negative number on buffer
616 overflow, which isn't the case. Replaced with Exim's string_format()
618 (b) There were several signed/unsigned issues. I just did the minimum
619 hacking in of casts. There is scope for a larger refactoring.
620 (c) The code used strcasecmp() which is not a standard C function.
621 Replaced with Exim's strcmpic() function.
622 (d) The code set only $1; it now sets $auth1 as well.
623 (e) A simple test gave the error "authentication client didn't specify
624 service in request". It would seem that Dovecot has changed its
625 interface. Fortunately there's a specification; I followed it and
626 changed what the client sends and it appears to be working now.
628 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
631 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
632 address A is aliased to B and C, where B exists and C does not. Without
633 -v the output is "A verified" because verification stops after a
634 successful redirection if more than one address is generated. However,
635 with -v the child addresses are also verified. Exim was outputting "A
636 failed to verify" and then showing the successful verification for C,
637 with its parentage. It now outputs "B failed to verify", showing B's
638 parentage before showing the successful verification of C.
640 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
641 look up a TXT record in a specific list after matching in a combined
644 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
645 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
646 they consult the DNS. I had assumed they would set it the way they
647 wanted; and indeed my experiments on Linux seem to show that in some
648 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
649 To be on the safe side, however, I have now made the interface to
650 host_find_byname() similar to host_find_bydns(), with an argument
651 containing the DNS resolver options. The host_find_byname() function now
652 sets these options at its start, just as host_find_bydns() does. The smtp
653 transport options dns_qualify_single and dns_search_parents are passed to
654 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
655 of host_find_byname() use the default settings of RES_DEFNAMES
656 (qualify_single) but not RES_DNSRCH (search_parents).
658 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
659 spool_read_header() do less string testing, by means of a preliminary
660 switch on the second character of optional "-foo" lines. (This is
661 overdue, caused by the large number of possibilities that now exist.
662 Originally there were few.) While I was there, I also converted the
663 str(n)cmp tests so they don't re-test the leading "-" and the first
664 character, in the hope this might squeeze out yet more improvement.
666 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
667 flag allowing group syntax was set by the header_syntax check but not
668 turned off, possible causing trouble later; (2) The flag was not being
669 set at all for the header_verify test, causing "group"-style headers to
670 be rejected. I have now set it in this case, and also caused header_
671 verify to ignore an empty address taken from a group. While doing this, I
672 came across some other cases where the code for allowing group syntax
673 while scanning a header line wasn't quite right (mostly, not resetting
674 the flag correctly in the right place). These bugs could have caused
675 trouble for malformed header lines. I hope it is now all correct.
677 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
678 with the "reply" argument non-NULL. The code, however (which originally
679 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
680 but it didn't always do it. This confused somebody who was copying the
681 code for some other use. I have removed all the tests.
683 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
684 feature that was used to support insecure browsers during the U.S. crypto
685 embargo. It requires special client support, and Exim is probably the
686 only MTA that supported it -- and would never use it because real RSA is
687 always available. This code has been removed, because it had the bad
688 effect of slowing Exim down by computing (never used) parameters for the
689 RSA_EXPORT functionality.
691 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
692 authenticator to fail if there's a tab character in the incoming data
693 (there should never be unless someone is messing about, as it's supposed
694 to be base64-encoded). Also added, on Timo's advice, the "secured" option
695 if the connection is using TLS or if the remote IP is the same as the
696 local IP, and the "valid-client-cert option" if a client certificate has
699 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
700 authenticators. This can be used for authorization after authentication
701 succeeds. (In the case of plaintext, it servers for both authentication
704 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
705 if any retry times were supplied.
707 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
708 connection, where there is no client IP address to check. In this
709 situation, the verify now always succeeds.
711 PH/25 Applied John Jetmore's -Mset patch.
713 PH/26 Added -bem to be like -Mset, but loading a message from a file.
715 PH/27 In a string expansion for a processed (not raw) header when multiple
716 headers of the same name were present, leading whitespace was being
717 removed from all of them, but trailing whitespace was being removed only
718 from the last one. Now trailing whitespace is removed from each header
719 before concatenation. Completely empty headers in a concatenation (as
722 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
723 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
725 PH/29 [Removed. This was a change that I later backed out, and forgot to
726 correct the ChangeLog entry (that I had efficiently created) before
727 committing the later change.]
729 PH/30 Exim was sometimes attempting to deliver messages that had suffered
730 address errors (4xx response to RCPT) over the same connection as other
731 messages routed to the same hosts. Such deliveries are always "forced",
732 so retry times are not inspected. This resulted in far too many retries
733 for the affected addresses. The effect occurred only when there were more
734 hosts than the hosts_max_try setting in the smtp transport when it had
735 the 4xx errors. Those hosts that it had tried were not added to the list
736 of hosts for which the message was waiting, so if all were tried, there
737 was no problem. Two fixes have been applied:
739 (i) If there are any address or message errors in an SMTP delivery, none
740 of the hosts (tried or untried) are now added to the list of hosts
741 for which the message is waiting, so the message should not be a
742 candidate for sending over the same connection that was used for a
743 successful delivery of some other message. This seems entirely
744 reasonable: after all the message is NOT "waiting for some host".
745 This is so "obvious" that I'm not sure why it wasn't done
746 previously. Hope I haven't missed anything, but it can't do any
747 harm, as the worst effect is to miss an optimization.
749 (ii) If, despite (i), such a delivery is accidentally attempted, the
750 routing retry time is respected, so at least it doesn't keep
751 hammering the server.
753 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
754 in ${readsocket because some servers need this prod.
756 PH/32 Added some extra debug output when updating a wait-xxx database.
758 PH/33 The hint "could be header name not terminated by colon", which has been
759 given for certain expansion errors for a long time, was not being given
760 for the ${if def:h_colon_omitted{... case.
762 PH/34 The spec says: "With one important exception, whenever a domain list is
763 being scanned, $domain contains the subject domain." There was at least
764 one case where this was not true.
766 PH/35 The error "getsockname() failed: connection reset by peer" was being
767 written to the panic log as well as the main log, but it isn't really
768 panic-worthy as it just means the connection died rather early on. I have
769 removed the panic log writing for the ECONNRESET error when getsockname()
772 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
773 runs only) independently of the message's sender address. This meant
774 that, if the 4xx error was in fact related to the sender, a different
775 message to the same recipient with a different sender could confuse
776 things. In particualar, this can happen when sending to a greylisting
777 server, but other circumstances could also provoke similar problems.
778 I have changed the default so that the retry time for these errors is now
779 based a combination of the sender and recipient addresses. This change
780 can be overridden by setting address_retry_include_sender=false in the
783 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
784 remote server are returned as part of bounce messages. This was not
785 happening for LMTP over a pipe (the lmtp transport), but now it is the
786 same for both kinds of LMTP.
788 PH/38 Despite being documented as not happening, Exim was rewriting addresses
789 in header lines that were in fact CNAMEs. This is no longer the case.
791 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
792 and queue runs started by the daemon processed all messages. This has
793 been fixed so that -R and -S can now usefully be given with -q<time>.
795 PH/40 Import PCRE release 6.7 (fixes some bugs).
797 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
799 PH/42 Give an error if -q is specified more than once.
801 PH/43 Renamed the variables $interface_address and $interface_port as
802 $received_ip_address and $received_port, to make it clear that these
803 values apply to message reception, and not to the outgoing interface when
804 a message is delivered. (The old names remain recognized, of course.)
806 PH/44 There was no timeout on the connect() call when using a Unix domain
807 socket in the ${readsocket expansion. There now is.
809 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
810 be meaningful with "accept".
812 SC/01 Eximstats V1.43
813 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
815 SC/02 Eximstats V1.44
816 Use a glob alias rather than an array ref in the generated
817 parser. This improves both readability and performance.
819 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
820 Collect SpamAssassin and rejection statistics.
821 Don't display local sender or destination tables unless
822 there is data to show.
823 Added average volumes into the top table text output.
825 SC/04 Eximstats V1.46
826 Collect data on the number of addresses (recipients)
827 as well as the number of messages.
829 SC/05 Eximstats V1.47
830 Added 'Message too big' to the list of mail rejection
831 reasons (thanks to Marco Gaiarin).
833 SC/06 Eximstats V1.48
834 Mainlog lines which have GMT offsets and are too short to
835 have a flag are now skipped.
837 SC/07 Eximstats V1.49 (Alain Williams)
838 Added the -emptyok flag.
840 SC/08 Eximstats V1.50
841 Fixes for obtaining the IP address from reject messages.
843 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
844 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
845 whitesspace changes from 4.64-PH/27
847 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
850 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
851 are found, allow negative numbers in numeric criteria)
853 JJ/06 exipick.20061117.2, added new $message_body_missing variable
855 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
856 to match changes made in 4.64-PH/43
858 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
860 PH/47 Put in an explicit test for a DNS lookup of an address record where the
861 "domain" is actually an IP address, and force a failure. This locks out
862 those revolvers/nameservers that support "A-for-A" lookups, in
863 contravention of the specifications.
865 PH/48 When a host name was looked up from an IP address, and the subsequent
866 forward lookup of the name timed out, the host name was left in
867 $sender_host_name, contrary to the specification.
869 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
870 restricted to single-key lookups, Exim was not diagnosing an error if
871 * or *@ was used with a query-style lookup.
873 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
875 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
876 long ago, but noone interested enough thought of it. Let's just say that
877 the "1.1" means that there are some new functions that weren't there at
878 some point in the past.
880 PH/51 Error processing for expansion failure of helo_data from an smtp
881 transport during callout processing was broken.
883 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
884 tested/used via the -bh/-bhc/-bs options.
886 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
887 bug, fixed in subsequent PCRE releases).
889 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
890 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
892 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
898 SC/01 Use a glob alias rather than an array ref in eximstats generated
899 parser. This improves both readability and performance.
901 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
902 Don't display local sender or destination tables in eximstats unless
903 there is data to show.
904 Added average volumes into the eximstats top table text output.
906 SC/03 Collect data on the number of addresses (recipients) as well
907 as the number of messages in eximstats.
909 TF/01 Correct an error in the documentation for the redirect router. Exim
910 does (usually) call initgroups() when daemonizing.
912 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
913 with consistent privilege compared to when running as a daemon.
915 TF/03 Note in the spec that $authenticated_id is not set for local
916 submissions from trusted users.
918 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
919 Thanks to Dean Brooks <dean@iglou.com> for the patch.
921 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
922 by adding some example configuration directives to the default
923 configuration file. A little bit of work is required to uncomment the
924 directives and define how usernames and passwords are checked, but
925 there is now a framework to start from.
927 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
928 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
929 without this. I don't know how relevant this is to other LDAP libraries.
931 PH/02 Add the verb name to the "unknown ACL verb" error.
933 PH/03 Magnus Holmgren's patch for filter_prepend_home.
935 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
937 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
938 directory not expanded when it should be if an expanded home directory
939 was set for the address (which is overridden by the transport).
941 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
944 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
945 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
946 because it is too late at that time, and has no effect.
948 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
949 security issue with \' (bugzilla #107). I could not use the
950 PQescapeStringConn() function, because it needs a PGconn value as one of
953 PH/08 When testing addresses using -bt, indicate those final addresses that
954 are duplicates that would not cause an additional delivery. At least one
955 person was confused, thinking that -bt output corresponded to deliveries.
956 (Suppressing duplicates isn't a good idea as you lose the information
957 about possibly different redirections that led to the duplicates.)
959 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
960 systems where poll() doesn't work, in particular OS X.
962 PH/10 Added more information to debugging output for retry time not reached.
964 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
965 operations in malware.c.
967 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
970 PH/13 If write_rejectlog was set false when logging was sent to syslog with
971 syslog_duplication set false, log lines that would normally be written
972 both the the main log and to the reject log were not written to syslog at
975 PH/14 In the default configuration, change the use of "message" in ACL warn
976 statements to "add_header".
978 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
979 not followed by a command (e.g. "seen endif").
981 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
982 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
985 PH/17 Added extra conditions to the default value of delay_warning_condition
989 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
990 { match{$h_precedence:}{(?i)bulk|list|junk} } \
991 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
994 The Auto-Submitted: and various List- headers are standardised, whereas I
995 don't think Precedence: ever was.
997 PH/18 Refactored debugging code in route_finduser() to show more information,
998 in particular, the error code if getpwnam() issues one.
1000 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1001 This is apparently needed in addition to the PH/07 change above to avoid
1002 any possible encoding problems.
1004 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1005 but not after initializing Perl.
1007 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1008 output them only if debugging. By default they are written stderr,
1009 apparently, which is not desirable.
1011 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1014 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1017 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1019 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1020 authenticated or an ident call has been made. Suppress the default
1021 values for $authenticated_id and $authenticated_sender (but permit -oMai
1022 and -oMas) when testing with -bh.
1024 PH/24 Re-jigged the order of the tests in the default configuration so that the
1025 tests for valid domains and recipients precede the DNS black list and CSA
1026 tests, on the grounds that those ones are more expensive.
1028 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1029 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1030 This bug exists in every version of Exim that I still have, right back to
1033 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1034 However, an attempt to turn on case-sensitivity in a regex key by
1035 including (?-i) didn't work because the subject string was already
1036 lowercased, and the effects were non-intuitive. It turns out that a
1037 one-line patch can be used to allow (?-i) to work as expected.
1043 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1044 other effects) broke the use of negated acl sub-conditions.
1046 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1049 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1050 "Deny" causes Exim to reject the incoming connection with a 554 error.
1051 Unfortunately, if there is a major crisis, such as a disk failure,
1052 tcp-wrappers gives "deny", whereas what one would like would be some
1053 kind of temporary error. A kludge has been added to help with this.
1054 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1055 554 error is used if errno is still zero or contains ENOENT (which occurs
1056 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1059 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1061 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1062 errors. Otherwise a message that provokes a temporary error (when other
1063 messages do not) can cause a whole host to time out.
1065 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1066 addresses were routed directly to files or pipes from a redirect router.
1067 File deliveries just didn't batch; pipe deliveries might have suffered
1070 PH/06 A failure to get a lock for a hints database would erroneously always say
1071 "Failed to get write lock", even when it was really a read lock.
1073 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1074 of 0600. This has been changed to use the value of the lockfile_mode
1075 option (which defaults to 0600).
1077 PH/08 Applied small patch from the Sieve maintainer.
1079 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1080 folder from quota calculations, a direct delivery into this folder messed
1081 up the contents of the maildirsize file. This was because the regex was
1082 used only to exclude .Trash (or whatever) when the size of the mailbox
1083 was calculated. There was no check that a delivery was happening into an
1084 excluded directory. This bug has been fixed by ignoring all quota
1085 processing for deliveries into excluded directories.
1087 PH/10 Added the maildirfolder_create_regex option to appendfile.
1093 PH/01 The code for finding all the local interface addresses on a FreeBSD
1094 system running IPv6 was broken. This may well have applied to all BSD
1095 systems, as well as to others that have similar system calls. The broken
1096 code found IPv4 interfaces correctly, but gave incorrect values for the
1097 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1098 that it would not match correctly against @[] and not recognize the IPv6
1101 PH/02 The ipliteral router was not recognizing addresses of the form user@
1102 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1104 PH/03 Added disable_ipv6.
1106 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1107 decoded form, because it is most often used to construct To: headers
1108 lines in autoreplies, and the decoded form may well be syntactically
1109 invalid. However, $reply_address has leading white space removed, and all
1110 newlines turned into spaces so that the autoreply transport does not
1113 PH/05 If group was specified without a user on a router, and no group or user
1114 was specified on a transport, the group from the router was ignored.
1116 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1117 for visible compile-time settings that can be used to change these
1118 numbers, for those that want even more. Backwards compatibility with old
1119 spool files has been maintained. However, going back to a previous Exim
1120 release will lost any variables that are in spool files.
1122 PH/07 Two small changes when running in the test harness: increase delay when
1123 passing a TCP/IP connection to a new process, in case the original
1124 process has to generate a bounce, and remove special handling of
1125 127.0.0.2 (sic), which is no longer necessary.
1127 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1128 be the same on different OS.
1130 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1133 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1134 whether --show-vars was specified or not
1136 JJ/02 exipick: Added support for new ACL variable spool format introduced
1139 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1140 syntactically invalid From: or Reply-to: line, and a filter used this to
1141 generate an autoreply, and therefore failed to obtain an address for the
1142 autoreply, Exim could try to deliver to a non-existent relative file
1143 name, causing unrelated and misleading errors. What now happens is that
1144 it logs this as a hard delivery error, but does not attempt to create a
1147 PH/11 The exinext utility has a -C option for testing purposes, but although
1148 the given file was scanned by exinext itself; it wasn't being passed on
1149 when Exim was called.
1151 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1152 an end-of-file indication when reading a command response.
1154 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1155 compiled. In many other places in Exim, IPv6 addresses are always
1156 recognized, so I have changed this. It also means that IPv4 domain
1157 literals of the form [IPV4:n.n.n.n] are now always recognized.
1159 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1160 used if the router is not running as root, for example, when verifying at
1161 ACL time, or when using -bh. The debugging output from this situation was
1162 non-existent - all you got was a failure to exec. I have made two
1165 (a) Failures to set uid/gid, the current directory, or a process leader
1166 in a subprocess such as that created by queryprogram now generate
1167 suitable debugging ouput when -d is set.
1169 (b) The queryprogram router detects when it is not running as root,
1170 outputs suitable debugging information if -d is set, and then runs
1171 the subprocess without attempting to change uid/gid.
1173 PH/15 Minor change to Makefile for building test_host (undocumented testing
1176 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1177 additional section of a DNS packet that returns MX or SRV records.
1178 Instead, it always explicitly searches for A/AAAA records. This avoids
1179 major problems that occur when a DNS server includes only records of one
1180 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1181 fixed another bug: if SRV records were looked up and the corresponding
1182 address records were *not* found in the additional section, the port
1183 values from the SRV records were lost.
1185 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1186 using the correct key (the original address) when searching the retry
1187 rules in order to find which one to use for generating the retry hint.
1189 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1190 adding the default one. Similarly, if it contains a Reply-To: header, the
1191 errors_reply_to option, if set, is not used.
1193 PH/19 When calculating a retry time, Exim used to measure the "time since
1194 failure" by looking at the "first failed" field in the retry record. Now
1195 it does not use this if it is later than than the arrival time of the
1196 message. Instead it uses the arrival time. This makes for better
1197 behaviour in cases where some deliveries succeed, thus re-setting the
1198 "first failed" field. An example is a quota failure for a huge message
1199 when small messages continue to be delivered. Without this change, the
1200 "time since failure" will always be short, possible causing more frequent
1201 delivery attempts for the huge message than are intended.
1202 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1204 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1205 $1, $2, $3) because the numerical variables can be reset during some
1206 expansion items (e.g. "match"), thereby losing the authentication data.
1208 PH/21 Make -bV show the size of off_t variables so that the test suite can
1209 decide whether to run tests for quotas > 2G.
1211 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1212 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1213 filecount value is greater than 2G or if a quota value is greater than 2G
1214 on a system where the size of off_t is not greater than 4, a panic error
1217 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1218 never match. The debug and -bh output now contains an explicit error
1219 message indicating a malformed IPv4 address or mask.
1221 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1222 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1223 PH/23 above applies.
1225 PH/25 Do not write to syslog when running in the test harness. The only
1226 occasion when this arises is a failure to open the main or panic logs
1227 (for which there is an explicit test).
1229 PH/26 Added the /no_tell option to "control=freeze".
1231 PH/27 If a host name lookup failed very early in a connection, for example, if
1232 the IP address matched host_lookup and the reverse lookup yielded a name
1233 that did not have a forward lookup, an error message of the form "no IP
1234 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1235 could be logged. Now it outputs the IP address instead of "NULL".
1237 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1238 allows the sender and the authenticated sender to be set when
1239 submitting a message from within Exim. Since child_open_exim() is
1240 documented for local_scan(), the new function should be too.
1242 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1243 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1244 results in an empty string is now treated as unset.
1246 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1248 PH/31 Added sender_verify_fail logging option.
1250 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1251 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1252 tidied the source and removed it altogether.
1254 PH/33 When a queue run was abandoned because the load average was too high, a
1255 log line was always written; now it is written only if the queue_run log
1256 selector is set. In addition, the log line for abandonment now contains
1257 information about the queue run such as the pid. This is always present
1258 in "start" and "stop" lines but was omitted from the "abandon" line.
1260 PH/34 Omit spaces between a header name and the colon in the error message that
1261 is given when verify = headers_syntax fails (if there are lots of them,
1262 the message gets confusing).
1264 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1265 names, as there are now some PTR records that contain slashes. This check
1266 is only to protect against broken name servers that fall over on strange
1267 characters, so the fact that it applies to all lookups doesn't matter.
1269 PH/36 Now that the new test suite is complete, we can remove some of the
1270 special code in Exim that was needed for the old test suite. For example,
1271 sorting DNS records because real resolvers return them in an arbitrary
1272 order. The new test suite's fake resolver always returns records in the
1275 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1276 bounces) except when queue_only is set, to avoid logging races between
1277 the different processes.
1279 PH/38 Panic-die if .include specifies a non-absolute path.
1281 PH/39 A tweak to the "H" retry rule from its user.
1283 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1284 a label. They prevented compilation on older perls.
1286 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1287 a warning to be raised on newish perls.
1289 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1290 on queue. Changes to match documented behaviour of showing count of
1291 messages matching specified criteria.
1293 PH/40 Changed the default ident timeout from 30s to 5s.
1295 PH/41 Added support for the use of login_cap features, on those BSD systems
1296 that have them, for controlling the resources used by pipe deliveries.
1298 PH/42 The content-scanning code uses fopen() to create files in which to put
1299 message data. Previously it was not paying any attention to the mode of
1300 the files. Exim runs with umask(0) because the rest of the code creates
1301 files with open(), and sets the required mode explicitly. Thus, these
1302 files were ending up world-writeable. This was not a big issue, because,
1303 being within the spool directory, they were not world-accessible. I have
1304 created a function called modefopen, which takes an additional mode
1305 argument. It sets umask(777), creates the file, chmods it to the required
1306 mode, then resets the umask. All the relevant calls to fopen() in the
1307 content scanning code have been changed to use this function.
1309 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1310 to 24 hours. This avoids potential overflow problems when processing G
1311 and H retry rules. I suspect nobody ever tinkers with this value.
1313 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1315 PH/45 When the plaintext authenticator is running as a client, the server's
1316 challenges are checked to ensure they are valid base64 strings. By
1317 default, the authentication attempt is cancelled if an invalid string is
1318 received. Setting client_ignore_invalid_base64 true ignores these errors.
1319 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1320 they are received. Thus, the responses can be made to depend on the
1321 challenges. If an invalid string is ignored, an empty string is placed in
1324 PH/46 Messages that are created by the autoreply transport now contains a
1325 References: header, in accordance with RFCs 2822 and 3834.
1327 PH/47 Added authenticated_sender_force to the smtp transport.
1329 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1331 PH/49 Installed latest patch from the Sieve maintainer.
1333 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1334 was also set, the appendfile transport was unnecessarily scanning a
1335 directory of message files (e.g. for maildir delivery) to find the count
1336 of files (along with the size), even though it did not need this
1337 information. It now does the scan only if it needs to find either the
1338 size of the count of files.
1340 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1342 PH/52 Two bugs concerned with error handling when the smtp transport is
1345 (i) Exim was not creating retry information for temporary errors given
1346 for individual recipients after the DATA command when the smtp transport
1347 was used in LMTP mode. This meant that they could be retried too
1348 frequently, and not timed out correctly.
1350 (ii) Exim was setting the flag that allows error details to be returned
1351 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1352 recipients that were returned after the DATA command.
1354 PH/53 This is related to PH/52, but is more general: for any failing address,
1355 when detailed error information was permitted to be returned to the
1356 sender, but the error was temporary, then after the final timeout, only
1357 "retry timeout exceeded" was returned. Now it returns the full error as
1358 well as "retry timeout exceeded".
1360 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1361 do this, and (what is worse) MTAs that accept it.
1363 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1364 will now be deprecated.
1366 PH/56 New os.c-cygwin from the Cygwin maintainer.
1368 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1369 formats (previously only available in exim formats via -bpr, -bpru,
1370 and -bpra. Now also available in native and exiqgrep formats)
1372 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1373 with very large, slow to parse queues
1375 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1377 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1379 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1380 responses to authentication challenges, though it was showing the
1381 challenges; (ii) I've removed the CR characters from the debug output for
1384 PH/58 Allow for the insertion of a newline as well as a space when a string
1385 is turned into more than one encoded-word during RFC 2047 encoding. The
1386 Sieve code now uses this.
1388 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1389 data_4xx, lost_connection, tls_required.
1391 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1392 message was being sent as an SMTP response.
1394 PH/61 Add -l and -k options to exicyclog.
1396 PH/62 When verifying, if an address was redirected to one new address, so that
1397 verification continued, and the new address failed or deferred after
1398 having set something in $address_data, the value of $address_data was not
1399 passed back to the ACL. This was different to the case when no
1400 redirection occurred. The value is now passed back in both cases.
1402 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1403 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1404 use login_cap.h, so on its own it isn't the distinguishing feature. The
1405 new name refers directly to the setclassresources() function.
1407 PH/65 Added configuration files for NetBSD3.
1409 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1411 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1412 is preferred over IPv4.
1414 PH/68 The bounce_return_message and bounce_return_body options were not being
1415 honoured for bounces generated during the reception of non-SMTP messages.
1416 In particular, this applied to messages rejected by the ACL. This bug has
1417 been fixed. However, if bounce_return_message is true and bounce_return_
1418 body is false, the headers that are returned for a non-SMTP message
1419 include only those that have been read before the error was detected.
1420 (In the case of an ACL rejection, they have all been read.)
1422 PH/69 The HTML version of the specification is now built in a directory called
1423 spec_html instead of spec.html, because the latter looks like a path with
1424 a MIME-type, and this confuses some software.
1426 PH/70 Catch two compiler warnings in sieve.c.
1428 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1429 function verify_get_ident() calls ip_connect() to connect a socket, but
1430 if the "connect()" function timed out, ip_connect() used to close the
1431 socket. However, verify_get_ident() also closes the socket later, and in
1432 between Exim writes to the log, which may get opened at this point. When
1433 the socket was closed in ip_connect(), the log could get the same file
1434 descriptor number as the socket. This naturally causes chaos. The fix is
1435 not to close the socket in ip_connect(); the socket should be closed by
1436 the function that creates it. There was only one place in the code where
1437 this was missing, in the iplookup router, which I don't think anybody now
1438 uses, but I've fixed it anyway.
1440 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1441 well as to direct DNS lookups. Otherwise the handling of names in host
1442 lists is inconsistent and therefore confusing.
1448 PH/01 Two changes to the default runtime configuration:
1450 (1) Move the checks for relay_from_hosts and authenticated clients from
1451 after to before the (commented out) DNS black list checks.
1453 (2) Add control=submission to the relay_from_hosts and authenticated
1454 clients checks, on the grounds that messages accepted by these
1455 statements are most likely to be submissions.
1457 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1459 (1) Generate an error if the third argument for the ${prvs expansion is
1462 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1465 (3) Reset the variables that are obtained from the first argument of
1466 ${prvscheck and used in the second argument before leaving the code,
1467 because their memory is reclaimed, so using them afterwards may do
1470 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1471 one (it's much easier than Tom thought :-).
1473 (5) Because of (4), we can now allow for the use of $prvscheck_result
1474 inside the third argument.
1476 PH/03 For some reason, the default setting of PATH when running a command from
1477 a pipe transport was just "/usr/bin". I have changed it to
1480 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1481 anything to be listed in the output from -bV.
1483 PH/05 When a filter generated an autoreply, the entire To: header line was
1484 quoted in the delivery log line, like this:
1486 => >A.N.Other <ano@some.domain> <original@ddress> ...
1488 This has been changed so that it extracts the operative address. There
1489 may be more than one such address. If so, they are comma-separated, like
1492 => >ano@some.domain,ona@other.domain <original@ddress> ...
1494 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1495 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1496 not being looked up in the rDNS to get a host name, Exim was showing the
1497 IP address twice in Received: lines, even though the IP addresses were
1498 identical. For example:
1500 Received: from [1.2.3.4] (helo=[1.2.3.4])
1502 However, if the real host name was known, it was omitting the HELO data
1503 if it matched the actual IP address. This has been tidied up so that it
1504 doesn't show the same IP address twice.
1506 PH/07 When both +timestamp and +memory debugging was on, the value given by
1507 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1508 called by the debug printing, thereby overwriting the timestamp buffer.
1509 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1511 PH/08 When the original message was included in an autoreply transport, it
1512 always said "this is a copy of the message, including all the headers",
1513 even if body_only or headers_only was set. It now gives an appropriate
1516 PH/09 Applied a patch from the Sieve maintainer which:
1518 o fixes some comments
1519 o adds the (disabled) notify extension core
1520 o adds some debug output for the result of if/elsif tests
1521 o points to the current vacation draft in the documentation
1522 and documents the missing references header update
1526 o fixes a bug in processing the envelope test (when testing
1527 multiple envelope elements, the last element determinted the
1530 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1531 Electronic Mail") by including:
1533 Auto-submitted: auto-generated
1535 in the messages that it generates (bounce messages and others, such as
1536 warnings). In the case of bounce messages for non-SMTP mesages, there was
1537 also a typo: it was using "Auto_submitted" (underscore instead of
1538 hyphen). Since every message generated by Exim is necessarily in response
1539 to another message, thes have all been changed to:
1541 Auto-Submitted: auto-replied
1543 in accordance with these statements in the RFC:
1545 The auto-replied keyword:
1547 - SHOULD be used on messages sent in direct response to another
1548 message by an automatic process,
1550 - MUST NOT be used on manually-generated messages,
1552 - MAY be used on Delivery Status Notifications (DSNs) and Message
1553 Disposition Notifications (MDNs),
1555 - MUST NOT be used on messages generated by automatic or periodic
1556 processes, except for messages which are automatic responses to
1559 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1560 to the default Received: header definition.
1562 PH/12 Added log selector acl_warn_skipped (default on).
1564 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1565 variables because (a) they are in the wrong storage pool and (b) even if
1566 they were copied, it wouldn't work properly because of the caching.
1568 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1569 checking when decoding. Apparently there are clients that generate
1570 overlong encoded strings. Why am I not surprised?
1572 PH/15 If the first argument of "${if match_address" was not empty, but did not
1573 contain an "@" character, Exim crashed. Now it writes a panic log message
1574 and treats the condition as false.
1576 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1578 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1579 extension "envelope-auth". The code is finished and in agreement with
1580 other implementations, but there is no documentation so far and in fact,
1581 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1582 not changing the active code.
1584 Print executed "if" and "elsif" statements when debugging is used. This
1585 helps a great deal to understand what a filter does.
1587 Document more things not specified clearly in RFC3028. I had all this
1588 sorted out, when out of a sudden new issues came to my mind. Oops."
1590 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1593 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1594 canonical form (as documented). However, after a host name lookup from
1595 the IP address, check_host() was doing a simple string comparison with
1596 addresses acquired from the DNS when checking that the found name did
1597 have the original IP as one of its addresses. Since any found IPv6
1598 addresses are likely to be in abbreviated form, the comparison could
1599 fail. Luckily, there already exists a function for doing the comparison
1600 by converting both addresses to binary, so now that is used instead of
1601 the text comparison.
1603 PH/20 There was another similar case to PH/19, when a complete host name was
1604 given in a host list; looking up its IP address could give an abbreviated
1605 form, whereas the current host's name might or might not be abbreviated.
1606 The same fix has been applied.
1612 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1613 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1616 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1617 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1619 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1621 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1622 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1623 Also, for "Auto-Submitted", treat anything other than "no" as
1624 non-personal, in accordance with RFC 3834. (Previously it treated
1625 anything starting "auto-" as non-personal.)
1627 TF/01 The control=submission/name=... option had a problem with syntax
1628 errors if the name included a slash character. The /name= option
1629 now slurps the rest of the string, so it can include any characters
1630 but it must come last in the list of options (after /sender_retain
1633 PH/05 Some modifications to the interface to the fake nameserver for the new
1641 TK/01 Added the "success_on_redirect" address verification option. See
1642 NewStuff for rationale and an example.
1644 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1646 PH/02 Patch to exigrep to allow it to work on syslog lines.
1648 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1649 fread() to skip over the body file's header line, because in Cygwin the
1650 header line is locked and is inaccessible.
1652 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1653 co-exist for some time) to make it clear that it is the Exim ID that is
1654 referenced, not the Message-ID: header line.
1656 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1657 string_format() function, because snprintf() does not exist on all
1660 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1661 ${stat: expansion item.
1663 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1664 protocol synchronization error", to keep the pedants happy.
1666 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1667 well as for IRIX systems, when gcc is being used. See the host.c source
1670 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1672 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1675 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1676 transport and to the smtp transport in LMTP mode.
1678 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1680 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1681 run a filter in a subprocess. This could lead to confusion in subsequent
1682 lookups in the parent process. There should also be a search_tidyup() at
1683 the end of the subprocess.
1685 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1686 only if the host matched helo_try_verify_hosts, which caused the
1687 verification to occur when the EHLO/HELO command was issued. The ACL just
1688 tested the remembered result. Now, if a previous verification attempt has
1689 not happened, "verify = helo" does it there and then.
1691 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1693 TK/03 Fix log output including CR from clamd.
1695 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1696 exist provoked a memory error which could cause a segfault.
1698 PH/15 Installed PCRE 6.2
1700 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1702 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1703 of the problem. Specifically, suggested +O2 rather than +O1 for the
1706 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1708 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1709 record, the port was not correctly logged when the outgoing_port log
1710 selector was set (it logged the transort's default port).
1712 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1713 fallback_hosts, and "hosts" in the smtp transport.
1715 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1716 host errors such as "Connection refused".
1718 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1719 authentication with radiusclient 0.4.9:
1721 - Error returned from rc_read_config was caught wrongly
1722 - Username/password not passed on to radius server due to wrong length.
1724 The presumption is that some radiusclient API changes for 4.51/PH/17
1725 were not taken care of correctly. The code is still untested by me (my
1726 Linux distribution still has 0.3.2 of radiusclient), but it was
1727 contributed by a Radius user.
1729 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1730 expanding the "port" option of the smtp transport.
1732 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1733 while reading a MIME header. Thanks to Tom Hughes for a patch.
1735 PH/24 Include config.h inside local_scan.h so that configuration settings are
1738 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1739 that in an ACL for RCPT (for example), you can examine exactly what was
1742 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1743 commands, but it was not correctly comparing the address with the actual
1744 client host address. Thus, it would show the EHLO address in Received:
1745 header lines when this was not necessary.
1747 PH/27 Added the % operator to ${eval:}.
1749 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1750 it should be ignoring failures (because with -C, for example, it has lost
1751 privilege). It wasn't ignoring creation failures other than "already
1754 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1757 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1758 because an input line was too long, either on its own, or by virtue of
1759 too many continuations, the temporary file was not being removed, and the
1760 return code was incorrect.
1762 PH/31 Missing "BOOL" in function definition in filtertest.c.
1764 PH/32 Applied Sieve patches from the maintainer.
1766 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1768 PH/33 Added "verify = not_blind".
1770 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1771 Local/Makefile (with some defaults set). These are used in built scripts
1772 such as exicyclog, but they have never been used in the exim_install
1773 script (though there are many overriding facilities there). I have
1774 arranged that the exim_install script now takes note of these two
1777 PH/35 Installed configuration files for Dragonfly.
1779 PH/36 When a locally submitted message by a trusted user did not contain a
1780 From: header, and the sender address was obtained from -f or from an SMTP
1781 MAIL command, and the trusted user did not use -F to supply a sender
1782 name, $originator_name was incorrectly used when constructing a From:
1783 header. Furthermore, $originator_name was used for submission mode
1784 messages from external hosts without From: headers in a similar way,
1785 which is clearly wrong.
1787 PH/37 Added control=suppress_local_fixups.
1789 PH/38 When log_selector = +received_sender was set, and the addition of the
1790 sender made the log line's construction buffer exactly full, or one byte
1791 less than full, an overflow happened when the terminating "\n" was
1794 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1795 when the result of a list match is failure because a DNS lookup failed.
1797 PH/40 RM_COMMAND is now used in the building process.
1799 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1800 the "build-* directories that it finds.
1802 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1803 address in a domain literal was a prefix of an interface address.
1805 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1806 when verifying a sender address, unless rewrite_headers is false.
1808 PH/44 Wrote a long comment about why errors_to addresses are verified as
1809 recipients, not senders.
1811 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1812 the ratelimit ACL was added.
1814 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1816 PH/46 Added extra information about PostgreSQL errors to the error string.
1818 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1819 avoiding the need to install special zones in a real server. This is
1820 backwards compatible; if it can't find the fake resolver, it drops back.
1821 Thus, both old and new test suites can be run.
1823 TF/02 Added util/ratelimit.pl
1825 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1826 clock is set back in time.
1828 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1829 Candler <B.Candler@pobox.com>.
1831 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1832 OK for addresses that are the result of redirections.
1834 PH/48 A number of further additions for the benefit of the new test suite,
1835 including a fake gethostbyname() that interfaces to the fake DNS resolver
1838 TF/06 The fix for widen_domains has also been applied to qualify_single and
1839 search_parents which are the other dnslookup options that can cause
1842 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1845 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1847 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1848 service" if it encountered an MX record with an empty target hostname.
1849 The message is now "an MX or SRV record indicated no SMTP service".
1851 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1852 if the DNS of the sending site is misconfigured. This is quite a
1853 common situation. This change restores the behaviour of treating a
1854 helo verification defer as a failure.
1856 PH/51 If self=fail was set on a router, the bounce message did not include the
1857 actual error message.
1863 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1865 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1866 command itself ended in error, the underlying message about the transport
1867 filter timeout was being overwritten with the pipe command error. Now the
1868 underlying error message should be appended to the second error message.
1870 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1872 PH/02 Reduce the amount of output that "make" produces by default. Full output
1873 can still be requested.
1875 PH/03 The warning log line about a condition test deferring for a "warn" verb
1876 was being output only once per connection, rather than after each
1877 occurrence (because it was using the same function as for successful
1878 "warn" verbs). This seems wrong, so I have changed it.
1880 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1881 it should not have, which might have caused a crash in the right
1882 circumstances, but probably never did.
1884 PH/04 Installed a modified version of Tony Finch's patch to make submission
1885 mode fix the return path as well as the Sender: header line, and to
1886 add a /name= option so that you can make the user's friendly name appear
1889 TF/03 Added the control = fakedefer ACL modifier.
1891 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1892 Mark Lowes for thorough testing.
1894 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1896 TK/03 Merged latest SRS patch from Miles Wilton.
1898 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1899 with the definition in sysexits.h (which is #included earlier).
1900 Fortunately, Exim does not actually use EX_OK. The code used to try to
1901 preserve the sysexits.h value, by assumimg that macro definitions were
1902 scanned for macro replacements. I have been disabused of this notion,
1903 so now the code just undefines EX_OK before #including unistd.h.
1905 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1906 in the smtp transport. When a block could not be written in a single
1907 write() function, the timeout was being re-applied to each part-write.
1908 This seems wrong - if the receiver was accepting one byte at a time it
1909 would take for ever. The timeout is now adjusted when this happens. It
1910 doesn't have to be particularly precise.
1912 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1913 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1915 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1916 without a domain if the check to <postmaster@domain> fails.
1918 SC/01 Eximstats: added -xls and the ability to specify output files
1919 (patch written by Frank Heydlauf).
1921 SC/02 Eximstats: use FileHandles for outputing results.
1923 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1925 SC/04 Eximstats: fixed display of large numbers with -nvr option
1927 SC/05 Eximstats: fixed merging of reports with empty tables.
1929 SC/06 Eximstats: added the -include_original_destination flag
1931 SC/07 Eximstats: removed tabs and trailing whitespace.
1933 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1935 TK/06 MBOX spool code: Add real "From " MBOX separator line
1936 so the .eml file is really in mbox format (even though
1937 most programs do not really care). Patch from Alex Miller.
1939 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1940 The latter is generated from $received_to and is only set if the
1941 message has one envelope recipient. SA can use these headers,
1942 obviously out-of-the-box. Patch from Alex Miller.
1944 PH/08 The ${def test on a variable was returning false if the variable's
1945 value was "0", contrary to what the specification has always said!
1946 The result should be true unless the variable is empty.
1948 PH/09 The syntax error of a character other than { following "${if
1949 def:variable_name" (after optional whitespace) was not being diagnosed.
1950 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1951 accidental colon was present, for example, could give incorrect results.
1953 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1954 result is used (not including appendfile, where other changes are about
1957 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1958 This involved changing a lot of size variables from int to off_t. It
1959 should work with maildirs and everything.
1961 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1962 spamd dying while we are connected to it.
1964 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1967 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1968 function for BDB 4.3.
1970 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1972 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1973 the sources for PCRE have been re-arranged and more files are now
1976 PH/15 The code I had for printing potentially long long variables in PH/11
1977 above was not the best (it lost precision). The length of off_t variables
1978 is now inspected at build time, and an appropriate printing format (%ld
1979 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1980 to be "long long int" or "long int". This is needed for the internal
1981 formatting function string_vformat().
1983 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1984 the configuration file to be ":syslog", then the script "guesses" where
1985 the logs files are, rather than using the compiled in default. In our
1986 case the guess is not the same as the compiled default, so the script
1987 suddenly stopped working when I started to use syslog. The patch checks
1988 to see if log_file_path is "". If so, it attempts to read it from exim
1989 with no configuration file to get the compiled in version, before it
1990 falls back to the previous guessing code."
1992 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1993 implementing BATV in an Exim configuration. See NewStuff for the gory
1996 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1997 Makefile that are specific to HP-UX.
1999 PH/18 If the "use_postmaster" option was set for a recipient callout together
2000 with the "random" option, the postmaster address was used as the MAIL
2001 FROM address for the random test, but not for the subsequent recipient
2002 test. It is now used for both.
2004 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2005 patch removes a few documentation additions to RFC 3028, because the
2006 latest draft now contains them. It adds the new en;ascii-case comparator
2007 and a new error check for 8bit text in MIME parts. Comparator and
2008 require names are now matched exactly. I enabled the subaddress
2009 extension, but it is not well tested yet (read: it works for me)."
2011 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2012 rework some of the code of TK/09 above to avoid the hardwired use of
2013 "%lld" and "long long". Replaced the call to snprintf() with a call to
2016 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2017 records point to non-existent hosts", "retry timeout exceeded", and
2018 "retry time not reached for any host after a long failure period".
2020 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2021 experimental DomainKeys support:
2023 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2024 (2) On an error such as an illegally used "control", the wrong name for
2025 the control was given.
2027 These problems did NOT occur unless DomainKeys support was compiled.
2029 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2031 PH/24 Added ${if match_ip condition.
2033 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2034 databases so that it will be absolutely obvious if a crash occurs in the
2035 DB library. This is a regular occurrence (often caused by mis-matched
2038 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2039 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2040 picked up on a user's system that detects such things. There doesn't seem
2041 to be a gcc warning option for this - only an attribute that has to be
2042 put on the function's prototype. It seems that in Fedora Core 4 they have
2043 set this on a number of new functions. No doubt there will be more in due
2046 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2047 specify a transport. However, if an address that was verified by such a
2048 router was the subject of a callout, Exim crashed because it tried to
2049 read the rcpt_include_affixes from the non-existent transport. Now it
2050 just assumes that the setting of that option is false. This bug was
2051 introduced by 4.51/PH/31.
2053 PH/28 Changed -d+all to exclude +memory, because that information is very
2054 rarely of interest, but it makes the output a lot bigger. People tend to
2055 do -d+all out of habit.
2057 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2058 code in os-type was giving problems when libc.so lives in lib64, like on
2061 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2062 aren't the modern standard, and it seems that some systems' include files
2063 don't always have them. Exim was already checking for some of the newer
2064 ones like T_AAAA, and defining it itself. I've added checks for all the
2065 record types that Exim uses.
2067 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2068 not automatically generating a new one, as it is supposed to. This
2069 prevented TLS from working. If the file did exist, but contained invalid
2070 data, a new version was generated, as expected. It was only the case of a
2071 non-existent file that was broken.
2073 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2074 with a change in libdomainkeys > 0.64.
2076 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2077 from DNS. If the selector record carries the flag, it now has
2078 precedence over the domain-wide flag.
2080 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2082 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2083 the use of a port name that isn't defined in /etc/services) occurred, the
2084 message was deferred as in a normal delivery, and thus remained on the
2085 spool, instead of being failed because of the mua_wrapper setting. This
2086 is now fixed, and I tidied up some of the mua_wrapper messages at the
2089 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2090 the messages in a hash of arrays rather than using individual hashes.
2091 This is a bit cleaner and results in dramatic memory savings, albeit
2092 at a slight CPU cost.
2094 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2095 as requested by Marc Sherman.
2097 SC/10 Eximstats: added histograms for user specified patterns as requested
2100 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2102 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2103 fopen() in the content-scanning modules that did not already have it.
2109 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2110 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2112 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2114 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2116 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2117 to test Sieve filters that use "vacation".
2119 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2120 that changes the way the GnuTLS parameters are stored in the cache file.
2121 The new format can be generated externally. For backward compatibility,
2122 if the data in the cache doesn't make sense, Exim assumes it has read an
2123 old-format file, and it generates new data and writes a new file. This
2124 means that you can't go back to an older release without removing the
2127 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2128 work if there are any delivery delays because "one_time" forces the
2129 parent to be marked "delivered", so its unseen clone is never tried
2130 again. For this reason, Exim now forbids the simultaneous setting of
2133 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2134 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2135 entry if you want to know the details. The fix, however, neglected to
2136 consider the case when local delivery batching is involved. The test for
2137 "previously delivered" was not happening when checking to see if an
2138 address could be batched with a previous (undelivered) one; under
2139 certain circumstances this could lead to multiple deliveries to the same
2142 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2143 in its include files, and this causes problems building Exim.
2145 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2146 header_syntax) but Exim was just ignoring anything given after a slash.
2147 In particular, this caused confusion with an attempt to use "verify =
2148 reverse_host_lookup/defer_ok". An error is now given when options are
2149 supplied for verify items that do not have them. (Maybe reverse_host_
2150 lookup should have a defer_ok option, but that's a different point.)
2152 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2153 defined by RFC 821) to 2048, because there were problems with some AUTH
2154 commands, and RFC 1869 says the size should be increased for extended
2155 SMTP commands that take arguments.
2157 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2160 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2161 "unknown" error; now it says that the functionality isn't in the binary.
2163 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2164 an address' error message when a string expansion fails (syntax or
2165 whatever). Otherwise the password may appear in the log. Following change
2166 PH/42 below, there is no longer a chance of it appearing in a bounce
2169 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2171 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2172 of its addresses was ever tried. (Bugzilla bug #2.)
2174 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2175 the result incorrectly in the debug output. (It correctly added a newline
2176 to what was transported.)
2178 TF/01 Added $received_time.
2180 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2181 commented out examples of how to interface to a virus scanner and to
2182 SpamAssassin. Also added commented examples of av_scanner and
2183 spamd_address settings.
2185 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2186 and controls are allowed in which ACLs. There were a couple of minor
2187 errors. Some of the entries in the conditions table (which is a table of
2188 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2189 negation of where the condition IS allowed.
2191 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2193 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2194 header file does not have a version number, so I've had to invent a new
2195 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2196 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2197 radiusclient), but it was contributed by a Radius user.
2199 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2200 files or directories, for OpenSSL.
2202 PH/19 When an Exim process that is running as root has to create an Exim log
2203 file, it does so in a subprocess that runs as exim:exim so as to get the
2204 ownership right at creation (otherwise, other Exim processes might see
2205 the file with the wrong ownership). There was no test for failure of this
2206 fork() call, which would lead to the process getting stuck as it waited
2207 for a non-existent subprocess. Forks do occasionally fail when resources
2208 run out. I reviewed all the other calls to fork(); they all seem to check
2211 PH/20 When checking for unexpected SMTP input at connect time (before writing
2212 the banner), Exim was not dealing correctly with a non-positive return
2213 from the read() function. If the client had disconnected by this time,
2214 the result was a log entry for a synchronization error with an empty
2215 string after "input=" when read() returned zero. If read() returned -1
2216 (an event I could not check), uninitialized data bytes were printed.
2217 There were reports of junk text (parts of files, etc) appearing after
2220 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2222 PH/22 Added support for macro redefinition, and (re)definition in between
2223 driver and ACL definitions.
2225 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2226 forgetting to use the resulting value; it was using the unexpanded value.
2228 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2229 hadn't been configured. The fix is from Juergen Kreileder, who
2230 understands it better than I do:
2232 "Here's what I see happening with three configured cyrus_sasl
2233 authenticators configured (plain, login, cram-md5):
2235 On startup auth_cyrus_sasl_init() gets called for each of these.
2236 This means three calls to sasl_listmech() without a specified mech_list.
2237 => SASL tests which mechs of all available mechs actually work
2238 => three warnings about OTP not working
2239 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2241 With the patch, sasl_listmech() also gets called three times. But now
2242 SASL's mech_list option is set to the server_mech specified in the the
2243 authenticator. Or in other words, the answer from sasl_listmech()
2244 gets limited to just the mech you're testing for (which is different
2246 => the return list contains just 'plain' or 'login', 'cram-md5' or
2247 nothing depending on the value of ob->server_mech.
2249 I've just tested the patch: Authentication still works fine,
2250 unavailable mechs specified in the exim configuration are still
2251 caught, and the auth.log warnings about OTP are gone."
2253 PH/25 When debugging is enabled, the contents of the command line are added
2254 to the debugging output, even when log_selector=+arguments is not
2257 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2258 answer is "GNU", and only if the return is "GNU/something" is the answer
2261 PH/27 $acl_verify_message is now set immediately after the failure of a
2262 verification in an ACL, and so is available in subsequent modifiers. In
2263 particular, the message can be preserved by coding like this:
2265 warn !verify = sender
2266 set acl_m0 = $acl_verify_message
2268 Previously, $acl_verify_message was set only while expanding "message"
2269 and "log_message" when a very denied access.
2271 PH/28 Modified OS/os.c-Linux with
2273 -#ifndef OS_LOAD_AVERAGE
2274 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2276 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2277 nomenclature these days.)
2279 PH/29 Installed patch from the Sieve maintainer that adds the options
2280 sieve_useraddress and sieve_subaddress to the redirect router.
2282 PH/30 In these circumstances:
2283 . Two addresses routed to the same list of hosts;
2284 . First host does not offer TLS;
2285 . First host accepts first address;
2286 . First host gives temporary error to second address;
2287 . Second host offers TLS and a TLS session is established;
2288 . Second host accepts second address.
2289 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2290 and peerdn, if requested) that were in fact used only for the second
2293 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2294 attention to any local part prefix or suffix that was matched by the
2295 router that accepted the address. It now behaves in the same way as it
2296 does for delivery: the affixes are removed from the local part unless
2297 rcpt_include_affixes is set on the transport.
2299 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2300 timeout during the DATA phase of an incoming message.
2302 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2303 applied a patch sanctioned by the Sieve maintainer.
2305 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2306 the uid or gid is negative. A case of a negative gid caused this to be
2307 noticed. The fix allows for either to be negative.
2309 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2310 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2313 PH/36 The change PH/12 above was broken. Fixed it.
2315 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2316 the grounds that routing the same address twice would always produce the
2317 same answer. This might have been true once, but it is certainly no
2318 longer true now. Routing a child address may depend on the previous
2319 routing that produced that child. Some complicated redirection strategies
2320 went wrong when messages had multiple recipients, and made Exim's
2321 behaviour dependent on the order in which the addresses were given.
2323 I have moved the duplicate checking until after the routing is complete.
2324 Exim scans the addresses that are assigned to local and remote
2325 transports, and removes any duplicates. This means that more work will be
2326 done, as duplicates will always all be routed, but duplicates are
2327 presumably rare, so I don't expect this is of any significance.
2329 For deliveries to pipes, files, and autoreplies, the duplicate checking
2330 still happens during the routing process, since they are not going to be
2333 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2334 It corrects a timeout issue with spamd. This is Ian's comment: "The
2335 background is that sometimes spamd either never reads data from a
2336 connection it has accepted, or it never writes response data. The exiscan
2337 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2338 blindly assumes that writes won't block so it may never time out."
2340 PH/39 Allow G after quota size as well as K and M.
2342 PH/40 The value set for $authenticated_id in an authenticator may not contain
2343 binary zeroes or newlines because the value is written to log lines and
2344 to spool files. There was no check on this. Now the value is run through
2345 the string_printing() function so that such characters are converted to
2346 printable escape sequences.
2348 PH/41 $message_linecount is a new variable that contains the total number of
2349 lines in the message. Compare $body_linecount, which is the count for the
2352 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2353 bounce and delay warning messages, except in certain special cases, which
2356 (a) An SMTP error message from a remote host;
2357 (b) A message specified in a :fail: redirection;
2358 (c) A message specified in a "fail" command in a system filter;
2359 (d) A message specified in a FAIL return from the queryprogram router;
2360 (e) A message specified by the cannot_route_message router option.
2362 In these cases only, Exim does include the error details in bounce and
2363 warning messages. There are also a few cases where bland messages such
2364 as "unrouteable address" or "local delivery error" are given.
2366 PH/43 $value is now also set for the "else" part of a ${run expansion.
2368 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2369 being worked on, but at least Exim now implements the latest version to
2372 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2373 process to complete was treated as a delivery failure, a timeout while
2374 writing the message to the pipe was logged, but erroneously treated as a
2375 successful delivery. Such timeouts include transport filter timeouts. For
2376 consistency with the overall process timeout, these timeouts are now
2377 treated as errors, giving rise to delivery failures by default. However,
2378 there is now a new Boolean option for the pipe transport called
2379 timeout_defer, which, if set TRUE, converts the failures into defers for
2380 both kinds of timeout. A transport filter timeout is now identified in
2383 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2384 systems where "make" and "gmake" are different, calling "gmake" at top
2385 level broke things. I've arranged for the value of $(MAKE) to be passed
2386 from the Makefile to this script so that it can call the same version of
2390 A note about Exim versions 4.44 and 4.50
2391 ----------------------------------------
2393 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2394 changes of various kinds. As a consequence, a big documentation update was
2395 needed. This delayed the release for rather longer than seemed good, especially
2396 in the light of a couple of (minor) security issues. Therefore, the changes
2397 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2398 release. So 4.44 and 4.50 are in effect two different branches that both start
2401 I have left the 4.50 change log unchanged; it contains all the changes since
2402 4.43. The change log for 4.44 is below; many of its items are identical to
2403 those for 4.50. This seems to be the most sensible way to preserve the
2404 historical information.
2410 1. Minor wording change to the doc/README.SIEVE file.
2412 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2413 computation of the current number of files was incorrect.
2415 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2416 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2419 4. Give more explanation in the error message when the command for a transport
2420 filter fails to execute.
2422 5. There are several places where Exim runs a non-Exim command in a
2423 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2424 was being done only for the command run by the queryprogram router. It is
2425 now done for all such subprocesses. The other cases are: ${run, transport
2426 filters, and the commands run by the lmtp and pipe transports.
2428 6. Added CONFIGURE_GROUP build-time option.
2430 7. Some older OS have a limit of 256 on the maximum number of file
2431 descriptors. Exim was using setrlimit() to set 1000 as a large value
2432 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2433 systems. I've change it so that if it can't get 1000, it tries for 256.
2435 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2436 was an oversight, and furthermore, ever since the addition of extra
2437 controls (e.g. 4.43/32), the checks on when to allow different forms of
2438 "control" were broken. There should now be diagnostics for all cases when a
2439 control that does not make sense is encountered.
2441 9. Added the /retain_sender option to "control=submission".
2443 10. $recipients is now available in the predata ACL (oversight).
2445 11. Tidy the search cache before the fork to do a delivery from a message
2446 received from the command line. Otherwise the child will trigger a lookup
2447 failure and thereby defer the delivery if it tries to use (for example) a
2448 cached ldap connection that the parent has called unbind on.
2450 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2451 of $address_data from the recipient verification was clobbered by the
2452 sender verification.
2454 13. The value of address_data from a sender verification is now available in
2455 $sender_address_data in subsequent conditions in the ACL statement.
2457 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2459 15. Added a new option "connect=<time>" to callout options, to set a different
2462 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2463 was its contents. (It was OK if the option was not defined at all.)
2465 17. A "Completed" log line is now written for messages that are removed from
2466 the spool by the -Mrm option.
2468 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2469 information about exactly what failed.
2471 19. Added -dd to debug only the daemon process.
2473 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2474 handles timeouts, both on the server side and network timeouts. Renamed the
2475 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2477 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2478 It is now set to "smtps".
2480 22. $host_address is now set to the target address during the checking of
2481 ignore_target_hosts.
2483 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2484 being passed; this would have caused $sender_host_name to have been used if
2485 matching the list had actually called for a host name (not very likely,
2486 since this list is usually IP addresses). A host name is now passed as
2489 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2490 code that specifies a non-restarting handler (typically sigaction() in
2491 modern systems) in an attempt to fix a rare and obscure crash bug.
2493 25. Narrowed the window for a race in the daemon that could cause it to ignore
2494 SIGCHLD signals. This is not a major problem, because they are used only to
2495 wake it up if nothing else does.
2497 26. A malformed maildirsize file could cause Exim to calculate negative values
2498 for the mailbox size or file count. Odd effects could occur as a result.
2499 The maildirsize information is now recalculated if the size or filecount
2502 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2503 support for a long time. Removed HAVE_SYS_VFS_H.
2505 28. Installed the latest version of exipick from John Jetmore.
2507 29. In an address list, if the pattern was not a regular expression, an empty
2508 subject address (from a bounce message) matched only if the pattern was an
2509 empty string. Non-empty patterns were not even tested. This was the wrong
2510 because it is perfectly reasonable to use an empty address as part of a
2511 database query. An empty address is now tested by patterns that are
2512 lookups. However, all the other forms of pattern expect the subject to
2513 contain a local part and a domain, and therefore, for them, an empty
2514 address still always fails if the pattern is not itself empty.
2516 30. Exim went into a mad DNS loop when attempting to do a callout where the
2517 host was specified on an smtp transport, and looking it up yielded more
2518 than one IP address.
2520 31. Re-factored the code for checking spool and log partition space into a
2521 function that finds that data and another that does the check. The former
2522 is then used to implement four new variables: $spool_space, $log_space,
2523 $spool_inodes, and $log_inodes.
2525 32. The RFC2047 encoding function was originally intended for short strings
2526 such as real names; it was not keeping to the 75-character limit for
2527 encoded words that the RFC imposes. It now respects the limit, and
2528 generates multiple encoded words if necessary. To be on the safe side, I
2529 have increased the buffer size for the ${rfc2047: expansion operator from
2532 33. It is now permitted to omit both strings after an "if" condition; if the
2533 condition is true, the result is "true". As before, when the second string
2534 is omitted, a false condition yields an empty string. This makes it less
2535 cumbersome to write custom ACL and router conditions.
2537 34. Failure to deliver a bounce message always caused it to be frozen, even if
2538 there was an errors_to setting on the router. The errors_to setting is now
2541 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2542 canonical form (fully expanded) before being placed in
2543 $sender_host_address.
2545 36. The table in the code that translates DNS record types into text (T_A to
2546 "A" for instance) was missing entries for NS and CNAME. It is just possible
2547 that this could have caused confusion if both these types were looked up
2548 for the same domain, because the text type is used as part of Exim's
2549 per-process caching. But the chance of anyone hitting this buglet seems
2552 37. The dnsdb lookup has been extended in a number of ways.
2554 (1) There is a new type, "zns", which walks up the domain tree until it
2555 finds some nameserver records. It should be used with care.
2557 (2) There is a new type, "mxh", which is like "mx" except that it returns
2558 just the host names, not the priorities.
2560 (3) It is now possible to give a list of domains (or IP addresses) to be
2561 looked up. The behaviour when one of the lookups defers can be
2562 controlled by a keyword.
2564 (4) It is now possible to specify the separator character for use when
2565 multiple records are returned.
2567 38. The dnslists ACL condition has been extended: it is now possible to supply
2568 a list of IP addresses and/or domains to be looked up in a particular DNS
2571 39. Added log_selector=+queue_time_overall.
2573 40. When running the queue in the test harness, wait just a tad after forking a
2574 delivery process, to get repeatability of debugging output.
2576 41. Include certificate and key file names in error message when GnuTLS fails
2577 to set them up, because the GnuTLS error message doesn't include the name
2578 of the failing file when there is a problem reading it.
2580 42. Allow both -bf and -bF in the same test run.
2582 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2584 44. The "Exiscan patch" is now merged into the mainline Exim source.
2586 45. Sometimes the final signoff response after QUIT could fail to get
2587 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2588 < 0 before doing a fflush(). This bug looks as though it goes back to the
2589 introduction of TLS in release 3.20, but "sometimes" must have been rare
2590 because the tests only now provoked it.
2592 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2593 (this can affect the format of dates).
2595 47. exim_tidydb, when checking for the continued existence of a message for
2596 which it has found a message-specific retry record, was not finding
2597 messages that were in split spool directories. Consequently, it was
2598 deleting retry records that should have stayed in existence.
2600 48. Steve fixed some bugs in eximstats.
2602 49. The SPA authentication driver was not abandoning authentication and moving
2603 on to the next authenticator when an expansion was forced to fail,
2604 contradicting the general specification for all authenticators. Instead it
2605 was generating a temporary error. It now behaves as specified.
2607 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2608 (the order specifies the preference for clients). The order is now AES256,
2609 AES128, 3DES, ARCFOUR128.
2611 51. Small patch to Sieve code - explicitly set From: when generating an
2614 52. Exim crashed if a remote delivery caused a very long error message to be
2615 recorded - for instance if somebody sent an entire SpamAssassin report back
2616 as a large number of 550 error lines. This bug was coincidentally fixed by
2617 increasing the size of one of Exim's internal buffers (big_buffer) that
2618 happened as part of the Exiscan merge. However, to be on the safe side, I
2619 have made the code more robust (and fixed the comments that describe what
2622 53. Now that there can be additional text after "Completed" in log lines (if
2623 the queue_time_overall log selector is set), a one-byte patch to exigrep
2624 was needed to allow it to recognize "Completed" as not the last thing in
2627 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2628 patch that reportedly fixes this has been added. I am not expert enough to
2629 create a test for it. This is what the patch creator wrote:
2631 "I found a little strange behaviour of ldap code when working with
2632 Windows 2003 AD Domain, where users was placed in more than one
2633 Organization Units. When I tried to give exim partial DN, the exit code
2634 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2635 But simultaneously result of request was absolutely normal ldap result,
2636 so I produce this patch..."
2638 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2639 REFERENCE, so I have modified the code to exclude the patch when that macro
2642 55. Some experimental protocols are using DNS PTR records for new purposes. The
2643 keys for these records are domain names, not reversed IP addresses. The
2644 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2645 leaves it alone. Component reversal etc. now happens only for IP addresses.
2648 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2650 57. Double the size of the debug message buffer (to 2048) so that more of very
2651 long debug lines gets shown.
2653 58. The exicyclog utility now does better if the number of log files to keep
2654 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2656 59. Two changes related to the smtp_active_hostname option:
2658 (1) $smtp_active_hostname is now available as a variable.
2659 (2) The default for smtp_banner uses $smtp_active_hostname instead
2660 of $primary_hostname.
2662 60. The host_aton() function is supposed to be passed a string that is known
2663 to be a valid IP address. However, in the case of IPv6 addresses, it was
2664 not checking this. This is a hostage to fortune. Exim now panics and dies
2665 if the condition is not met. A case was found where this could be provoked
2666 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2667 components; fortuitously, this particular loophole had already been fixed
2668 by change 4.50/55 above.
2670 If there are any other similar loopholes, the new check in host_aton()
2671 itself should stop them being exploited. The report I received stated that
2672 data on the command line could provoke the exploit when Exim was running as
2673 exim, but did not say which command line option was involved. All I could
2674 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2675 running as the user.
2678 61. There was a buffer overflow vulnerability in the SPA authentication code
2679 (which came originally from the Samba project). I have added a test to the
2680 spa_base64_to_bits() function which I hope fixes it.
2683 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2684 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2686 63. The daemon start-up calls getloadavg() while still root for those OS that
2687 need the first call to be done as root, but it missed one case: when
2688 deliver_queue_load_max is set with deliver_drop_privilege. This is
2689 necessary for the benefit of the queue runner, because there is no re-exec
2690 when deliver_drop_privilege is set.
2692 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2693 This has been fixed.
2695 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2696 was in use, was not putting the data itself into the right store pool;
2697 consequently, it could be overwritten for a subsequent message in the same
2698 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2701 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2703 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2704 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2705 the function were treating the return as a boolean value, which happened to
2706 work because 0=false and not-0=true, but is not correct code.
2708 68. The host_aton() function was not handling scoped IPv6 addresses (those
2709 with, for example, "%eth0" on the end) correctly.
2711 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2712 negated items (that is, ~something) in unsigned ints. Some compilers
2713 apparently mutter when there is no cast.
2715 70. If an address verification called from an ACL failed, and did not produce a
2716 user-specific message (i.e. there was only a "system" message), nothing was
2717 put in $acl_verify_message. In this situation, it now puts the system
2720 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2721 session; change 4.31/43 added the unwanted input to the log line - except
2722 that it did not do this in the start of session case. It now does.
2724 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2725 This is wrong and can cause the other end to generate a synchronization
2726 error if it is another Exim or anything else that does the synchronization
2727 check. A QUIT command is no longer sent after a timeout.
2729 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2730 during host lookups.
2732 74. The defer_ok option of callout verification was not working if it was used
2733 when verifying addresses in header lines, that is, for this case:
2735 verify = header_sender/callout=defer_ok
2737 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2738 those file descriptors could be used for SMTP connections. If anything
2739 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2740 could be sent to the SMTP client, causing chaos. The daemon now opens
2741 stdin, stdout, and stderr to /dev/null when it puts itself into the
2744 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2745 log by default. The user can override this with suitable Perl magic.
2747 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2748 the log message when discard triggers, was not working for the DATA ACL or
2749 for the non-SMTP ACL.
2751 78. Error message wording change in sieve.c.
2753 79. If smtp_accept_max_per_host was set, the number of connections could be
2754 restricted to fewer than expected, because the daemon was trying to set up
2755 a new connection before checking whether the processes handling previous
2756 connections had finished. The check for completed processes is now done
2757 earlier. On busy systems, this bug wouldn't be noticed because something
2758 else would have woken the daemon, and it would have reaped the completed
2761 80. If a message was submitted locally by a user whose login name contained one
2762 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2763 It caused a spool format error. I have fixed the spool reading code. A
2764 related problem was that the "from" clause in the Received: line became
2765 illegal because of the space(s). It is now covered by ${quote_local_part.
2767 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2770 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2772 83. Patch from Sieve maintainer to fix unterminated string problem in
2773 "vacation" handling.
2775 84. Some minor changes to the Linux configuration files to help with other
2776 OS variants using glibc.
2778 85. One more patch for Sieve to update vacation handling to latest spec.
2781 ----------------------------------------------------
2782 See the note above about the 4.44 and 4.50 releases.
2783 ----------------------------------------------------
2789 1. Change 4.43/35 introduced a bug that caused file counts to be
2790 incorrectly computed when quota_filecount was set in an appendfile
2793 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2794 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2797 3. Give more explanation in the error message when the command for a transport
2798 filter fails to execute.
2800 4. There are several places where Exim runs a non-Exim command in a
2801 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2802 was being done only for the command run by the queryprogram router. It is
2803 now done for all such subprocesses. The other cases are: ${run, transport
2804 filters, and the commands run by the lmtp and pipe transports.
2806 5. Some older OS have a limit of 256 on the maximum number of file
2807 descriptors. Exim was using setrlimit() to set 1000 as a large value
2808 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2809 systems. I've change it so that if it can't get 1000, it tries for 256.
2811 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2812 was an oversight, and furthermore, ever since the addition of extra
2813 controls (e.g. 4.43/32), the checks on when to allow different forms of
2814 "control" were broken. There should now be diagnostics for all cases when a
2815 control that does not make sense is encountered.
2817 7. $recipients is now available in the predata ACL (oversight).
2819 8. Tidy the search cache before the fork to do a delivery from a message
2820 received from the command line. Otherwise the child will trigger a lookup
2821 failure and thereby defer the delivery if it tries to use (for example) a
2822 cached ldap connection that the parent has called unbind on.
2824 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2825 of $address_data from the recipient verification was clobbered by the
2826 sender verification.
2828 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2829 was its contents. (It was OK if the option was not defined at all.)
2831 11. A "Completed" log line is now written for messages that are removed from
2832 the spool by the -Mrm option.
2834 12. $host_address is now set to the target address during the checking of
2835 ignore_target_hosts.
2837 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2838 being passed; this would have caused $sender_host_name to have been used if
2839 matching the list had actually called for a host name (not very likely,
2840 since this list is usually IP addresses). A host name is now passed as
2843 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2844 code that specifies a non-restarting handler (typically sigaction() in
2845 modern systems) in an attempt to fix a rare and obscure crash bug.
2847 15. Narrowed the window for a race in the daemon that could cause it to ignore
2848 SIGCHLD signals. This is not a major problem, because they are used only to
2849 wake it up if nothing else does.
2851 16. A malformed maildirsize file could cause Exim to calculate negative values
2852 for the mailbox size or file count. Odd effects could occur as a result.
2853 The maildirsize information is now recalculated if the size or filecount
2856 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2857 support for a long time. Removed HAVE_SYS_VFS_H.
2859 18. Updated exipick to current release from John Jetmore.
2861 19. Allow an empty sender to be matched against a lookup in an address list.
2862 Previously the only cases considered were a regular expression, or an
2865 20. Exim went into a mad DNS lookup loop when doing a callout where the
2866 host was specified on the transport, if the DNS lookup yielded more than
2869 21. The RFC2047 encoding function was originally intended for short strings
2870 such as real names; it was not keeping to the 75-character limit for
2871 encoded words that the RFC imposes. It now respects the limit, and
2872 generates multiple encoded words if necessary. To be on the safe side, I
2873 have increased the buffer size for the ${rfc2047: expansion operator from
2876 22. Failure to deliver a bounce message always caused it to be frozen, even if
2877 there was an errors_to setting on the router. The errors_to setting is now
2880 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2881 canonical form (fully expanded) before being placed in
2882 $sender_host_address.
2884 24. Updated eximstats to version 1.33
2886 25. Include certificate and key file names in error message when GnuTLS fails
2887 to set them up, because the GnuTLS error message doesn't include the name
2888 of the failing file when there is a problem reading it.
2890 26. Expand error message when OpenSSL has problems setting up cert/key files.
2893 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2894 (this can affect the format of dates).
2896 28. exim_tidydb, when checking for the continued existence of a message for
2897 which it has found a message-specific retry record, was not finding
2898 messages that were in split spool directories. Consequently, it was
2899 deleting retry records that should have stayed in existence.
2901 29. eximstats updated to version 1.35
2902 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2903 1.35 - bugfix such that pie charts by volume are generated correctly
2905 30. The SPA authentication driver was not abandoning authentication and moving
2906 on to the next authenticator when an expansion was forced to fail,
2907 contradicting the general specification for all authenticators. Instead it
2908 was generating a temporary error. It now behaves as specified.
2910 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2911 (the order specifies the preference for clients). The order is now AES256,
2912 AES128, 3DES, ARCFOUR128.
2914 31. Small patch to Sieve code - explicitly set From: when generating an
2917 32. Exim crashed if a remote delivery caused a very long error message to be
2918 recorded - for instance if somebody sent an entire SpamAssassin report back
2919 as a large number of 550 error lines. This bug was coincidentally fixed by
2920 increasing the size of one of Exim's internal buffers (big_buffer) that
2921 happened as part of the Exiscan merge. However, to be on the safe side, I
2922 have made the code more robust (and fixed the comments that describe what
2925 33. Some experimental protocols are using DNS PTR records for new purposes. The
2926 keys for these records are domain names, not reversed IP addresses. The
2927 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2928 leaves it alone. Component reversal etc. now happens only for IP addresses.
2931 34. The host_aton() function is supposed to be passed a string that is known
2932 to be a valid IP address. However, in the case of IPv6 addresses, it was
2933 not checking this. This is a hostage to fortune. Exim now panics and dies
2934 if the condition is not met. A case was found where this could be provoked
2935 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2936 components; fortuitously, this particular loophole had already been fixed
2937 by change 4.50/55 or 4.44/33 above.
2939 If there are any other similar loopholes, the new check in host_aton()
2940 itself should stop them being exploited. The report I received stated that
2941 data on the command line could provoke the exploit when Exim was running as
2942 exim, but did not say which command line option was involved. All I could
2943 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2944 running as the user.
2947 35. There was a buffer overflow vulnerability in the SPA authentication code
2948 (which came originally from the Samba project). I have added a test to the
2949 spa_base64_to_bits() function which I hope fixes it.
2952 36. The daemon start-up calls getloadavg() while still root for those OS that
2953 need the first call to be done as root, but it missed one case: when
2954 deliver_queue_load_max is set with deliver_drop_privilege. This is
2955 necessary for the benefit of the queue runner, because there is no re-exec
2956 when deliver_drop_privilege is set.
2958 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2959 was in use, was not putting the data itself into the right store pool;
2960 consequently, it could be overwritten for a subsequent message in the same
2961 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2964 38. Sometimes the final signoff response after QUIT could fail to get
2965 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2966 < 0 before doing a fflush(). This bug looks as though it goes back to the
2967 introduction of TLS in release 3.20, but "sometimes" must have been rare
2968 because the tests only now provoked it.
2974 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2975 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2976 Now it returns an integer. A number of calls were still expecting a T/F
2977 return. Fortuitously, in all cases, the tests worked in OK situations,
2978 which is the norm. However, things would have gone wrong on any write
2979 failures on the smtp file descriptor. This function is used when sending
2980 messages over SMTP and also when doing verify callouts.
2982 2. When Exim is called to do synchronous delivery of a locally submitted
2983 message (the -odf or -odi options), it no longer closes stderr before doing
2986 3. Implemented the mua_wrapper option.
2988 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2990 5. Implemented the functions header_remove(), header_testname(),
2991 header_add_at_position(), and receive_remove_recipient(), and exported them
2994 6. If an ACL "warn" statement specified the addition of headers, Exim already
2995 inserted X-ACL-Warn: at the start if there was no header name. However, it
2996 was not making this test for the second and subsequent header lines if
2997 there were newlines in the string. This meant that an invalid header could
2998 be inserted if Exim was badly configured.
3000 7. Allow an ACL "warn" statement to add header lines at the start or after all
3001 the Received: headers, as well as at the end.
3003 8. Added the rcpt_4xx retry error code.
3005 9. Added postmaster_mailfrom=xxx to callout verification option.
3007 10. Added mailfrom=xxxx to the callout verification option, for verify=
3010 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3011 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3013 12. Inserted some casts to stop certain compilers warning when using pointer
3014 differences as field lengths or precisions in printf-type calls (mostly
3015 affecting debugging statements).
3017 13. Added optional readline() support for -be (dynamically loaded).
3019 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3020 same clock tick as a message's arrival, so that its received time was the
3021 same as the "first fail" time on the retry record, and that message
3022 remained on the queue past the ultimate address timeout, every queue runner
3023 would try a delivery (because it was past the ultimate address timeout) but
3024 after another failure, the ultimate address timeout, which should have then
3025 bounced the address, did not kick in. This was a "< instead of <=" error;
3026 in most cases the first failure would have been in the next clock tick
3027 after the received time, and all would be well.
3029 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3030 being recognized when the domain list was tested by the match_domain
3031 condition in an expansion string.
3033 16. Added the ${str2b64: operator.
3035 17. Exim was always calling setrlimit() to set a large limit for the number of
3036 processes, without checking whether the existing limit was already
3037 adequate. (It did check for the limit on file descriptors.) Furthermore,
3038 errors from getrlimit() and setrlimit() were being ignored. Now they are
3039 logged to the main and panic logs, but Exim does carry on, to try to do its
3040 job under whatever limits there are.
3042 18. Imported PCRE 5.0.
3044 19. Trivial typo in log message " temporarily refused connection" (the leading
3047 20. If the log selector return_path_on_delivery was set and an address was
3048 redirected to /dev/null, the delivery process crashed because it assumed
3049 that a return path would always be set for a "successful" delivery. In this
3050 case, the whole delivery is bypassed as an optimization, and therefore no
3053 21. Internal re-arrangement: the function for sending a challenge and reading
3054 a response while authentication was assuming a zero-terminated challenge
3055 string. It's now changed to take a pointer and a length, to allow for
3056 binary data in such strings.
3058 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3060 23. Exim was not respecting finduser_retries when seeking the login of the
3061 uid under which it was called; it was always trying 10 times. (The default
3062 setting of finduser_retries is zero.) Also, it was sleeping after the final
3063 failure, which is pointless.
3065 24. Implemented tls_on_connect_ports.
3067 25. Implemented acl_smtp_predata.
3069 26. If the domain in control=submission is set empty, Exim assumes that the
3070 authenticated id is a complete email address when it generates From: or
3071 Sender: header lines.
3073 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3074 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3075 chown and chgrp in /bin and hostname in /usr/bin.
3077 28. Exim was keeping the "process log" file open after each use, just as it
3078 does for the main log. This opens the possibility of it remaining open for
3079 long periods when the USR1 signal hits a daemon. Occasional processlog
3080 errors were reported, that could have been caused by this. Anyway, it seems
3081 much more sensible not to leave this file open at all, so that is what now
3084 29. The long-running daemon process does not normally write to the log once it
3085 has entered its main loop, and it closes the log before doing so. This is
3086 so that log files can straightforwardly be renamed and moved. However,
3087 there are a couple of unusual error situations where the daemon does write
3088 log entries, and I had neglected to close the log afterwards.
3090 30. The text of an SMTP error response that was received during a remote
3091 delivery was being truncated at 512 bytes. This is too short for some of
3092 the long messages that one sometimes sees. I've increased the limit to
3095 31. It is now possible to make retry rules that apply only when a message has a
3096 specific sender, in particular, an empty sender.
3098 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3099 it possible to be selective about when SMTP synchronization is enforced.
3101 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3103 32. Implemented hosts_connection_nolog.
3105 33. Added an ACL for QUIT.
3107 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3110 35. Added mailbox_size and mailbox_filecount to appendfile.
3112 36. Added control = no_multiline_responses to ACLs.
3114 37. There was a bug in the logic of the code that waits for the clock to tick
3115 in the case where the clock went backwards by a substantial amount such
3116 that the microsecond fraction of "now" was more than the microsecond
3117 fraction of "then" (but the whole seconds number was less).
3119 38. Added support for the libradius Radius client library this is found on
3120 FreeBSD (previously only the radiusclient library was supported).
3126 1. When certain lookups returned multiple values in the form name=value, the
3127 quoting of the values was not always being done properly. Specifically:
3128 (a) If the value started with a double quote, but contained no whitespace,
3130 (b) If the value contained whitespace other than a space character (i.e.
3131 tabs or newlines or carriage returns) it was not quoted.
3132 This fix has been applied to the mysql and pgsql lookups by writing a
3133 separate quoting function and calling it from the lookup code. The fix
3134 should probably also be applied to nisplus, ibase and oracle lookups, but
3135 since I cannot test any of those, I have not disturbed their existing code.
3137 2. A hit in the callout cache for a specific address caused a log line with no
3138 reason for rejecting RCPT. Now it says "Previous (cached) callout
3139 verification failure".
3141 3. There was an off-by-one bug in the queryprogram router. An over-long
3142 return line was truncated at 256 instead of 255 characters, thereby
3143 overflowing its buffer with the terminating zero. As well as fixing this, I
3144 have increased the buffer size to 1024 (and made a note to document this).
3146 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3147 when Exim is waiting for an SMTP response from a remote server, Exim
3148 restarts its select() call on the socket, thereby resetting its timeout.
3149 This is not a problem when such interrupts are rare. Somebody set up a cron
3150 job to run exiwhat every 2 minutes, which is less than the normal select()
3151 timeout (5 or 10 minutes). This meant that the select() timeout never
3152 kicked in because it was always reset. I have fixed this by comparing the
3153 time when an interrupt arrives with the time at the start of the first call
3154 to select(). If more time than the timeout has elapsed, the interrupt is
3155 treated as a timeout.
3157 5. Some internal re-factoring in preparation for the addition of Sieve
3158 extensions (by MH). In particular, the "personal" test is moved to a
3159 separate function, and given an option for scanning Cc: and Bcc: (which is
3160 not set for Exim filters).
3162 6. When Exim created an email address using the login of the caller as the
3163 local part (e.g. when creating a From: or Sender: header line), it was not
3164 quoting the local part when it contained special characters such as @.
3166 7. Installed new OpenBSD configuration files.
3168 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3169 try to make them clearer.
3171 9. Callout options, other than the timeout value, were being ignored when
3172 verifying sender addresses in header lines. For example, when using
3174 verify = header_sender/callout=no_cache
3176 the cache was (incorrectly) being used.
3178 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3179 only the Cygwin environment.
3181 11. When return_path_on_delivery was set as a log selector, if different remote
3182 addresses in the same message used different return paths and parallel
3183 remote delivery occurred, the wrong values would sometimes be logged.
3184 (Whenever a remote delivery process finished, the return path value from
3185 the most recently started remote delivery process was logged.)
3187 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3188 lines when AUTH and/or TLS are in use. This is the "received protocol"
3189 field. Exim used to use "asmtp" for authenticated SMTP, without any
3190 indication (in the protocol name) for TLS use. Now it follows the RFC and
3191 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3192 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3193 names appear in log lines as well as in Received: header lines.
3195 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3196 extensions, and comparison tests, and to fix some bugs.
3198 14. Changes to the "personal" filter test:
3200 (1) The test was buggy in that it was just doing the equivalent of
3201 "contains" tests on header lines. For example, if a user's address was
3202 anne@some.where, the "personal" test would incorrectly be true for
3204 To: susanne@some.where
3206 This test is now done by extracting each address from the header in turn,
3207 and checking the entire address. Other tests that are part of "personal"
3208 are now done using regular expressions (for example, to check local parts
3209 of addresses in From: header lines).
3211 (2) The list of non-personal local parts in From: addresses has been
3212 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3213 taken from the Sieve specification recommendations.
3215 (3) If the message contains any header line starting with "List-" it is
3216 treated as non-personal.
3218 (4) The test for "circular" in the Subject: header line has been removed
3219 because it now seems ill-conceived.
3221 15. Minor typos in src/EDITME comments corrected.
3223 16. Installed latest exipick from John Jetmore.
3225 17. If headers_add on a router specified a text string that was too long for
3226 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3227 of string_sprintf() is now avoided.
3229 18. $message_body_size was not set (it was always zero) when running the DATA
3230 ACL and the local_scan() function.
3232 19. For the "mail" command in an Exim filter, no default was being set for
3233 the once_repeat time, causing a random time value to be used if "once" was
3234 specified. (If the value happened to be <= 0, no repeat happened.) The
3235 default is now 0s, meaning "never repeat". The "vacation" command was OK
3236 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3237 (I found it when inspecting the code).
3239 20. There is now an overall timeout for performing a callout verification. It
3240 defaults to 4 times the callout timeout, which applies to individual SMTP
3241 commands during the callout. The overall timeout applies when there is more
3242 than one host that can be tried. The timeout is checked before trying the
3243 next host. This prevents very long delays if there are a large number of
3244 hosts and all are timing out (e.g. when the network connections are timing
3245 out). The value of the overall timeout can be changed by specifying an
3246 additional sub-option for "callout", called "maxwait". For example:
3248 verify = sender/callout=5s,maxwait=20s
3250 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3251 to the end before writing, but this should make it even safer).
3253 22. Exim was forgetting that it had advertised PIPELINING for the second and
3254 subsequent messages on an SMTP connection. It was also not resetting its
3255 memory on STARTTLS and an internal HELO.
3257 23. When Exim logs an SMTP synchronization error within a session, it now
3258 records whether PIPELINING has been advertised or not.
3260 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3261 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3262 rather than long int.
3264 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3266 26. Added the never_mail option to autoreply.
3272 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3273 crash if the getsockname() call failed; this can happen if a connection is
3274 closed very soon after it is established. The problem was simply in the
3275 order in which certain operations were done, causing Exim to try to write
3276 to the SMTP stream before it had set up the file descriptor. The bug has
3277 been fixed by making things happen in the correct order.
3283 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3284 before the connection was closed, thus losing the rejection response.
3286 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3287 some early Solaris releases, but causes trouble in current releases where
3288 socklen_t is defined.
3290 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3293 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3296 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3297 that releasing the top of it at the end releases what was used for sub-
3298 expansions (unless the block got too big). However, discard this block if
3299 the first thing is a variable or header, so that we can use its block when
3300 it is dynamic (useful for very large $message_headers, for example).
3302 6. Lookups now cache *every* query, not just the most recent. A new, separate
3303 store pool is used for this. It can be recovered when all lookup caches are
3304 flushed. Lookups now release memory at the end of their result strings.
3305 This has involved some general refactoring of the lookup sources.
3307 7. Some code has been added to the store_xxx() functions to reduce the amount
3308 of flapping under certain conditions.
3310 8. log_incoming_interface used to affect only the <= reception log lines. Now
3311 it causes the local interface and port to be added to several more SMTP log
3312 lines, for example "SMTP connection from", and rejection lines.
3314 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3316 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3318 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3319 could be overwritten at the end of the current message (or the start of a
3320 new message if it was set in a HELO ACL). The value is now preserved for
3321 the duration of the SMTP connection.
3323 12. If a transport had a headers_rewrite setting, and a matching header line
3324 contained an unqualified address, that address was qualified, even if it
3325 did not match any rewriting rules. The underlying bug was that the values
3326 of the flags that permit the existence of unqualified sender and recipient
3327 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3328 non-local messages, and by -bnq for local messages) were not being
3329 preserved with the message after it was received.
3331 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3332 "next input=" as part of the text comprising the host identity instead of
3333 the correct text. The code was using the same buffer for two different
3334 strings. However, depending on which order the printing function evaluated
3335 its arguments, the bug did not always show up. Under Linux, for example, my
3336 test suite worked just fine.
3338 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3339 doesn't work with some older versions of Perl. It has been changed to "my",
3340 which in any case is probably the better facility to use.
3342 15. A really picky compiler found some instances of statements for creating
3343 error messages that either had too many or two few arguments for the format
3346 16. The size of the buffer for calls to the DNS resolver has been increased
3347 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3348 for addresses that have a lot of PTR records. This alleviates a problem; it
3349 does not fully solve it.
3351 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3352 buffer now truncates the list and logs the incident, which is the same
3353 action as happens when Exim is looking up a host name and its aliases.
3354 Previously in this situation something unpredictable would happen;
3355 sometimes it was "internal error: store_reset failed".
3357 18. If a server dropped the connection unexpectedly when an Exim client was
3358 using GnuTLS and trying to read a response, the client delivery process
3359 crashed while trying to generate an error log message.
3361 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3362 string, for example:
3364 warn message = H1: something\nH2: something
3366 the text was added as a single header line from Exim's point of view
3367 though it ended up OK in the delivered message. However, searching for the
3368 second and subsequent header lines using $h_h2: did not work. This has been
3369 fixed. Similarly, if a system filter added multiple headers in this way,
3370 the routers could not see them.
3372 20. Expanded the error message when iplsearch is called with an invalid key to
3373 suggest using net-iplsearch in a host list.
3375 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3376 ACLs are no longer actually imposed (and a message to that effect is
3379 22. If a "gecos" field in a passwd entry contained escaped characters, in
3380 particular, if it contained a \" sequence, Exim got it wrong when building
3381 a From: or a Sender: header from that name. A second bug also caused
3382 incorrect handling when an unquoted " was present following a character
3383 that needed quoting.
3385 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3386 was not being matched caselessly.
3388 24. Arranged for all hyphens in the exim.8 source to be escaped with
3391 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3392 properly. Recipient callout cache records were still being keyed to include
3393 the sender, even when use_sender was set false. This led to far more
3394 callouts that were necessary. The sender is no longer included in the key
3395 when use_sender is false.
3397 26. Added "control = submission" modifier to ACLs.
3399 27. Added the ${base62d: operator to decode base 62 numbers.
3401 28. dnsdb lookups can now access SRV records.
3403 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3404 the configuration file.
3406 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3407 (-v) mode. This makes the output for a verbose queue run more intelligible.
3409 31. Added a use_postmaster feature to recipient callouts.
3411 32. Added the $body_zerocount variable, containing the number of binary zero
3412 bytes in the message body.
3414 33. The time of last modification of the "new" subdirectory is now used as the
3415 "mailbox time last read" when there is a quota error for a maildir
3418 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3420 35. Added +ignore_unknown as a special item in host lists.
3422 36. Code for decoding IPv6 addresses in host lists is now included, even if
3423 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3424 address was recognized as an IP address, but was then not correctly decoded
3425 into binary, causing unexpected and incorrect effects when compared with
3432 1. Very minor rewording of debugging text in manualroute to say "list of
3433 hosts" instead of "hostlist".
3435 2. If verify=header_syntax was set, and a header line with an unqualified
3436 address (no domain) and a large number of spaces between the end of the
3437 name and the colon was received, the reception process suffered a buffer
3438 overflow, and (when I tested it) crashed. This was caused by some obsolete
3439 code that should have been removed. The fix is to remove it!
3441 3. When running in the test harness, delay a bit after writing a bounce
3442 message to get a bit more predictability in the log output.
3444 4. Added a call to search_tidyup() just before forking a reception process. In
3445 theory, someone could use a lookup in the expansion of smtp_accept_max_
3446 per_host which, without the tidyup, could leave open a database connection.
3448 5. Added the variables $recipient_data and $sender_data which get set from a
3449 lookup success in an ACL "recipients" or "senders" condition, or a router
3450 "senders" option, similar to $domain_data and $local_part_data.
3452 6. Moved the writing of debug_print from before to after the "senders" test
3455 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3456 problems for message scanning, either using a data ACL, or using
3457 local_scan() because the Received: header was not generated till after they
3458 were called (in order to set the time as the time of reception completion).
3459 I have revised the way this works. The header is now generated after the
3460 body is received, but before the ACL or local_scan() are called. After they
3461 are run, the timestamp in the header is updated.
3467 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3468 before starting a queue runner without re-exec. This happened only when
3469 deliver_drop_privilege was set or when the Exim user was set to root. The
3470 effect of the bug was that timeouts during subsequent deliveries caused
3471 crashes instead of being properly handled. The handler is now left at its
3472 default (and expected) setting.
3474 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3475 message, again when deliver_drop_privilege is set or Exim is run as root.
3476 The bug described in (1) was not present in this case, but the tidying up
3477 of the other signals was missing. I have made the two cases consistent.
3479 3. The ignore_target_hosts setting on a manualroute router was being ignored
3480 for hosts that were looked up using the /MX notation.
3482 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3485 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3486 operated on the sender address. After changing the $sender_address to <>
3487 for the sender address verify, Exim was re-instated it as the original
3488 (before rewriting) address, but remembering that it had rewritten it, so it
3489 wasn't rewriting it again. This bug also had the effect of breaking the
3490 sender address verification caching when the sender address was rewritten.
3492 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3493 This has been changed so that if the ip literal address matches
3494 ignore_target_hosts, the router declines.
3496 7. Added expansion conditions match_domain, match_address, and match_local_
3497 part (NOT match_host).
3499 8. The placeholder for the Received: header didn't have a length field set.
3501 9. Added code to Exim itself and to exim_lock to test for a specific race
3502 condition that could lead to file corruption when using MBX delivery. The
3503 issue is with the lockfile that is created in /tmp. If this file is removed
3504 after a process has opened it but before that process has acquired a lock,
3505 there is the potential for a second process to recreate the file and also
3506 acquire a lock. This could lead to two Exim processes writing to the file
3507 at the same time. The added code performs the same test as UW imapd; it
3508 checks after acquiring the lock that its file descriptor still refers to
3509 the same named file.
3511 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3512 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3515 11. Added the smtp_active_hostname option. If used, this will typically be made
3516 to depend on the incoming interface address. Because $interface_address is
3517 not set up until the daemon has forked a reception process, error responses
3518 that can happen earlier (such as "too many connections") no longer contain
3521 12. If an expansion in a condition on a "warn" statement fails because a lookup
3522 defers, the "warn" statement is abandoned, and the next ACL statement is
3523 processed. Previously this caused the whole ACL to be aborted.
3525 13. Added the iplsearch lookup type.
3527 14. Added ident_timeout as a log selector.
3529 15. Added tls_certificate_verified as a log selector.
3531 16. Added a global option tls_require_ciphers (compare the smtp transport
3532 option of the same name). This controls incoming TLS connections.
3534 17. I finally figured out how to make tls_require_ciphers do a similar thing
3535 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3536 before starting the TLS session.
3538 18. Tabs are now shown as \t in -bP output.
3540 19. If the log selector return_path_on_delivery was set, Exim crashed when
3541 bouncing a message because it had too many Received: header lines.
3543 20. If two routers both had headers_remove settings, and the first one included
3544 a superfluous trailing colon, the final name in the first list and the
3545 first name in the second list were incorrectly joined into one item (with a
3546 colon in the middle).
3552 1. Added -C and -D options to the exinext utility, mainly to make it easier
3553 to include in the automated testing, but these could be helpful when
3554 multiple configurations are in use.
3556 2. The exinext utility was not formatting the output nicely when there was
3557 an alternate port involved in the retry record key, nor when there was a
3558 message id as well (for retries that were specific to a specific message
3559 and a specific host). It was also confused by IPv6 addresses, because of
3560 the additional colons they contain. I have fixed the IPv4 problem, and
3561 patched it up to do a reasonable job for IPv6.
3563 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3564 delivery, the log line now contains "pipelined" if PIPELINING was used.
3566 4. An SMTP transport process used to panic and die if the bind() call to set
3567 an explicit outgoing interface failed. This has been changed; it is now
3568 treated in the same way as a connect() failure.
3570 5. A reference to $sender_host_name in the part of a conditional expansion
3571 that was being skipped was still causing a DNS lookup. This no longer
3574 6. The def: expansion condition was not recognizing references to header lines
3575 that used bh_ and bheader_.
3577 7. Added the _cache feature to named lists.
3579 8. The code for checking quota_filecount in the appendfile transport was
3580 allowing one more file than it should have been.
3582 9. For compatibility with Sendmail, the command line option
3590 and sets the incoming protocol and host name (for trusted callers). The
3591 host name and its colon can be omitted when only the protocol is to be set.
3592 Note the Exim already has two private options, -pd and -ps, that refer to
3593 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3594 "s", but I don't think that's a major issue.
3596 10. A number of refactoring changes to the code, none of which should affect
3599 (a) The number of logging options was getting close to filling up the
3600 32-bit word that was used as a bit map. I have split them into two classes:
3601 those that are passed in the argument to log_write(), and those that are
3602 only ever tested independently outside of that function. These are now in
3603 separate 32-bit words, so there is plenty of room for expansion again.
3604 There is no change in the user interface or the logging behaviour.
3606 (b) When building, for example, log lines, the code previously used a
3607 macro that called string_cat() twice, in order to add two strings. This is
3608 not really sufficiently general. Furthermore, there was one instance where
3609 it was actually wrong because one of the argument was used twice, and in
3610 one call a function was used. (As it happened, calling the function twice
3611 did not affect the overall behaviour.) The macro has been replaced by a
3612 function that can join an arbitrary number of extra strings onto a growing
3615 (c) The code for expansion conditions now uses a table and a binary chop
3616 instead of a serial search (which was left over from when there were very
3617 few conditions). Also, it now recognizes conditions like "pam" even when
3618 the relevant support is not compiled in: a suitably worded error message is
3619 given if an attempt is made to use such a condition.
3621 11. Added ${time_interval:xxxxx}.
3623 12. A bug was causing one of the ddress fields not to be passed back correctly
3624 from remote delivery subprocesses. The field in question was not being
3625 subsequently used, so this caused to problems in practice.
3627 13. Added new log selectors queue_time and deliver_time.
3629 14. Might have fixed a bug in maildirsizefile handling that threw up
3630 "unexpected character" debug warnings, and recalculated the data
3631 unnecessarily. In any case, I expanded the warning message to give more
3634 15. Added the message "Restricted characters in address" to the statements in
3635 the default ACL that block characters like @ and % in local parts.
3637 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3638 Three changes have been made:
3640 (a) There was a serious bug; a negative response to MAIL caused the whole
3641 recipient domain to be cached as invalid, thereby blocking all messages
3642 to all local parts at the same domain, from all senders. This bug has
3643 been fixed. The domain is no longer cached after a negative response to
3644 MAIL if the sender used is not empty.
3646 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3649 (c) A new callout option, "use_sender" has been added for people who want
3650 the modified behaviour.
3656 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3659 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3660 indeed breaks things for older releases.
3662 3. Added additional logging to the case where there is a problem reading data
3663 from a filter that is running in a subprocess using a pipe, in order to
3664 try to track down a specific problem.
3666 4. Testing facility fudge: when running in the test harness and attempting
3667 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3668 getting "No route to host". Convert this to a timeout.
3670 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3673 6. Some OS don't have socklen_t but use size_t instead. This affects the
3674 fifth argument of getsockopt() amongst other things. This is now
3675 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3676 can be set for individual OS. I have set it for SunOS5, OSF1, and
3677 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3678 some earlier ones do not.
3680 7. Change 4.30/15 was not doing the test caselessly.
3682 8. The standard form for an IPv6 address literal was being rejected by address
3683 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3684 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3685 this, as well as the form without the "IPv6" on the front (but only when
3686 address literals are enabled, of course).
3688 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3690 10. Exim crashed if a message with an empty sender address specified by -f
3691 encountered a router with an errors_to setting. This could be provoked only
3692 by a command such as
3696 where an empty string was supplied; "<>" did not hit this bug.
3698 11. Installed PCRE release 4.5.
3700 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3701 remained set. It is now erased.
3703 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3704 times from message ids (which are base 36 rather than the normal 62).
3706 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3707 were being counted as actual protocol errors, and logged if the log
3708 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3709 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3710 and DATA following a set of rejected RCPTs do not count as protocol errors.
3711 In other words, Exim assumes they were pipelined, though this may not
3712 actually be the case. Of course, in all cases the client gets an
3713 appropriate error code.
3715 15. If a lookup fails in an ACL condition, a message about the failure may
3716 be available; it is used if testing the ACL cannot continue, because most
3717 such messages specify what the cause of the deferral is. However, some
3718 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3719 that caused an old message to be retained and used if a later statement
3720 caused a defer, replacing the real cause of the deferral.
3722 16. If an IP address had so many PTR records that the DNS lookup buffer
3723 was not large enough to hold them, Exim could crash while trying to process
3724 the truncated data. It now detects and logs this case.
3726 17. Further to 4.21/58, another change has been made: if (and only if) the
3727 first line of a message (the first header line) ends with CRLF, a bare LF
3728 in a subsequent header line has a space inserted after it, so as not to
3729 terminate the header.
3731 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3732 unnecessarily, used atoi() instead of strtol(), and didn't check the
3733 termination when getting file sizes from file names by regex.
3735 19. Completely re-implemented the support for maildirsize files, in the light
3736 of a number of problems with the previous contributed implementation
3737 (4.30/29). In particular:
3739 . If the quota is zero, the maildirsize file is maintained, but no quota is
3742 . If the maildir directory does not exist, it is created before any attempt
3743 to write a maildirsize file.
3745 . The quota value in the file is just a cache; if the quota is changed in
3746 the transport, the new value overrides.
3748 . A regular expression is available for excluding directories from the
3751 20. The autoreply transport checks the characters in options that define the
3752 message's headers; it allows continued headers, but it was checking with
3753 isspace() after an embedded newline instead of explicitly looking for a
3756 21. If all the "regular" hosts to which an address was routed had passed their
3757 expiry times, and had not reached their retry times, the address was
3758 bounced, even if fallback hosts were defined. Now Exim should go on to try
3761 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3762 equivalent code in the SMTP transport. Some hosts send humungous responses
3763 to HELO/EHLO, more than 1024 it seems.
3765 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3766 gives compiler warnings in some environments. I've now done it "properly",
3769 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3770 (because of problems with the built-in one) was declared to return uschar *
3771 instead of char *, causing compiler failure.
3773 25. Fixed a file descriptor leak when processing alias/forward files.
3775 26. Fixed a minor format string issue in dbfn.c.
3777 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3779 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3780 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3783 29. When a sender address is verified, it is cached, to save repeating the test
3784 when there is more than one recipient in a message. However, when the
3785 verification involves a callout, it is possible for different callout
3786 options to be set for different recipients. It is too complicated to keep
3787 track of this in the cache, so now Exim always runs a verification when a
3788 callout is required, relying on the callout cache for the optimization.
3789 The overhead is duplication of the address routing, but this should not be
3792 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3793 to be verified with callout=postmaster, and the main callout worked but the
3794 postmaster check failed, the verification correctly failed. However, if a
3795 subsequent RCPT command asked for sender verification *without* the
3796 postmaster check, incorrect caching caused this verification also to fail,
3799 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3800 it was not caching the DNS options (qualify_single, search_parents) that
3801 were used when the lookup failed. A subsequent lookup with different
3802 options therefore always gave the same answer, though there were cases
3803 where it should not have. (Example: a "domains = !$mx_any" option on a
3804 dnslookup router: the "domains" option is always processed without any
3805 widening, but the router might have qualify_single set.) Now Exim uses the
3806 cached value only when the same options are set.
3808 32. Added John Jetmore's "exipick" utility to the distribution.
3810 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3811 than a timeout (e.g. a certificate is required, and is not provided), an
3812 Exim server now closes the connection immediately. Previously it waited for
3813 the client to close - but if the client is SSL, it seems that they each
3814 wait for each other, leading to a delay before one of them times out.
3816 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3817 maintained 0.8.x compatibility because I don't think many are using it, and
3818 it is clearly obsolete.
3820 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3823 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3824 client certificate was expired. A simple patch fixes this, though I don't
3825 understand the full logic of why the verify callback is called multiple
3828 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3829 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3830 which causes problems with some clients (such as the Certicom SSL Plus
3831 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3832 disables the coutermeasure allowing Eudora to connect."
3834 38. Exim was not checking that a write() to a log file succeeded. This could
3835 lead to Bad Things if a log got too big, in particular if it hit a file
3836 size limit. Exim now panics and dies if it cannot write to a log file, just
3837 as it does if it cannot open a log file.
3839 39. Modified OS/Makefile-Linux so that it now contains
3841 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3843 The two -D definitions ensure that Exim is compiled with large file
3844 support, which makes it possible to handle log files that are bigger than
3847 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3848 instance) a domain was checked against a named list that involved a lookup,
3849 causing $domain_data to be set, then another domain was checked against the
3850 same list, then the first domain was re-checked, the value of $domain_data
3851 after the final check could be wrong. In particular, if the second check
3852 failed, it could be set empty. This bug probably also applied to
3855 41. The strip_trailing_dot option was not being applied to the address given
3856 with the -f command-line option.
3858 42. The code for reading a message's header from the spool was incrementing
3859 $received_count, but never initializing it. This meant that the value was
3860 incorrect (doubled) while delivering a message in the same process in which
3861 it was received. In the most common configuration of Exim, this never
3862 happens - a fresh exec is done - but it can happen when
3863 deliver_drop_privilege is set.
3865 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3866 it now includes up to 150 characters of the unexpected data in the log
3869 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3870 and building data strings. The size of both of these buffers was 10 000
3871 bytes - far larger than anybody would *ever* want, thought I. Needless to
3872 say, somebody hit the limit. I have increased the maximum line length to
3873 20 000 and the maximum data length of concatenated lines to 100 000. I have
3874 also fixed two bugs, because there was no checking on these buffers. Tsk,
3875 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3876 buffer is too small.
3878 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3879 lsearch lookups. Now it does.
3881 46. When parsing a route_list item in a manualroute router, a fixed-length
3882 buffer was used for the list of hosts. I made this 1024 bytes long,
3883 thinking that nobody would ever have a list of hosts that long. Wrong.
3884 Somebody had a whole pile of complicated expansion conditions, and the
3885 string was silently truncated, leading to an expansion error. It turns out
3886 that it is easier to change to an unlimited length (owing to other changes
3887 that have happened since this code was originally written) than to build
3888 structure for giving a limitation error. The length of the item that
3889 expands into the list of hosts is now unlimited.
3891 47. The lsearch lookup could not handle data where the length of text line was
3892 more than 4095 characters. Such lines were truncated, leading to shortened
3893 data being returned. It should now handle lines of any length.
3895 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3896 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3899 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3900 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3901 no longer happens in comment lines. A list of replacements is now placed
3902 at the head of all of the source files, except those whose only change is
3903 to replace PERL_COMMAND in the very first #! line.
3905 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3906 messages on the queue, with a bottom-up merge sort, using code contributed
3907 by Michael Haardt. This should make operations like -bp somewhat faster on
3908 large queues. It won't affect queue runners, except when queue_run_in_order
3911 51. Installed eximstats 1.31 in the distribution.
3913 52. Added support for SRV lookups to the dnslookup router.
3915 53. If an ACL referred to $message_body or $message_body_end, the value was not
3916 reset for any messages that followed in the same SMTP session.
3918 54. The store-handling optimization for building very long strings was not
3919 differentiating between the different store pools. I don't think this
3920 actually made any difference in practice, but I've tidied it.
3922 55. While running the routers to verify a sender address, $sender_address
3923 was still set to the sender address. This is wrong, because when routing to
3924 send a bounce to the sender, it would be empty. Therefore, I have changed
3925 it so that, while verifying a sender address, $sender_address is set to <>.
3926 (There is no change to what happens when verifying a recipient address.)
3928 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3929 target A or AAAA records (if not already returned) without resetting the
3930 qualify_single or search_parents options of the DNS resolver. These are
3931 inappropriate in this case because the targets of MX and SRV records must
3932 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3933 target that, when qualified, matched something in the local domain. These
3934 two options are now turned off when doing these lookups.
3936 57. It seems that at least some releases of Reiserfs (which does not have the
3937 concept of a fixed number of inodes) returns zero and not -1 for the
3938 number of available inodes. This interacted badly with check_spool_inodes,
3939 which assumed that -1 was the "no such thing" setting. What I have done is
3940 to check that the total number of inodes is greater than zero before doing
3941 the test of how many are available.
3943 58. When a "warn" ACL statement has a log_message modifier, the message is
3944 remembered, and not repeated. This is to avoid a lot of repetition when a
3945 message has many recipients that cause the same warning to be written.
3946 Howewer, Exim was preserving the list of already written lines for an
3947 entire SMTP session, which doesn't seem right. The memory is now reset if a
3948 new message is started.
3950 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3951 debugging output unless log_rewrite was also set.
3953 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3954 of (int)(handle) when we know that handle contains (void *)(-1).
3956 61. The Exim daemon panic-logs an error return when it closes the incoming
3957 connection. However "connection reset by peer" seems to be common, and
3958 isn't really an error worthy of noting specially, so that particular error
3961 62. When Exim is trying to find all the local interfaces, it used to panic and
3962 die if the ioctl to get the interface flags failed. However, it seems that
3963 on at least one OS (Solaris 9) it is possible to have an interface that is
3964 included in the list of interfaces, but for which you get a failure error
3965 for this call. This happens when the interface is not "plumbed" into a
3966 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3967 failure of the "get flags" call assumes that the interface is down.
3969 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3970 makes life easier for people who are doing arithmetic on fields extracted
3971 from dates, where you often get leading zeros that should not be
3972 interpreted as octal.
3974 64. Added qualify_domain to the redirect router, to override the global
3977 65. If a pathologically long header line contained very many addresses (the
3978 report of this problem mentioned 10 000) and each of them was rewritten,
3979 Exim could use up a very large amount of memory. (It kept on making new
3980 copies of the header line as it rewrote, and never released the old ones.)
3981 At the expense of a bit more processing, the header rewriting function has
3982 been changed so that it no longer eats memory in this way.
3984 66. The generation of the Received: header has been moved from the time that a
3985 message starts to be received, to the time that it finishes. The timestamp
3986 in the Received: header should now be very close to that of the <= log
3987 line. There are two side-effects of this change:
3989 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3990 logged header lines no longer include the local Received: line, because
3991 it has not yet been created. The same applies to a copy of the message
3992 that is returned to a non-SMTP sender when a message is rejected.
3994 (b) When a filter file is tested using -bf, no additional Received: header
3995 is added to the test message. After some thought, I decided that this
3998 This change does not affect the value of $received_for. It is still set
3999 after address rewriting, but before local_scan() is called.
4001 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4003 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4004 gave an unhelpful panic error message, and a defer error. I have managed to
4005 change this behaviour so that it now rejects any supplied certificate,
4006 which seems right, as the list of acceptable certificates is empty.
4008 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4009 gave an unhelpful defer error. I have not managed to make this reject any
4010 supplied certificates, but the error message it gives is "no certificate
4011 supplied", which is not helpful.
4013 70. exigrep's output now also includes lines that are not associated with any
4014 message, but which match the given pattern. Implemented by a patch from
4015 Martin Sluka, which also tidied up the Perl a bit.
4017 71. Recipient callout verification, like sender verification, was using <> in
4018 the MAIL FROM command. This isn't really the right thing, since the actual
4019 sender may affect whether the remote host accepts the recipient or not. I
4020 have changed it to use the actual sender in the callout; this means that
4021 the cache record is now keyed on a recipient/sender pair, not just the
4022 recipient address. There doesn't seem to be a real danger of callout loops,
4023 since a callout by the remote host to check the sender would use <>.
4024 [SEE ABOVE: changed after hitting problems.]
4026 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4027 temporary errors. However, in the case of such a code being given after
4028 the end of a data transmission (i.e. after ".") Exim was failing to write
4029 a retry record for the message. (Yes, there was some broken host that was
4030 actually sending 8xx at this point.)
4032 73. An unknown lookup type in a host list could cause Exim to panic-die when
4033 the list was checked. (An example that provoked this was putting <; in the
4034 middle of a list instead of at the start.) If this happened during a DATA
4035 ACL check, a -D file could be left lying around. This kind of configuration
4036 error no longer causes Exim to die; instead it causes a defer errror. The
4037 incident is still logged to the main and panic logs.
4039 74. Buglet left over from Exim 3 conversion. The message "too many messages
4040 in one connection" was written to the rejectlog but not the mainlog, except
4041 when address rewriting (yes!) was being logged.
4043 75. Added write_rejectlog option.
4045 76. When a system filter was run not as root (that is, when system_filter_user
4046 was set), the values of the $n variables were not being returned to the
4047 main process; thus, they were not subsequently available in the $sn
4050 77. Added +return_path_on_delivery log selector.
4052 78. A connection timeout was being treated differently from recipients deferred
4053 when testing hosts_max_try with a message that was older than the host's
4054 retry timeout. (The host should not be counted, thus allowing all hosts to
4055 be tried at least once before bouncing.) This may have been the cause of an
4056 occasionally reported bug whereby a message would remain on the queue
4057 longer than the retry timeout, but would be bounced if a delivery was
4058 forced. I say "may" because I never totally pinned down the problem;
4059 setting up timeout/retry tests is difficult. See also the next item.
4061 79. The ultimate address timeout was not being applied to errors that involved
4062 a combination of host plus message (for example, a timeout on a MAIL
4063 command). When an address resolved to a number of possible hosts, and they
4064 were not all tried for each delivery (e.g. because of hosts_max_try), a
4065 message could remain on the queue longer than the retry timeout.
4067 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4070 81. Fixed an obscure SMTP outgoing bug which required at least the following
4071 conditions: (a) there was another message waiting for the same server;
4072 (b) the server returned 5xx to all RCPT commands in the first message so
4073 that the message was not completed; (c) the server dropped the connection
4074 or gave a negative response to the RSET that Exim sends to abort the
4075 transaction. The observed case was a dropped connection after DATA that had
4076 been sent in pipelining mode. That is, the server had advertised PIPELINING
4077 but was not implementing it correctly. The effect of the bug was incorrect
4078 behaviour, such as trying another host, and this could lead to a crash.
4084 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4085 and daemon.c were passed as pointers to ints; they should have been
4086 pointers to socklen_t variables (which are typically unsigned ints).
4088 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4091 3. Fixed a really odd bug that affected only the testing scheme; patching a
4092 certain fixed string in the binary changed the value of another string that
4093 happened to be identical to the end of the original first string.
4095 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4096 name", it returns that address as the IP address. On some operating
4097 systems (e.g. Solaris), it also passes back the IP address string as the
4098 "host name". However, on others (e.g. Linux), it passes back an empty
4099 string. Exim wasn't checking for this, and was changing the host name to an
4100 empty string, assuming it had been canonicized.
4102 5. Although rare, it is permitted to have more than one PTR record for a given
4103 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4104 all the names associated with an address, because they do in Solaris.
4105 However, it seems that they do not in Linux for data that comes from the
4106 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4107 I found this out when I moved to a new Linux workstation and tried to run
4108 the Exim test suite.
4110 To get round this problem I have changed the code so that it now does its
4111 own call to the DNS to look up PTR records when searching for a host name.
4112 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4113 addresses that are only in /etc/hosts are still found.
4115 This behaviour is, however, controlled by an option called host_lookup_
4116 order, which defaults to "bydns:byaddr". If people want to use the other
4117 order, or indeed, just use one or the other means of lookup, they can
4118 specify it in this variable.
4120 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4121 some operating systems, this comes back from gethostbyaddr() as an empty
4122 string, and this is what Exim used to test for. However, it seems that in
4123 other systems, "." is yielded. Exim now tests for this case too.
4125 7. The values of check_spool_space and check_log_space are now held internally
4126 as a number of kilobytes instead of an absolute number of bytes. If a
4127 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4128 kilobyte. This means that much larger values can be stored.
4130 8. Exim monitor: an attempt to get the action menu when not actually pointing
4131 at a message produces an empty menu entitled "No message selected". This
4132 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4133 no entries in it ("Shell widget menu has zero width and/or height"). So I
4134 have added a single, blank menu entry in this case.
4136 9. Added ${quote_local_part.
4138 10. MIME decoding is now applied to the contents of Subject: header lines when
4141 11. Now that a reference to $sender_host_address automatically causes a reverse
4142 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4143 host lookup before query-style lookups in lists that might use this
4144 variable. This has therefore been abolished, and the "net-" prefix is no
4145 longer necessary for query-style lookups.
4147 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4148 have been a typo for LFLAGS, so it has been changed.
4150 13. The install script calls Exim with "-C /dev/null" in order to find the
4151 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4152 to be output. Howeve, since Exim outputs its version number before the
4153 error, it didn't break the script. It just looked ugly. I fixed this by
4154 always allowing "-C /dev/null" if the caller is root.
4156 14. Ignore overlarge ACL variable number when reading spool file - insurance
4157 against a later release with more variables having written the file.
4159 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4160 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4161 this, as well as the form without the "IPv6" on the front.
4163 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4164 OS/Makefile-Darwin file.
4166 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4167 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4169 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4170 variable. However, some OS use ints for inodes. I've added cast to long int
4171 to get rid of the compiler warning.
4173 19. I had forgotten to lock out "/../" in configuration file names when
4174 ALT_CONFIG_PREFIX was set.
4176 20. Routers used for verification do not need to specify transports. However,
4177 if such a router generated a host list, and callout was configured, Exim
4178 crashed, because it could not find a port number from the (non-existent)
4179 transport. It now assumes port 25 in this circumstance.
4181 21. Added the -t option to exigrep.
4183 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4184 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4185 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4186 them is set, LOOKUP_LSEARCH is forced.
4188 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4190 24. Added sender and host information to the "rejected by local_scan()" log
4191 line; previously there was no indication of these.
4193 25. Added .include_if_exists.
4195 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4196 Linux. It turns out that not all file systems support this. Apparently some
4197 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4198 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4199 not supported on the file descriptor, is now ignored when Exim is trying to
4200 sync a directory. This applies only to Linux.
4202 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4204 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4205 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4208 29. Added support for maildirsize files from supplied patch (modified a bit).
4210 30. The use of :fail: followed by an empty string could lead Exim to respond to
4211 sender verification failures with (e.g.):
4213 550 Verification failed for <xxx>
4214 550 Sender verify failed
4216 where the first response line was missing the '-' that indicates it is not
4217 the final line of the response.
4219 31. The loop for finding the name of the user that called Exim had a hardwired
4220 limit of 10; it now uses the value of finduser_retries, which is used for
4221 all other user lookups.
4223 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4226 33. Exim was neglecting to zero errno before one call of strtol() when
4227 expanding a string and expecting an integer value. On some systems this
4228 resulted in spurious "integer overflow" errors. Also, it was casting the
4229 result into an int without checking.
4231 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4232 did not work. The code looks as if it has *never* worked, though it appears
4233 to have been documented since at least releast 1.62. I have made it work.
4235 35. The "timeout_DNS" error in retry rules, also documented since at least
4236 1.62, also never worked. As it isn't clear exactly what this means, and
4237 clearly it isn't a major issue, I have abolished the feature by treating it
4238 as "timeout", and writing a warning to the main and panic logs.
4240 36. The display of retry rules for -brt wasn't always showing the error code
4243 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4244 timeout_connect_A, timeout_connect_MX.
4246 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4247 to the empty sender.
4249 39. The daemon was not analysing the content of -oX till after it had closed
4250 stderr and disconnected from the controlling terminal. This meant that any
4251 syntax errors were only noted on the panic log, and the return code from
4252 the command was 0. By re-arranging the code a little, I've made the
4253 decoding happen first, so such errors now appear on stderr, and the return
4254 code is 1. However, the actual setting up of the sockets still happens in
4255 the disconnected process, so errors there are still only recorded on the
4258 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4259 connections (as happens on some IP stacks) was logged at start up time as
4260 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4261 it from "IPv6 and IPv4", which means that two separate sockets are being
4264 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4265 says whether AF_INET or AF_INET6 was passed as an argument.
4267 42. Exiwhat output was messed up when time zones were included in log
4270 43. Exiwhat now gives more information about the daemon's listening ports,
4271 and whether -tls-on-connect was used.
4273 44. The "port" option of the smtp transport is now expanded.
4275 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4276 silently ignored. Now an error message is written to the main and panic
4279 46. There's a new ACL modifier called "logwrite" which writes to a log file
4280 as soon as it is encountered.
4282 47. Added $local_user_uid and $local_user_gid at routing time.
4284 48. Exim crashed when trying to verify a sender address that was being
4287 49. Exim was recognizing only a space character after ".include". It now also
4288 recognizes a tab character.
4290 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4291 extracting the relevant information from the specification. The man page no
4292 longer contains scrambled data for the -d option, and I've added a section
4293 at the front about calling Exim under different names.
4295 51. Added "extra_headers" argument to the "mail" command in filter files.
4297 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4300 53. Installed eximstats 1.29.
4302 54. Added transport_filter_timeout as a generic transport option.
4304 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4305 Cc: header lines. This was required by RFC 822, but it not required by RFC
4308 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4309 incoming messages that did not have them. Now it does so only if the
4310 message originates locally, that is, if there is no associated remote host
4311 address. When Resent- header lines are present, this applies to the Resent-
4312 lines rather than the non-Resent- lines.
4314 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4315 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4317 58. Messages for configuration errors now include the name of the main
4318 configuration file - useful now that there may be more than one file in a
4319 list (.included file names were always shown).
4321 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4322 for those rare installations that do not start the daemon as root or run it
4323 setuid root. I've cut out the call to initgroups() if the daemon is not
4326 60. The Exim user and group can now be bound into the binary as text strings
4327 that are looked up at the start of Exim's processing.
4329 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4331 62. Added $mailstore_basename variable.
4333 63. Installed patch to sieve.c from Michael Haardt.
4335 64. When Exim failed to open the panic log after failing to open the main log,
4336 the original message it was trying to log was written to stderr and debug
4337 output, but if they were not available (the usual case in production), it
4338 was lost. Now it is written to syslog before the two lines that record the
4339 failures to open the logs.
4341 65. Users' Exim filters run in subprocesses under the user's uid. It is
4342 possible for a "deliver" command or an alias in a "personal" command to
4343 provoke an address rewrite. If logging of address rewriting is configured,
4344 this fails because the process is not running as root or exim. There may be
4345 a better way of dealing with this, but for the moment (because 4.30 needs
4346 to be released), I have disabled address rewrite logging when running a
4347 filter in a non-root, non-exim process.
4353 1. The buildconfig auxiliary program wasn't quoting the value set for
4354 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4355 not defined. This bug was masked in 4.22 by the effect that was fixed in
4358 2. Some messages that were rejected after a message id was allocated were
4359 shown as "incomplete" by exigrep. It no longer does this for messages that
4360 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4362 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4363 have allow_domain_literals set, the ID did not get logged in the <= line.
4364 Domain literals are now always recognized in Message-ID: header lines.
4366 4. The first argument for a ${extract expansion item is the key name or field
4367 number. Leading and trailing spaces in this item were not being ignored,
4368 causing some misleading effects.
4370 5. When deliver_drop_privilege was set, single queue runner processes started
4371 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4372 same command in the process it spins off) were not dropping privilege.
4374 6. When the daemon running as "exim" started a queue runner, it always
4375 re-executed Exim in the spun-off process. This is a waste of effort when
4376 deliver_drop_privilege is set. The new process now just calls the
4377 queue-runner function directly.
4383 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4386 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4387 ignored. Though the use of -oP was forcing the writing of a pid file, it
4388 was always written to the default place.
4390 3. If the message "no IP address found for host xxxx" is generated during
4391 incoming verification, it is now followed by identification of the incoming
4392 connection (so you can more easily find what provoked it).
4394 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4396 5. Added some features to "harden" Exim a bit more against certain attacks:
4398 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4399 be put in Local/Makefile. This is like the never_users runtime option,
4400 but it cannot be overridden. The default setting is "root".
4402 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4403 prefix string with which any file named in a -C command line option
4406 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4407 is retained for -C and -D only if the caller of Exim is root. Without
4408 it, the exim user may also use -C and -D and retain privilege.
4410 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4411 command line option is disabled.
4413 6. Macro names set by the -D option must start with an upper case letter, just
4414 like macro names defined in the configuration file.
4416 7. Added "dereference=" facility to LDAP.
4418 8. Two instances of the typo "uknown" in the source files are fixed.
4420 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4421 the Configure-Makefile script screwed up while processing it.
4423 10. Incorporated PCRE 4.4.
4425 11. The SMTP synchronization check was not operating right at the start of an
4426 SMTP session. For example, it could not catch a HELO sent before the client
4427 waited for the greeting. There is now a check for outstanding input at the
4428 point when the greeting is written. Because of the duplex, asynchronous
4429 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4430 way, but not yet received, when the check is performed.
4432 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4433 on TCP/IP sockets, because this apparently causes some broken clients to
4436 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4437 unchanged) from the Cygwin maintainer.
4439 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4440 was supported instead of testing for mbx. Effectively a typo.
4442 15. The spa authenticator server code was not checking that the input it
4443 received was valid base64.
4445 16. The debug output line for the "set" modifier in ACLs was not showing the
4446 name of the variable that was being set.
4448 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4450 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4451 lookup on its own. Something else was needed to trigger the lookup. For
4452 example, a match in host_lookup or the need for a host name in a host list.
4453 Now, if $sender_host_name is referenced and the host name has not yet been
4454 looked up, a lookup is performed. If the lookup fails, the variable remains
4455 empty, and $host_lookup_failed is set to "1".
4457 19. Added "eqi" as a case-independent comparison operator.
4459 20. The saslauthd authentication condition could segfault if neither service
4460 nor realm was specified.
4462 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4463 error message that was logged was misleading, and incoming SMTP
4464 connections were dropped. The message is now more accurate, and temporary
4465 errors are given to SMTP connections.
4467 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4468 (except RSET) with a 421 error, until QUIT is received. However, it was
4469 failing to send a response to QUIT.
4471 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4472 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4473 ACL is now run after the helo_try_verify_hosts code.
4475 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4476 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4477 case-independent, so other case variants are also recognized). Apparently
4478 some systems use these upper case variants.
4480 25. If more than two messages were waiting for the same host, and a transport
4481 filter was specified for the transport, Exim sent two messages over the
4482 same TCP/IP connection, and then failed with "socket operation on non-
4483 socket" when it tried to send the third.
4485 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4487 27. The extern definition of crypt16() in expand.c was not being excluded when
4488 the OS had its own crypt16() function.
4490 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4491 as a preferred synonym for return_size_limit, both as an option and as an
4494 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4496 30. Changed the default configuration ACL to relax the local part checking rule
4497 for addresses that are not in any local domains. For these addresses,
4498 slashes and pipe symbols are allowed within local parts, but the sequence
4499 /../ is explicitly forbidden.
4501 31. SPA server authentication was not clearing the challenge buffer before
4504 32. log_message in a "warn" ACL statement was writing to the reject log as
4505 well as to the main log, which contradicts the documentation and doesn't
4506 seem right (because no rejection is happening). So I have stopped it.
4508 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4509 However, I am unable to do any testing of this.
4511 34. Fixed an infelicity in the appendfile transport. When checking directories
4512 for a mailbox, to see if any needed to be created, it was accidentally
4513 using path names with one or more superfluous leading slashes; tracing
4514 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4516 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4517 contents are added to the log line that is written for every discarded
4518 recipient. (Previously a log_message setting was ignored.)
4520 36. The ${quote: operator now quotes the string if it is empty.
4522 37. The install script runs exim in order to find its version number. If for
4523 some reason other than non-existence or emptiness, which it checks, it
4524 could not run './exim', it was installing it with an empty version number,
4525 i.e. as "exim-". This error state is now caught, and the installation is
4528 38. An argument was missing from the function that creates an error message
4529 when Exim fails to connect to the socket for saslauthd authentication.
4530 This could cause Exim to crash, or give a corrupted message.
4532 39. Added isip, isip4, and isip6 to ${if conditions.
4534 40. The ACL variables $acl_xx are now saved with the message, and can be
4535 accessed later in routers, transports, and filters.
4537 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4538 strings in the file are not string-expanded.
4540 42. If a MAIL command specified a SIZE value that was too large to fit into an
4541 int variable, the check against message_size_limit failed. Such values are
4542 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4543 day this will have to be increased, but I don't think I want to be around
4544 when emails are that large.
4551 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4552 iconv() is not standard in FreeBSD.
4554 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4555 IPv6 enabled. The observed symptom was a segmentation fault on return from
4556 the function os_common_find_running_interfaces() in src/os.c.
4558 3. In the check_special_case() function in daemon.c I had used "errno" as an
4559 argument name, which causes warnings on some systems. This was basically a
4560 typo, since it was named "eno" in the comments!
4562 4. The code that waits for the clock to tick (at a resolution of some fraction
4563 of a second) so as to ensure message-id uniqueness was always waiting for
4564 at least one whole tick, when it could have waited for less. [This is
4565 almost certainly not relevant at current processor speeds, where it is
4566 unlikely to ever wait at all. But we try to future-proof.]
4568 5. The function that sleeps for a time interval that includes fractions of a
4569 second contained a race. It did not block SIGALRM between setting the
4570 timer, and suspending (a couple of lines later). If the interval was short
4571 and the sigsuspend() was delayed until after it had expired, the suspension
4572 never ended. On busy systems this could lead to processes getting stuck for
4575 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4576 process, before it forks any delivery processes. The open lookup caching
4577 mechanism meant that the open file or database connection was passed into
4578 the delivery process. The problem was that delivery processes always tidy
4579 up cached lookup data. This could cause a problem for the next delivery
4580 process started by the queue runner, because the external queue runner
4581 process does not know about the closure. So the next delivery process
4582 still has data in the lookup cache. In the case of a file lookup, there was
4583 no problem because closing a file descriptor in a subprocess doesn't affect
4584 the parent. However, if the lookup was caching a connection to a database,
4585 the connection was closed, and the second delivery process was likely to
4586 see errors such as "PGSQL: query failed: server closed the connection
4587 unexpectedly". The problem has been fixed by closing all cached lookups
4588 in a queue runner before running a delivery process.
4590 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4591 seem to have the "const" qualifier which it has on other OS. I've
4594 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4595 *of the same type* (client or server) with the same public name that an
4596 error should be diagnosed.
4598 9. When Exim looked up a host name for an IP address, but failed to find the
4599 original IP address when looking up the host name (a safety check), it
4600 output the message "<ip address> does not match any IP for NULL", which was
4601 confusing, to say the least. The bug was that the host name should have
4602 appeared instead of "NULL".
4604 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4605 user that is built into the binary, and the -C or -D options is used, root
4606 privilege is dropped before the configuration file is read. In addition,
4607 logging is switched to stderr instead of the normal log files. If the
4608 configuration then re-defines the Exim user, the unprivileged environment
4609 is probably not what is expected, so Exim logs a panic warning message (but
4612 However, if deliver_drop_privilege is set, the unprivileged state may well
4613 be exactly what is intended, so the warning has been cut out in that case,
4614 and Exim is allowed to try to write to its normal log files.
4620 1. smtp_return_error_details was not giving details for temporary sender
4621 or receiver verification errors.
4623 2. Diagnose a configuration error if two authenticators have the same public
4626 3. Exim used not to create the message log file for a message until the first
4627 delivery attempt. This could be confusing when incoming messages were held
4628 for policy or load reasons. The message log file is now created at the time
4629 the message is received, and an initial "Received" line is written to it.
4631 4. The automatically generated man page for command line options had a minor
4632 bug that caused no ill effects; however, a more serious problem was that
4633 the procedure for building the man page automatically didn't always
4634 operate. Consequently, release 4.20 contains an out-of-date version. This
4635 shouldn't happen again.
4637 5. When building Exim with embedded Perl support, the script that builds the
4638 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4639 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4641 6. The freeze_tell option was not being used for messages that were frozen on
4642 arrival, either by an ACL or by local_scan().
4644 7. Added the smtp_incomplete_transaction log selector.
4646 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4647 was accepting AUTH without a new EHLO.
4649 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4650 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4653 10. Logging of TCP/IP connections (when configured) now happens in the main
4654 daemon process instead of the child process, so that the TCP/IP connection
4655 count is more accurate (but it can never be perfect).
4657 11. The use of "drop" in a nested ACL was not being handled correctly in the
4658 outer ACL. Now, if condition failure induced by the nested "drop" causes
4659 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4660 or "require"), the connection is dropped.
4662 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4663 that yield "discard" can now be used with an "accept" or a "discard" verb,
4664 but an error is generated for any others (because I can't see a useful way
4665 to define what should happen).
4667 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4668 are now processed in the same way as lines in the Exim configuration file.
4669 In particular, continuation lines are supported.
4671 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4673 15. Added -ti meaning -t -i.
4675 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4676 domains, and warn by logging if others are found.
4678 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4679 returned from GIFCONF when Exim is trying to find the list of interfaces on
4680 a host. The code in os.c has been modified to copy each ifreq to an aligned
4681 structure in all cases.
4683 Also, in some cases, the returned ifreq's were being copied to a 'struct
4684 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4685 means the last couple of bytes of an IPv6 address could be chopped if the
4686 ifreq contained only a normal sockaddr (14 bytes storage).
4688 18. Named domain lists were not supported in the hosts_treat_as_local option.
4689 An entry such as +xxxx was not recognized, and was treated as a literal
4692 19. Ensure that header lines added by a DATA ACL are included in the reject log
4693 if the ACL subsequently rejects the message.
4695 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4696 MD5 (which is deprecated).
4698 21. When testing a filter file using -bf, Exim was writing a message when it
4699 took the sender from a "From " line in the message, but it was not doing so
4700 when it took $return_path from a Return-Path: header line. It now does.
4702 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4703 with a valid header line field name (a series of printing characters
4704 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4706 23. Changed "disc" in the source to "disk" to conform to the documentation and
4707 the book and for uniformity.
4709 24. Ignore Sendmail's -Ooption=value command line item.
4711 25. When execve() failed while trying to run a command in a pipe transport,
4712 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4713 could be confused with a return value of 69 from the command itself. This
4714 has been changed to 127, the value the shell returns if it is asked to run
4715 a non-existent command. The wording for the related log line suggests a
4716 non-existent command as the problem.
4718 26. If received_header_text expands to an empty string, do not add a Received:
4719 header line to the message. (Well, it adds a token one on the spool, but
4720 marks it "old" so that it doesn't get used or transmitted.)
4722 27. Installed eximstats 1.28 (addition of -nt option).
4724 28. There was no check for failure on the call to getsockname() in the daemon
4725 code. This can fail if there is a shortage of resources on the system, with
4726 ENOMEM, for example. A temporary error is now given on failure.
4728 29. Contrary to the C standard, it seems that in some environments, the
4729 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4730 program. Exim now does this explicitly; it affects the formatting of
4731 timestamps using strftime().
4733 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4734 complaints. I've now initialized all the variables, to avoid this.
4736 32. Header lines added by a system filter were not being "seen" during
4737 transport-time rewrites.
4739 33. The info_callback() function passed to OpenSSL is set up with type void
4740 (*)(SSL *, int, int), as described somewhere. However, when calling the
4741 function (actually a macro) that sets it up, the type void(*)() is
4742 expected. I've put in a cast to prevent warnings from picky compilers.
4744 34. If a DNS black list lookup found a CNAME record, but there were no A
4745 records associated with the domain it pointed at, Exim crashed.
4747 35. If a DNS black list lookup returned more than one A record, Exim ignored
4748 all but the first. It now scans all returned addresses if a particular IP
4749 value is being sought. In this situation, the contents of the
4750 $dnslist_value variable are a list of all the addresses, separated by a
4753 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4754 to accept a host name and all its aliases if the forward lookup for any of
4755 them yielded the IP address of the incoming connection. Now it accepts only
4756 those names whose forward lookup yields the correct IP address. Any other
4757 names are discarded. This closes a loophole whereby a rogue DNS
4758 administrator could create reverse DNS records to break through a
4759 wildcarded host restriction in an ACL.
4761 37. If a user filter or a system filter that ran in a subprocess used any of
4762 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4763 the wrong values were passed to the pipe command ($thisaddress had the
4764 value of $0, $0 had the value of $1, etc). This bug was introduced by
4765 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4767 38. Improved the line breaking for long SMTP error messages from ACLs.
4768 Previously, if there was no break point between 40 and 75 characters, Exim
4769 left the rest of the message alone. Two changes have been made: (a) I've
4770 reduced the minimum length to 35 characters; (b) if it can't find a break
4771 point between 35 and 75 characters, it looks ahead and uses the first one
4772 that it finds. This may give the occasional overlong line, but at least the
4773 remaining text gets split now.
4775 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4776 file descriptors might be low, and that setting 1000 would always raise it.
4777 It turns out that in some environments, the limit is already over 1000 and
4778 that lowering it causes trouble. So now Exim takes care not to decrease it.
4780 40. When delivering a message, the value of $return_path is set to $sender_
4781 address at the start of routing (routers may change the value). By an
4782 oversight, this default was not being set up when an address was tested by
4783 -bt or -bv, which affected the outcome if any router or filter referred to
4786 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4787 log only when "message" or "log_message" are set. However, if one of the
4788 conditions was an address verification, or a call to a nested ACL, the
4789 messages generated by the underlying test were being passed through. This
4790 no longer happens. The underlying message is available in $acl_verify_
4791 message for both "message" and "log_message" expansions, so it can be
4792 passed through if needed.
4794 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4795 new expansion $bh_ to give the encoded byte string without charset
4796 translation. Translation happens only if iconv() is available; HAVE_ICONV
4797 indicates this at build time. HEADERS_CHARSET gives the charset to
4798 translate to; headers_charset can change it in the configuration, and
4799 "headers charset" can change it in an individual filter file.
4801 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4802 that creates RFC 2047 encoded "words" labels them as that charset instead
4803 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4804 expansion operator; (ii) when Exim creates a From: line for a local
4805 message; (iii) when a header line is rewritten to include a "phrase" part.
4807 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4808 buggy, causing it to skip the first lines of messages whose message ID
4809 ended in 'D'. This would not have bitten before Exim release 4.14, because
4810 message IDs were unlikely to end in 'D' before then. The effect was to have
4811 incorrect size information for certain domains.
4813 45. #include "config.h" was missing at the start of the crypt16.c module. This
4814 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4817 46. If there was a timeout during a "random" callout check, Exim treated it as
4818 a failure of the random address, and carried on sending RSET and the real
4819 address. If the delay was just some slowness somewhere, the response to the
4820 original RCPT would be taken as a response to RSET and so on, causing
4821 mayhem of various kinds.
4823 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4824 when I implemented it. It didn't allow for the fact that some option values
4825 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4826 the right test for positive values.
4828 48. Domain names in DNS records are case-independent. Exim always looks them up
4829 in lower case. Some resolvers return domain names in exactly the case they
4830 appear in the zone file, that is, they may contain uppercase letters. Not
4831 all resolvers do this - some return always lower case. Exim was treating a
4832 change of case by a resolver as a change of domain, similar to a widening
4833 of a domain abbreviation. This triggered its re-routing code and so it was
4834 trying to route what was effectively the same domain again. This normally
4835 caused routing to fail (because the router wouldn't handle the domain
4836 twice). Now Exim checks for this case specially, and just changes the
4837 casing of the domain that it ultimately uses when it transmits the message
4840 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4843 50. If a filter generated a file delivery with a non-absolute name (possible if
4844 no home directory exists for the router), the forbid_file option was not
4847 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4848 the existing equality matching.
4850 52. Exim was using ints instead of ino_t variables in some places where it was
4851 dealing with inode numbers.
4853 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4854 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4855 called TMPDIR, and if it finds it is different, it changes its value.
4857 54. The smtp_printf() function is now made available to local_scan() so
4858 additional output lines can be written before returning. There is also an
4859 smtp_fflush() function to enable the detection of a dropped connection.
4860 The variables smtp_input and smtp_batched_input are exported to
4863 55. Changed the default runtime configuration: the message "Unknown user"
4864 has been removed from the ACL, and instead placed on the localuser router,
4865 using the cannot_route_message feature. This means that any verification
4866 failures that generate their own messages won't get overridden. Similarly,
4867 the "Unrouteable address" message that was in the ACL for unverifiable
4868 relay addresses has also been removed.
4870 56. Added hosts_avoid_esmtp to the smtp transport.
4872 57. The exicyclog script was not checking for the esoteric option
4873 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4874 will work only if exicyclog is run under the appropriate euid.
4876 58. Following a discussion on the list, the rules by which Exim recognises line
4877 endings on incoming messages have been changed. The -dropcr and drop_cr
4878 options are now no-ops, retained only for backwards compatibility. The
4879 following line terminators are recognized: LF CRLF CR. However, special
4880 processing applies to CR:
4882 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4883 nor a local message in the state where . is a terminator.
4885 (ii) If a bare CR is encountered in a header line, an extra space is added
4886 after the line terminator so as not to end the header. The reasoning
4887 behind this is that bare CRs in header lines are most likely either
4888 to be mistakes, or people trying to play silly games.
4890 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4891 was being incorrectly given as 18 bytes larger than it should have been.
4892 This is a VOB (very old bug).
4894 60. This may never have affected anything current, but just in case it has:
4895 When the local host is found other than at the start of a list of hosts,
4896 the local host, those with the same MX, and any that follow, are discarded.
4897 When the list in question was part of a longer list of hosts, the following
4898 hosts (not currently being processed) were also being discarded. This no
4899 longer happens. I'm not sure if this situation could ever has previously
4902 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4905 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4906 header. This is something that is recommended in a new Internet Draft, and
4907 is something that is documented as being done by Sendmail. There are two
4908 possible values. For messages generated by the autoreply transport, Exim
4911 Auto-Submitted: auto-replied
4913 whereas for all other generated messages (e.g. bounces) it adds
4915 Auto-Submitted: auto-generated
4917 63. The "personal" condition in filters now includes a test for the
4918 Auto-Submitted: header. If it contains the string "auto-" the message it
4919 not considered personal.
4921 64. Added rcpt_include_affixes as a generic transport option.
4923 65. Added queue_only_override (default true).
4925 66. Added the syslog_duplication option.
4927 67. If what should have been the first header line of a message consisted of
4928 a space followed by a colon, Exim was mis-interpreting it as a header line.
4929 It isn't of course - it is syntactically invalid and should therefore be
4930 treated as the start of the message body. The misbehaviour could have
4931 caused a number of strange effects, including loss of data in subsequent
4932 header lines, and spool format errors.
4934 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4935 client host had authenticated. This control can now be exercised by an ACL
4936 for more flexibility.
4938 69. By default, callouts do not happen when testing with -bh. There is now a
4939 variant, -bhc, which does actually run the callout code, including
4940 consulting and updating the callout cache.
4942 70. Added support for saslauthd authentication, courtesy of Alexander
4945 71. If statvfs() failed on the spool or log directories while checking their
4946 size for availability, Exim confusingly gave the error "space shortage".
4947 Furthermore, in debugging mode it crashed with a floating point exception.
4948 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4949 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4950 serious problem, Exim now writes to the main and panic logs when this
4951 happens, with details of the failure. It then refuses to accept the
4952 incoming message, giving the message "spool directory problem" or "log
4953 directory problem" with a 421 code for SMTP messages.
4955 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4956 0, 1, and 2 exist, because some OS complain for execs without them (see
4957 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4958 descriptors. However, the code omitted to check that the open succeeded,
4959 causing mysterious errors if for some reason the permissions on /dev/null
4960 got screwed. Now Exim writes a message to the main and panic logs, and
4961 bombs out if it can't open /dev/null.
4963 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4964 interact so that it is all more flexible. It is supposed to remain
4965 backwards compatible. Also added extra_local_interfaces.
4967 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4968 to bomb out with an assertion failure - to the client this appears as a
4969 connection drop. This problem occurs in the part of the code that was taken
4970 from the Samba project. Fortunately, the assertion is in a very simple
4971 function, so I have fixed this by reproducing the function inline in the
4972 one place where it is called, and arranging for authentication to fail
4973 instead of killing the process with assert().
4975 75. The SPA client code was not working when the server requested OEM rather
4976 than Unicode encoding.
4978 76. Added code to make require_files with a specific uid setting more usable in
4979 the case where statting the file as root fails - usually a non-root-mounted
4980 NFS file system. When this happens and the failure is EACCES, Exim now
4981 forks a subprocess and does the per-uid checking as the relevant uid.
4983 77. Added process_log_path.
4985 78. If log_file_path was not explicitly set, a setting of check_log_space or
4986 check_log_inodes was ignored.
4988 79. If a space check for the spool or log partitions fails, the incident is now
4989 logged. Of course, in the latter case the data may get lost...
4991 80. Added the %p formatting code to string_format() so that it can be used to
4992 print addresses in debug_print(). Adjusted all the address printing in the
4993 debugging in store.c to use %p rather than %d.
4995 81. There was a concern that a line of code in smtp_in.c could overflow a
4996 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4997 initially expressed, the concern was not well-founded, because trailing
4998 spaces are removed early. However, if the trailing spaces were followed by
4999 a NULL, they did not get removed, so the overflow was possible. Two fixes
5002 (a) I re-wrote the offending code in a cleaner fashion.
5003 (b) If an incoming SMTP command contains a NULL character, it is rejected
5006 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5007 runs initgroups(), so that if the Exim user is in any additional groups,
5008 they will be used during message reception.
5014 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob