1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.585 2009/10/30 15:14:04 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
149 TK/01 Add preliminary DKIM support. Currently requires a forked version of
150 ALT-N's libdkim that I have put here:
151 http://duncanthrax.net/exim-experimental/
153 Note to Michael Haardt: I had to rename some vars in sieve.c. They
154 were called 'true' and it seems that C99 defines that as a reserved
155 keyword to be used with 'bool' variable types. That means you could
156 not include C99-style headers which use bools without triggering
157 build errors in sieve.c.
159 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
160 as mailq or other aliases. Changed the --help handling significantly
161 to do whats expected. exim_usage() emits usage/help information.
163 SC/01 Added the -bylocaldomain option to eximstats.
165 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
167 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
169 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
175 PH/01 Another patch from the Sieve maintainer.
177 PH/02 When an IPv6 address is converted to a string for single-key lookup
178 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
179 dots are used instead of colons so that keys in lsearch files need not
180 contain colons. This was done some time before quoting was made available
181 in lsearch files. However, iplsearch files do require colons in IPv6 keys
182 (notated using the quote facility) so as to distinguish them from IPv4
183 keys. This meant that lookups for IP addresses in host lists did not work
184 for iplsearch lookups.
186 This has been fixed by arranging for IPv6 addresses to be expressed with
187 colons if the lookup type is iplsearch. This is not incompatible, because
188 previously such lookups could never work.
190 The situation is now rather anomolous, since one *can* have colons in
191 ordinary lsearch keys. However, making the change in all cases is
192 incompatible and would probably break a number of configurations.
194 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
197 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
198 conversion specification without a maximum field width, thereby enabling
199 a rogue spamd server to cause a buffer overflow. While nobody in their
200 right mind would setup Exim to query an untrusted spamd server, an
201 attacker that gains access to a server running spamd could potentially
202 exploit this vulnerability to run arbitrary code as the Exim user.
204 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
205 $primary_hostname instead of what libspf2 thinks the hosts name is.
207 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
208 a directory entry by the name of the lookup key. Previously, if a
209 symlink pointed to a non-existing file or a file in a directory that
210 Exim lacked permissions to read, a lookup for a key matching that
211 symlink would fail. Now it is enough that a matching directory entry
212 exists, symlink or not. (Bugzilla 503.)
214 PH/03 The body_linecount and body_zerocount variables are now exported in the
217 PH/04 Added the $dnslist_matched variable.
219 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
220 This means they are set thereafter only if the connection becomes
223 PH/06 Added the client_condition to authenticators so that some can be skipped
224 by clients under certain conditions.
226 PH/07 The error message for a badly-placed control=no_multiline_responses left
227 "_responses" off the end of the name.
229 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
231 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
232 (without spaces) instead of just copying the configuration text.
234 PH/10 Added the /noupdate option to the ratelimit ACL condition.
236 PH/11 Added $max_received_linelength.
238 PH/12 Added +ignore_defer and +include_defer to host lists.
240 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
241 way in which PCRE > 7.0 is built.
243 PH/14 Implemented queue_only_load_latch.
245 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
246 MAIL command. The effect was to mangle the value on 64-bit systems.
248 PH/16 Another patch from the Sieve maintainer.
250 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
252 PH/18 If a system quota error occurred while trying to create the file for
253 a maildir delivery, the message "Mailbox is full" was not appended to the
254 bounce if the delivery eventually timed out. Change 4.67/27 below applied
255 only to a quota excession during the actual writing of the file.
257 PH/19 It seems that peer DN values may contain newlines (and other non-printing
258 characters?) which causes problems in log lines. The DN values are now
259 passed through string_printing() before being added to log lines.
261 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
262 and InterBase are left for another time.)
264 PH/21 Added message_body_newlines option.
266 PH/22 Guard against possible overflow in moan_check_errorcopy().
268 PH/23 POSIX allows open() to be a macro; guard against that.
270 PH/24 If the recipient of an error message contained an @ in the local part
271 (suitably quoted, of course), incorrect values were put in $domain and
272 $local_part during the evaluation of errors_copy.
278 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
279 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
282 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
283 issue a MAIL command.
285 PH/02 In an ACL statement such as
287 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
289 if a client was not listed at all, or was listed with a value other than
290 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
291 the condition was not true (as it should be), so access was not denied.
292 The bug was that the ! inversion was incorrectly passed on to the second
293 item. This has been fixed.
295 PH/03 Added additional dnslists conditions == and =& which are different from
296 = and & when the dns lookup returns more than one IP address.
298 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
299 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
301 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
302 FSYNC, which compiles an option called disable_fsync that allows for
303 bypassing fsync(). The documentation is heavily laced with warnings.
305 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
307 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
308 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
309 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
310 including adding "make clean"; (3) Added -fPIC when compiling the test
311 dynamically loaded module, to get rid of a warning.
313 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
314 message fails, move_frozen_messages = true and ignore_bounce_errors_after
315 = 0s. The bug is otherwise harmless.
317 PH/07 There was a bug in the dovecot authenticator such that the value of
318 $auth1 could be overwritten, and so not correctly preserved, after a
319 successful authentication. This usually meant that the value preserved by
320 the server_setid option was incorrect.
322 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
324 PH/09 Installed PCRE release 7.0.
326 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
327 run for batched SMTP input. It is now run at the start of every message
328 in the batch. While fixing this I discovered that the process information
329 (output by running exiwhat) was not always getting set for -bs and -bS
330 input. This is fixed, and it now also says "batched" for BSMTP.
332 PH/11 Added control=no_pipelining.
334 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
335 patch, slightly modified), and move the expansion of helo_data till after
336 the connection is made in the smtp transport (so it can use these
339 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
341 PH/14 Added log_selector = +pid.
343 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
345 PH/16 Add ${if forany and ${if forall.
347 PH/17 Added dsn_from option to vary the From: line in DSNs.
349 PH/18 Flush SMTP output before performing a callout, unless control =
350 no_callout_flush is set.
352 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
353 was true (the default) a successful delivery failed to delete the retry
354 item, thus causing premature timeout of the address. The bug is now
357 PH/20 Added hosts_avoid_pipelining to the smtp transport.
359 PH/21 Long custom messages for fakedefer and fakereject are now split up
360 into multiline reponses in the same way that messages for "deny" and
361 other ACL rejections are.
363 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
364 with slight modification.
366 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
367 draft, changing the syntax and factoring some duplicate code".
369 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
370 for deliveries of the second and subsequent messages over the same SMTP
373 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
374 ${reduce, with only minor "tidies".
376 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
378 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
379 expansion side effects.
381 PH/27 When a message times out after an over-quota error from an Exim-imposed
382 quota, the bounce message says "mailbox is full". This message was not
383 being given when it was a system quota that was exceeded. It now should
386 MH/03 Made $recipients available in local_scan(). local_scan() already has
387 better access to the recipient list through recipients_list[], but
388 $recipients can be useful in postmaster-provided expansion strings.
390 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
391 in the case of a MAIL command with additional options following the
392 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
393 were accidentally chopped off.
395 PH/29 SMTP synchronization checks are implemented when a command is read -
396 there is a check that no more input is waiting when there shouldn't be
397 any. However, for some commands, a delay in an ACL can mean that it is
398 some time before the response is written. In this time, more input might
399 arrive, invalidly. So now there are extra checks after an ACL has run for
400 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
401 pipelining has not been advertised.
403 PH/30 MH's patch to allow iscntrl() characters to be list separators.
405 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
406 returned in the SMTP response when smtp_return_error_details was false.
409 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
410 instead of the C I/O that was originally supplied, because problems were
413 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
414 Exim which did not show up earlier: it was assuming that a call to
415 SSL_CTX_set_info_callback() might give an error value. In fact, there is
416 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
417 was a macro that became an assignment, so it seemed to work. This has
418 changed to a proper function call with a void return, hence the compile
419 error. Exim's code has been fixed.
421 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
424 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
426 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
428 PH/37 If a message is not accepted after it has had an id assigned (e.g.
429 because it turns out to be too big or there is a timeout) there is no
430 "Completed" line in the log. When some messages of this type were
431 selected by exigrep, they were listed as "not completed". Others were
432 picked up by some special patterns. I have improved the selection
433 criteria to be more general.
435 PH/38 The host_find_failed option in the manualroute router can now be set
436 to "ignore", to completely ignore a host whose IP address cannot be
437 found. If all hosts are ignored, the behaviour is controlled by the new
438 host_all_ignored option.
440 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
441 homing or because of multiple MX records with /mx) generated more than
442 one IP address, and the following item turned out to be the local host,
443 all the secondary addresses of the first item were incorrectly removed
444 from the list, along with the local host and any following hosts (which
445 is what is supposed to happen).
447 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
448 whoever called Exim into the -H file. In the case of the daemon it was
449 behaving confusingly. When first started, it used values for whoever
450 started the daemon, but after a SIGHUP it used the Exim user (because it
451 calls itself on a restart). I have changed the code so that it now always
454 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
455 message are rejected with the same error (e.g. no authentication or bad
456 sender address), and a DATA command is nevertheless sent (as can happen
457 with PIPELINING or a stupid MUA), the error message that was given to the
458 RCPT commands is included in the rejection of the DATA command. This is
459 intended to be helpful for MUAs that show only the final error to their
462 PH/42 Another patch from the Sieve maintainer.
464 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
465 Eximstats - Fixed some broken HTML links and added missing column headers
467 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
468 columns for Rejects, Temp Rejects, Ham, and Spam rows.
470 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
472 PH/43 Yet another patch from the Sieve maintainer.
474 PH/44 I found a way to check for a TCP/IP connection going away before sending
475 the response to the final '.' that terminates a message, but only in the
476 case where the client has not sent further data following the '.'
477 (unfortunately, this is allowed). However, in many cases there won't be
478 any further data because there won't be any more messages to send. A call
479 to select() can be used: if it shows that the input is "ready", there is
480 either input waiting, or the socket has been closed. An attempt to read
481 the next input character can distinguish the two cases. Previously, Exim
482 would have sent an OK response which the client would never have see.
483 This could lead to message repetition. This fix should cure that, at
484 least in a lot of common cases.
486 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
487 advertised in response to EHLO.
493 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
494 fixed by 4.65/MH/01 (is this a record?) are fixed:
496 (i) An empty string was always treated as zero by the numeric comparison
497 operators. This behaviour has been restored.
499 (ii) It is documented that the numeric comparison operators always treat
500 their arguments as decimal numbers. This was broken in that numbers
501 starting with 0 were being interpreted as octal.
503 While fixing these problems I realized that there was another issue that
504 hadn't been noticed. Values of message_size_limit (both the global option
505 and the transport option) were treated as octal if they started with 0.
506 The documentation was vague. These values are now always treated as
507 decimal, and I will make that clear in the documentation.
513 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
514 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
517 MH/01 Don't check that the operands of numeric comparison operators are
518 integers when their expansion is in "skipping" mode (fixes bug
519 introduced by 4.64-PH/07).
521 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
522 child addresses, Exim now panics and dies. Previously, because the count
523 is held in a short int, deliveries were likely to be lost. As such a
524 large number of recipients for a single message is ridiculous
525 (performance will be very, very poor), I have chosen to impose a limit
526 rather than extend the field.
532 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
533 leftover -K file (the existence of which was triggered by #402).
534 While we were at it, introduced process PID as part of the -K
535 filename. This should rule out race conditions when creating
538 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
539 processing considerably. Previous code took too long for large mails,
540 triggering a timeout which in turn triggers #401.
542 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
543 in the DK code in transports.c. sendfile() is not really portable,
544 hence the _LINUX specificness.
546 TF/01 In the add_headers option to the mail command in an Exim filter,
547 there was a bug that Exim would claim a syntax error in any
548 header after the first one which had an odd number of characters
551 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
552 callout verification, Exim cached a "reject" for the entire domain. This
553 is correct for most verifications, but it is not correct for a recipient
554 verification with use_sender or use_postmaster set, because in that case
555 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
556 case of MAIL FROM:<> rejection from other early rejections (e.g.
557 rejection of HELO). When verifying a recipient using a non-null MAIL
558 address, the cache is ignored if it shows MAIL FROM:<> rejection.
559 Whatever the result of the callout, the value of the domain cache is
560 left unchanged (for any other kind of callout, getting as far as trying
561 RCPT means that the domain itself is ok).
563 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
566 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
567 manifest itself as EPIPE rather than ECONNECT. When tidying away a
568 session, the daemon ignores ECONNECT errors and logs others; it now
569 ignores EPIPE as well.
571 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
572 (quoted-printable decoding).
574 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
575 later the small subsequent patch to fix an introduced bug.
577 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
579 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
581 PH/08 An error is now given if message_size_limit is specified negative.
583 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
584 to be given (somewhat) arbitrary names.
586 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
589 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
590 miscellaneous code fixes
592 PH/10 Added the log_reject_target ACL modifier to specify where to log
595 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
596 hostname. This is wrong, because it relates to the incoming message (and
597 probably the interface on which it is arriving) and not to the outgoing
598 callout (which could be using a different interface). This has been
599 changed to use the value of the helo_data option from the smtp transport
600 instead - this is what is used when a message is actually being sent. If
601 there is no remote transport (possible with a router that sets up host
602 addresses), $smtp_active_hostname is used.
604 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
605 tweaks were necessary in order to get it to work (see also 21 below):
606 (a) The code assumed that strncpy() returns a negative number on buffer
607 overflow, which isn't the case. Replaced with Exim's string_format()
609 (b) There were several signed/unsigned issues. I just did the minimum
610 hacking in of casts. There is scope for a larger refactoring.
611 (c) The code used strcasecmp() which is not a standard C function.
612 Replaced with Exim's strcmpic() function.
613 (d) The code set only $1; it now sets $auth1 as well.
614 (e) A simple test gave the error "authentication client didn't specify
615 service in request". It would seem that Dovecot has changed its
616 interface. Fortunately there's a specification; I followed it and
617 changed what the client sends and it appears to be working now.
619 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
622 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
623 address A is aliased to B and C, where B exists and C does not. Without
624 -v the output is "A verified" because verification stops after a
625 successful redirection if more than one address is generated. However,
626 with -v the child addresses are also verified. Exim was outputting "A
627 failed to verify" and then showing the successful verification for C,
628 with its parentage. It now outputs "B failed to verify", showing B's
629 parentage before showing the successful verification of C.
631 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
632 look up a TXT record in a specific list after matching in a combined
635 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
636 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
637 they consult the DNS. I had assumed they would set it the way they
638 wanted; and indeed my experiments on Linux seem to show that in some
639 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
640 To be on the safe side, however, I have now made the interface to
641 host_find_byname() similar to host_find_bydns(), with an argument
642 containing the DNS resolver options. The host_find_byname() function now
643 sets these options at its start, just as host_find_bydns() does. The smtp
644 transport options dns_qualify_single and dns_search_parents are passed to
645 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
646 of host_find_byname() use the default settings of RES_DEFNAMES
647 (qualify_single) but not RES_DNSRCH (search_parents).
649 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
650 spool_read_header() do less string testing, by means of a preliminary
651 switch on the second character of optional "-foo" lines. (This is
652 overdue, caused by the large number of possibilities that now exist.
653 Originally there were few.) While I was there, I also converted the
654 str(n)cmp tests so they don't re-test the leading "-" and the first
655 character, in the hope this might squeeze out yet more improvement.
657 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
658 flag allowing group syntax was set by the header_syntax check but not
659 turned off, possible causing trouble later; (2) The flag was not being
660 set at all for the header_verify test, causing "group"-style headers to
661 be rejected. I have now set it in this case, and also caused header_
662 verify to ignore an empty address taken from a group. While doing this, I
663 came across some other cases where the code for allowing group syntax
664 while scanning a header line wasn't quite right (mostly, not resetting
665 the flag correctly in the right place). These bugs could have caused
666 trouble for malformed header lines. I hope it is now all correct.
668 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
669 with the "reply" argument non-NULL. The code, however (which originally
670 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
671 but it didn't always do it. This confused somebody who was copying the
672 code for some other use. I have removed all the tests.
674 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
675 feature that was used to support insecure browsers during the U.S. crypto
676 embargo. It requires special client support, and Exim is probably the
677 only MTA that supported it -- and would never use it because real RSA is
678 always available. This code has been removed, because it had the bad
679 effect of slowing Exim down by computing (never used) parameters for the
680 RSA_EXPORT functionality.
682 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
683 authenticator to fail if there's a tab character in the incoming data
684 (there should never be unless someone is messing about, as it's supposed
685 to be base64-encoded). Also added, on Timo's advice, the "secured" option
686 if the connection is using TLS or if the remote IP is the same as the
687 local IP, and the "valid-client-cert option" if a client certificate has
690 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
691 authenticators. This can be used for authorization after authentication
692 succeeds. (In the case of plaintext, it servers for both authentication
695 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
696 if any retry times were supplied.
698 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
699 connection, where there is no client IP address to check. In this
700 situation, the verify now always succeeds.
702 PH/25 Applied John Jetmore's -Mset patch.
704 PH/26 Added -bem to be like -Mset, but loading a message from a file.
706 PH/27 In a string expansion for a processed (not raw) header when multiple
707 headers of the same name were present, leading whitespace was being
708 removed from all of them, but trailing whitespace was being removed only
709 from the last one. Now trailing whitespace is removed from each header
710 before concatenation. Completely empty headers in a concatenation (as
713 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
714 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
716 PH/29 [Removed. This was a change that I later backed out, and forgot to
717 correct the ChangeLog entry (that I had efficiently created) before
718 committing the later change.]
720 PH/30 Exim was sometimes attempting to deliver messages that had suffered
721 address errors (4xx response to RCPT) over the same connection as other
722 messages routed to the same hosts. Such deliveries are always "forced",
723 so retry times are not inspected. This resulted in far too many retries
724 for the affected addresses. The effect occurred only when there were more
725 hosts than the hosts_max_try setting in the smtp transport when it had
726 the 4xx errors. Those hosts that it had tried were not added to the list
727 of hosts for which the message was waiting, so if all were tried, there
728 was no problem. Two fixes have been applied:
730 (i) If there are any address or message errors in an SMTP delivery, none
731 of the hosts (tried or untried) are now added to the list of hosts
732 for which the message is waiting, so the message should not be a
733 candidate for sending over the same connection that was used for a
734 successful delivery of some other message. This seems entirely
735 reasonable: after all the message is NOT "waiting for some host".
736 This is so "obvious" that I'm not sure why it wasn't done
737 previously. Hope I haven't missed anything, but it can't do any
738 harm, as the worst effect is to miss an optimization.
740 (ii) If, despite (i), such a delivery is accidentally attempted, the
741 routing retry time is respected, so at least it doesn't keep
742 hammering the server.
744 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
745 in ${readsocket because some servers need this prod.
747 PH/32 Added some extra debug output when updating a wait-xxx database.
749 PH/33 The hint "could be header name not terminated by colon", which has been
750 given for certain expansion errors for a long time, was not being given
751 for the ${if def:h_colon_omitted{... case.
753 PH/34 The spec says: "With one important exception, whenever a domain list is
754 being scanned, $domain contains the subject domain." There was at least
755 one case where this was not true.
757 PH/35 The error "getsockname() failed: connection reset by peer" was being
758 written to the panic log as well as the main log, but it isn't really
759 panic-worthy as it just means the connection died rather early on. I have
760 removed the panic log writing for the ECONNRESET error when getsockname()
763 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
764 runs only) independently of the message's sender address. This meant
765 that, if the 4xx error was in fact related to the sender, a different
766 message to the same recipient with a different sender could confuse
767 things. In particualar, this can happen when sending to a greylisting
768 server, but other circumstances could also provoke similar problems.
769 I have changed the default so that the retry time for these errors is now
770 based a combination of the sender and recipient addresses. This change
771 can be overridden by setting address_retry_include_sender=false in the
774 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
775 remote server are returned as part of bounce messages. This was not
776 happening for LMTP over a pipe (the lmtp transport), but now it is the
777 same for both kinds of LMTP.
779 PH/38 Despite being documented as not happening, Exim was rewriting addresses
780 in header lines that were in fact CNAMEs. This is no longer the case.
782 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
783 and queue runs started by the daemon processed all messages. This has
784 been fixed so that -R and -S can now usefully be given with -q<time>.
786 PH/40 Import PCRE release 6.7 (fixes some bugs).
788 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
790 PH/42 Give an error if -q is specified more than once.
792 PH/43 Renamed the variables $interface_address and $interface_port as
793 $received_ip_address and $received_port, to make it clear that these
794 values apply to message reception, and not to the outgoing interface when
795 a message is delivered. (The old names remain recognized, of course.)
797 PH/44 There was no timeout on the connect() call when using a Unix domain
798 socket in the ${readsocket expansion. There now is.
800 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
801 be meaningful with "accept".
803 SC/01 Eximstats V1.43
804 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
806 SC/02 Eximstats V1.44
807 Use a glob alias rather than an array ref in the generated
808 parser. This improves both readability and performance.
810 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
811 Collect SpamAssassin and rejection statistics.
812 Don't display local sender or destination tables unless
813 there is data to show.
814 Added average volumes into the top table text output.
816 SC/04 Eximstats V1.46
817 Collect data on the number of addresses (recipients)
818 as well as the number of messages.
820 SC/05 Eximstats V1.47
821 Added 'Message too big' to the list of mail rejection
822 reasons (thanks to Marco Gaiarin).
824 SC/06 Eximstats V1.48
825 Mainlog lines which have GMT offsets and are too short to
826 have a flag are now skipped.
828 SC/07 Eximstats V1.49 (Alain Williams)
829 Added the -emptyok flag.
831 SC/08 Eximstats V1.50
832 Fixes for obtaining the IP address from reject messages.
834 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
835 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
836 whitesspace changes from 4.64-PH/27
838 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
841 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
842 are found, allow negative numbers in numeric criteria)
844 JJ/06 exipick.20061117.2, added new $message_body_missing variable
846 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
847 to match changes made in 4.64-PH/43
849 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
851 PH/47 Put in an explicit test for a DNS lookup of an address record where the
852 "domain" is actually an IP address, and force a failure. This locks out
853 those revolvers/nameservers that support "A-for-A" lookups, in
854 contravention of the specifications.
856 PH/48 When a host name was looked up from an IP address, and the subsequent
857 forward lookup of the name timed out, the host name was left in
858 $sender_host_name, contrary to the specification.
860 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
861 restricted to single-key lookups, Exim was not diagnosing an error if
862 * or *@ was used with a query-style lookup.
864 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
866 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
867 long ago, but noone interested enough thought of it. Let's just say that
868 the "1.1" means that there are some new functions that weren't there at
869 some point in the past.
871 PH/51 Error processing for expansion failure of helo_data from an smtp
872 transport during callout processing was broken.
874 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
875 tested/used via the -bh/-bhc/-bs options.
877 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
878 bug, fixed in subsequent PCRE releases).
880 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
881 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
883 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
889 SC/01 Use a glob alias rather than an array ref in eximstats generated
890 parser. This improves both readability and performance.
892 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
893 Don't display local sender or destination tables in eximstats unless
894 there is data to show.
895 Added average volumes into the eximstats top table text output.
897 SC/03 Collect data on the number of addresses (recipients) as well
898 as the number of messages in eximstats.
900 TF/01 Correct an error in the documentation for the redirect router. Exim
901 does (usually) call initgroups() when daemonizing.
903 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
904 with consistent privilege compared to when running as a daemon.
906 TF/03 Note in the spec that $authenticated_id is not set for local
907 submissions from trusted users.
909 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
910 Thanks to Dean Brooks <dean@iglou.com> for the patch.
912 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
913 by adding some example configuration directives to the default
914 configuration file. A little bit of work is required to uncomment the
915 directives and define how usernames and passwords are checked, but
916 there is now a framework to start from.
918 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
919 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
920 without this. I don't know how relevant this is to other LDAP libraries.
922 PH/02 Add the verb name to the "unknown ACL verb" error.
924 PH/03 Magnus Holmgren's patch for filter_prepend_home.
926 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
928 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
929 directory not expanded when it should be if an expanded home directory
930 was set for the address (which is overridden by the transport).
932 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
935 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
936 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
937 because it is too late at that time, and has no effect.
939 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
940 security issue with \' (bugzilla #107). I could not use the
941 PQescapeStringConn() function, because it needs a PGconn value as one of
944 PH/08 When testing addresses using -bt, indicate those final addresses that
945 are duplicates that would not cause an additional delivery. At least one
946 person was confused, thinking that -bt output corresponded to deliveries.
947 (Suppressing duplicates isn't a good idea as you lose the information
948 about possibly different redirections that led to the duplicates.)
950 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
951 systems where poll() doesn't work, in particular OS X.
953 PH/10 Added more information to debugging output for retry time not reached.
955 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
956 operations in malware.c.
958 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
961 PH/13 If write_rejectlog was set false when logging was sent to syslog with
962 syslog_duplication set false, log lines that would normally be written
963 both the the main log and to the reject log were not written to syslog at
966 PH/14 In the default configuration, change the use of "message" in ACL warn
967 statements to "add_header".
969 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
970 not followed by a command (e.g. "seen endif").
972 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
973 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
976 PH/17 Added extra conditions to the default value of delay_warning_condition
980 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
981 { match{$h_precedence:}{(?i)bulk|list|junk} } \
982 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
985 The Auto-Submitted: and various List- headers are standardised, whereas I
986 don't think Precedence: ever was.
988 PH/18 Refactored debugging code in route_finduser() to show more information,
989 in particular, the error code if getpwnam() issues one.
991 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
992 This is apparently needed in addition to the PH/07 change above to avoid
993 any possible encoding problems.
995 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
996 but not after initializing Perl.
998 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
999 output them only if debugging. By default they are written stderr,
1000 apparently, which is not desirable.
1002 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1005 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1008 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1010 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1011 authenticated or an ident call has been made. Suppress the default
1012 values for $authenticated_id and $authenticated_sender (but permit -oMai
1013 and -oMas) when testing with -bh.
1015 PH/24 Re-jigged the order of the tests in the default configuration so that the
1016 tests for valid domains and recipients precede the DNS black list and CSA
1017 tests, on the grounds that those ones are more expensive.
1019 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1020 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1021 This bug exists in every version of Exim that I still have, right back to
1024 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1025 However, an attempt to turn on case-sensitivity in a regex key by
1026 including (?-i) didn't work because the subject string was already
1027 lowercased, and the effects were non-intuitive. It turns out that a
1028 one-line patch can be used to allow (?-i) to work as expected.
1034 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1035 other effects) broke the use of negated acl sub-conditions.
1037 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1040 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1041 "Deny" causes Exim to reject the incoming connection with a 554 error.
1042 Unfortunately, if there is a major crisis, such as a disk failure,
1043 tcp-wrappers gives "deny", whereas what one would like would be some
1044 kind of temporary error. A kludge has been added to help with this.
1045 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1046 554 error is used if errno is still zero or contains ENOENT (which occurs
1047 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1050 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1052 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1053 errors. Otherwise a message that provokes a temporary error (when other
1054 messages do not) can cause a whole host to time out.
1056 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1057 addresses were routed directly to files or pipes from a redirect router.
1058 File deliveries just didn't batch; pipe deliveries might have suffered
1061 PH/06 A failure to get a lock for a hints database would erroneously always say
1062 "Failed to get write lock", even when it was really a read lock.
1064 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1065 of 0600. This has been changed to use the value of the lockfile_mode
1066 option (which defaults to 0600).
1068 PH/08 Applied small patch from the Sieve maintainer.
1070 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1071 folder from quota calculations, a direct delivery into this folder messed
1072 up the contents of the maildirsize file. This was because the regex was
1073 used only to exclude .Trash (or whatever) when the size of the mailbox
1074 was calculated. There was no check that a delivery was happening into an
1075 excluded directory. This bug has been fixed by ignoring all quota
1076 processing for deliveries into excluded directories.
1078 PH/10 Added the maildirfolder_create_regex option to appendfile.
1084 PH/01 The code for finding all the local interface addresses on a FreeBSD
1085 system running IPv6 was broken. This may well have applied to all BSD
1086 systems, as well as to others that have similar system calls. The broken
1087 code found IPv4 interfaces correctly, but gave incorrect values for the
1088 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1089 that it would not match correctly against @[] and not recognize the IPv6
1092 PH/02 The ipliteral router was not recognizing addresses of the form user@
1093 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1095 PH/03 Added disable_ipv6.
1097 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1098 decoded form, because it is most often used to construct To: headers
1099 lines in autoreplies, and the decoded form may well be syntactically
1100 invalid. However, $reply_address has leading white space removed, and all
1101 newlines turned into spaces so that the autoreply transport does not
1104 PH/05 If group was specified without a user on a router, and no group or user
1105 was specified on a transport, the group from the router was ignored.
1107 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1108 for visible compile-time settings that can be used to change these
1109 numbers, for those that want even more. Backwards compatibility with old
1110 spool files has been maintained. However, going back to a previous Exim
1111 release will lost any variables that are in spool files.
1113 PH/07 Two small changes when running in the test harness: increase delay when
1114 passing a TCP/IP connection to a new process, in case the original
1115 process has to generate a bounce, and remove special handling of
1116 127.0.0.2 (sic), which is no longer necessary.
1118 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1119 be the same on different OS.
1121 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1124 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1125 whether --show-vars was specified or not
1127 JJ/02 exipick: Added support for new ACL variable spool format introduced
1130 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1131 syntactically invalid From: or Reply-to: line, and a filter used this to
1132 generate an autoreply, and therefore failed to obtain an address for the
1133 autoreply, Exim could try to deliver to a non-existent relative file
1134 name, causing unrelated and misleading errors. What now happens is that
1135 it logs this as a hard delivery error, but does not attempt to create a
1138 PH/11 The exinext utility has a -C option for testing purposes, but although
1139 the given file was scanned by exinext itself; it wasn't being passed on
1140 when Exim was called.
1142 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1143 an end-of-file indication when reading a command response.
1145 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1146 compiled. In many other places in Exim, IPv6 addresses are always
1147 recognized, so I have changed this. It also means that IPv4 domain
1148 literals of the form [IPV4:n.n.n.n] are now always recognized.
1150 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1151 used if the router is not running as root, for example, when verifying at
1152 ACL time, or when using -bh. The debugging output from this situation was
1153 non-existent - all you got was a failure to exec. I have made two
1156 (a) Failures to set uid/gid, the current directory, or a process leader
1157 in a subprocess such as that created by queryprogram now generate
1158 suitable debugging ouput when -d is set.
1160 (b) The queryprogram router detects when it is not running as root,
1161 outputs suitable debugging information if -d is set, and then runs
1162 the subprocess without attempting to change uid/gid.
1164 PH/15 Minor change to Makefile for building test_host (undocumented testing
1167 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1168 additional section of a DNS packet that returns MX or SRV records.
1169 Instead, it always explicitly searches for A/AAAA records. This avoids
1170 major problems that occur when a DNS server includes only records of one
1171 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1172 fixed another bug: if SRV records were looked up and the corresponding
1173 address records were *not* found in the additional section, the port
1174 values from the SRV records were lost.
1176 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1177 using the correct key (the original address) when searching the retry
1178 rules in order to find which one to use for generating the retry hint.
1180 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1181 adding the default one. Similarly, if it contains a Reply-To: header, the
1182 errors_reply_to option, if set, is not used.
1184 PH/19 When calculating a retry time, Exim used to measure the "time since
1185 failure" by looking at the "first failed" field in the retry record. Now
1186 it does not use this if it is later than than the arrival time of the
1187 message. Instead it uses the arrival time. This makes for better
1188 behaviour in cases where some deliveries succeed, thus re-setting the
1189 "first failed" field. An example is a quota failure for a huge message
1190 when small messages continue to be delivered. Without this change, the
1191 "time since failure" will always be short, possible causing more frequent
1192 delivery attempts for the huge message than are intended.
1193 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1195 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1196 $1, $2, $3) because the numerical variables can be reset during some
1197 expansion items (e.g. "match"), thereby losing the authentication data.
1199 PH/21 Make -bV show the size of off_t variables so that the test suite can
1200 decide whether to run tests for quotas > 2G.
1202 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1203 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1204 filecount value is greater than 2G or if a quota value is greater than 2G
1205 on a system where the size of off_t is not greater than 4, a panic error
1208 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1209 never match. The debug and -bh output now contains an explicit error
1210 message indicating a malformed IPv4 address or mask.
1212 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1213 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1214 PH/23 above applies.
1216 PH/25 Do not write to syslog when running in the test harness. The only
1217 occasion when this arises is a failure to open the main or panic logs
1218 (for which there is an explicit test).
1220 PH/26 Added the /no_tell option to "control=freeze".
1222 PH/27 If a host name lookup failed very early in a connection, for example, if
1223 the IP address matched host_lookup and the reverse lookup yielded a name
1224 that did not have a forward lookup, an error message of the form "no IP
1225 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1226 could be logged. Now it outputs the IP address instead of "NULL".
1228 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1229 allows the sender and the authenticated sender to be set when
1230 submitting a message from within Exim. Since child_open_exim() is
1231 documented for local_scan(), the new function should be too.
1233 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1234 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1235 results in an empty string is now treated as unset.
1237 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1239 PH/31 Added sender_verify_fail logging option.
1241 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1242 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1243 tidied the source and removed it altogether.
1245 PH/33 When a queue run was abandoned because the load average was too high, a
1246 log line was always written; now it is written only if the queue_run log
1247 selector is set. In addition, the log line for abandonment now contains
1248 information about the queue run such as the pid. This is always present
1249 in "start" and "stop" lines but was omitted from the "abandon" line.
1251 PH/34 Omit spaces between a header name and the colon in the error message that
1252 is given when verify = headers_syntax fails (if there are lots of them,
1253 the message gets confusing).
1255 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1256 names, as there are now some PTR records that contain slashes. This check
1257 is only to protect against broken name servers that fall over on strange
1258 characters, so the fact that it applies to all lookups doesn't matter.
1260 PH/36 Now that the new test suite is complete, we can remove some of the
1261 special code in Exim that was needed for the old test suite. For example,
1262 sorting DNS records because real resolvers return them in an arbitrary
1263 order. The new test suite's fake resolver always returns records in the
1266 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1267 bounces) except when queue_only is set, to avoid logging races between
1268 the different processes.
1270 PH/38 Panic-die if .include specifies a non-absolute path.
1272 PH/39 A tweak to the "H" retry rule from its user.
1274 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1275 a label. They prevented compilation on older perls.
1277 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1278 a warning to be raised on newish perls.
1280 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1281 on queue. Changes to match documented behaviour of showing count of
1282 messages matching specified criteria.
1284 PH/40 Changed the default ident timeout from 30s to 5s.
1286 PH/41 Added support for the use of login_cap features, on those BSD systems
1287 that have them, for controlling the resources used by pipe deliveries.
1289 PH/42 The content-scanning code uses fopen() to create files in which to put
1290 message data. Previously it was not paying any attention to the mode of
1291 the files. Exim runs with umask(0) because the rest of the code creates
1292 files with open(), and sets the required mode explicitly. Thus, these
1293 files were ending up world-writeable. This was not a big issue, because,
1294 being within the spool directory, they were not world-accessible. I have
1295 created a function called modefopen, which takes an additional mode
1296 argument. It sets umask(777), creates the file, chmods it to the required
1297 mode, then resets the umask. All the relevant calls to fopen() in the
1298 content scanning code have been changed to use this function.
1300 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1301 to 24 hours. This avoids potential overflow problems when processing G
1302 and H retry rules. I suspect nobody ever tinkers with this value.
1304 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1306 PH/45 When the plaintext authenticator is running as a client, the server's
1307 challenges are checked to ensure they are valid base64 strings. By
1308 default, the authentication attempt is cancelled if an invalid string is
1309 received. Setting client_ignore_invalid_base64 true ignores these errors.
1310 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1311 they are received. Thus, the responses can be made to depend on the
1312 challenges. If an invalid string is ignored, an empty string is placed in
1315 PH/46 Messages that are created by the autoreply transport now contains a
1316 References: header, in accordance with RFCs 2822 and 3834.
1318 PH/47 Added authenticated_sender_force to the smtp transport.
1320 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1322 PH/49 Installed latest patch from the Sieve maintainer.
1324 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1325 was also set, the appendfile transport was unnecessarily scanning a
1326 directory of message files (e.g. for maildir delivery) to find the count
1327 of files (along with the size), even though it did not need this
1328 information. It now does the scan only if it needs to find either the
1329 size of the count of files.
1331 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1333 PH/52 Two bugs concerned with error handling when the smtp transport is
1336 (i) Exim was not creating retry information for temporary errors given
1337 for individual recipients after the DATA command when the smtp transport
1338 was used in LMTP mode. This meant that they could be retried too
1339 frequently, and not timed out correctly.
1341 (ii) Exim was setting the flag that allows error details to be returned
1342 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1343 recipients that were returned after the DATA command.
1345 PH/53 This is related to PH/52, but is more general: for any failing address,
1346 when detailed error information was permitted to be returned to the
1347 sender, but the error was temporary, then after the final timeout, only
1348 "retry timeout exceeded" was returned. Now it returns the full error as
1349 well as "retry timeout exceeded".
1351 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1352 do this, and (what is worse) MTAs that accept it.
1354 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1355 will now be deprecated.
1357 PH/56 New os.c-cygwin from the Cygwin maintainer.
1359 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1360 formats (previously only available in exim formats via -bpr, -bpru,
1361 and -bpra. Now also available in native and exiqgrep formats)
1363 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1364 with very large, slow to parse queues
1366 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1368 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1370 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1371 responses to authentication challenges, though it was showing the
1372 challenges; (ii) I've removed the CR characters from the debug output for
1375 PH/58 Allow for the insertion of a newline as well as a space when a string
1376 is turned into more than one encoded-word during RFC 2047 encoding. The
1377 Sieve code now uses this.
1379 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1380 data_4xx, lost_connection, tls_required.
1382 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1383 message was being sent as an SMTP response.
1385 PH/61 Add -l and -k options to exicyclog.
1387 PH/62 When verifying, if an address was redirected to one new address, so that
1388 verification continued, and the new address failed or deferred after
1389 having set something in $address_data, the value of $address_data was not
1390 passed back to the ACL. This was different to the case when no
1391 redirection occurred. The value is now passed back in both cases.
1393 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1394 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1395 use login_cap.h, so on its own it isn't the distinguishing feature. The
1396 new name refers directly to the setclassresources() function.
1398 PH/65 Added configuration files for NetBSD3.
1400 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1402 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1403 is preferred over IPv4.
1405 PH/68 The bounce_return_message and bounce_return_body options were not being
1406 honoured for bounces generated during the reception of non-SMTP messages.
1407 In particular, this applied to messages rejected by the ACL. This bug has
1408 been fixed. However, if bounce_return_message is true and bounce_return_
1409 body is false, the headers that are returned for a non-SMTP message
1410 include only those that have been read before the error was detected.
1411 (In the case of an ACL rejection, they have all been read.)
1413 PH/69 The HTML version of the specification is now built in a directory called
1414 spec_html instead of spec.html, because the latter looks like a path with
1415 a MIME-type, and this confuses some software.
1417 PH/70 Catch two compiler warnings in sieve.c.
1419 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1420 function verify_get_ident() calls ip_connect() to connect a socket, but
1421 if the "connect()" function timed out, ip_connect() used to close the
1422 socket. However, verify_get_ident() also closes the socket later, and in
1423 between Exim writes to the log, which may get opened at this point. When
1424 the socket was closed in ip_connect(), the log could get the same file
1425 descriptor number as the socket. This naturally causes chaos. The fix is
1426 not to close the socket in ip_connect(); the socket should be closed by
1427 the function that creates it. There was only one place in the code where
1428 this was missing, in the iplookup router, which I don't think anybody now
1429 uses, but I've fixed it anyway.
1431 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1432 well as to direct DNS lookups. Otherwise the handling of names in host
1433 lists is inconsistent and therefore confusing.
1439 PH/01 Two changes to the default runtime configuration:
1441 (1) Move the checks for relay_from_hosts and authenticated clients from
1442 after to before the (commented out) DNS black list checks.
1444 (2) Add control=submission to the relay_from_hosts and authenticated
1445 clients checks, on the grounds that messages accepted by these
1446 statements are most likely to be submissions.
1448 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1450 (1) Generate an error if the third argument for the ${prvs expansion is
1453 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1456 (3) Reset the variables that are obtained from the first argument of
1457 ${prvscheck and used in the second argument before leaving the code,
1458 because their memory is reclaimed, so using them afterwards may do
1461 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1462 one (it's much easier than Tom thought :-).
1464 (5) Because of (4), we can now allow for the use of $prvscheck_result
1465 inside the third argument.
1467 PH/03 For some reason, the default setting of PATH when running a command from
1468 a pipe transport was just "/usr/bin". I have changed it to
1471 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1472 anything to be listed in the output from -bV.
1474 PH/05 When a filter generated an autoreply, the entire To: header line was
1475 quoted in the delivery log line, like this:
1477 => >A.N.Other <ano@some.domain> <original@ddress> ...
1479 This has been changed so that it extracts the operative address. There
1480 may be more than one such address. If so, they are comma-separated, like
1483 => >ano@some.domain,ona@other.domain <original@ddress> ...
1485 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1486 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1487 not being looked up in the rDNS to get a host name, Exim was showing the
1488 IP address twice in Received: lines, even though the IP addresses were
1489 identical. For example:
1491 Received: from [1.2.3.4] (helo=[1.2.3.4])
1493 However, if the real host name was known, it was omitting the HELO data
1494 if it matched the actual IP address. This has been tidied up so that it
1495 doesn't show the same IP address twice.
1497 PH/07 When both +timestamp and +memory debugging was on, the value given by
1498 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1499 called by the debug printing, thereby overwriting the timestamp buffer.
1500 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1502 PH/08 When the original message was included in an autoreply transport, it
1503 always said "this is a copy of the message, including all the headers",
1504 even if body_only or headers_only was set. It now gives an appropriate
1507 PH/09 Applied a patch from the Sieve maintainer which:
1509 o fixes some comments
1510 o adds the (disabled) notify extension core
1511 o adds some debug output for the result of if/elsif tests
1512 o points to the current vacation draft in the documentation
1513 and documents the missing references header update
1517 o fixes a bug in processing the envelope test (when testing
1518 multiple envelope elements, the last element determinted the
1521 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1522 Electronic Mail") by including:
1524 Auto-submitted: auto-generated
1526 in the messages that it generates (bounce messages and others, such as
1527 warnings). In the case of bounce messages for non-SMTP mesages, there was
1528 also a typo: it was using "Auto_submitted" (underscore instead of
1529 hyphen). Since every message generated by Exim is necessarily in response
1530 to another message, thes have all been changed to:
1532 Auto-Submitted: auto-replied
1534 in accordance with these statements in the RFC:
1536 The auto-replied keyword:
1538 - SHOULD be used on messages sent in direct response to another
1539 message by an automatic process,
1541 - MUST NOT be used on manually-generated messages,
1543 - MAY be used on Delivery Status Notifications (DSNs) and Message
1544 Disposition Notifications (MDNs),
1546 - MUST NOT be used on messages generated by automatic or periodic
1547 processes, except for messages which are automatic responses to
1550 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1551 to the default Received: header definition.
1553 PH/12 Added log selector acl_warn_skipped (default on).
1555 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1556 variables because (a) they are in the wrong storage pool and (b) even if
1557 they were copied, it wouldn't work properly because of the caching.
1559 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1560 checking when decoding. Apparently there are clients that generate
1561 overlong encoded strings. Why am I not surprised?
1563 PH/15 If the first argument of "${if match_address" was not empty, but did not
1564 contain an "@" character, Exim crashed. Now it writes a panic log message
1565 and treats the condition as false.
1567 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1569 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1570 extension "envelope-auth". The code is finished and in agreement with
1571 other implementations, but there is no documentation so far and in fact,
1572 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1573 not changing the active code.
1575 Print executed "if" and "elsif" statements when debugging is used. This
1576 helps a great deal to understand what a filter does.
1578 Document more things not specified clearly in RFC3028. I had all this
1579 sorted out, when out of a sudden new issues came to my mind. Oops."
1581 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1584 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1585 canonical form (as documented). However, after a host name lookup from
1586 the IP address, check_host() was doing a simple string comparison with
1587 addresses acquired from the DNS when checking that the found name did
1588 have the original IP as one of its addresses. Since any found IPv6
1589 addresses are likely to be in abbreviated form, the comparison could
1590 fail. Luckily, there already exists a function for doing the comparison
1591 by converting both addresses to binary, so now that is used instead of
1592 the text comparison.
1594 PH/20 There was another similar case to PH/19, when a complete host name was
1595 given in a host list; looking up its IP address could give an abbreviated
1596 form, whereas the current host's name might or might not be abbreviated.
1597 The same fix has been applied.
1603 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1604 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1607 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1608 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1610 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1612 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1613 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1614 Also, for "Auto-Submitted", treat anything other than "no" as
1615 non-personal, in accordance with RFC 3834. (Previously it treated
1616 anything starting "auto-" as non-personal.)
1618 TF/01 The control=submission/name=... option had a problem with syntax
1619 errors if the name included a slash character. The /name= option
1620 now slurps the rest of the string, so it can include any characters
1621 but it must come last in the list of options (after /sender_retain
1624 PH/05 Some modifications to the interface to the fake nameserver for the new
1632 TK/01 Added the "success_on_redirect" address verification option. See
1633 NewStuff for rationale and an example.
1635 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1637 PH/02 Patch to exigrep to allow it to work on syslog lines.
1639 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1640 fread() to skip over the body file's header line, because in Cygwin the
1641 header line is locked and is inaccessible.
1643 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1644 co-exist for some time) to make it clear that it is the Exim ID that is
1645 referenced, not the Message-ID: header line.
1647 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1648 string_format() function, because snprintf() does not exist on all
1651 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1652 ${stat: expansion item.
1654 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1655 protocol synchronization error", to keep the pedants happy.
1657 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1658 well as for IRIX systems, when gcc is being used. See the host.c source
1661 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1663 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1666 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1667 transport and to the smtp transport in LMTP mode.
1669 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1671 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1672 run a filter in a subprocess. This could lead to confusion in subsequent
1673 lookups in the parent process. There should also be a search_tidyup() at
1674 the end of the subprocess.
1676 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1677 only if the host matched helo_try_verify_hosts, which caused the
1678 verification to occur when the EHLO/HELO command was issued. The ACL just
1679 tested the remembered result. Now, if a previous verification attempt has
1680 not happened, "verify = helo" does it there and then.
1682 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1684 TK/03 Fix log output including CR from clamd.
1686 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1687 exist provoked a memory error which could cause a segfault.
1689 PH/15 Installed PCRE 6.2
1691 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1693 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1694 of the problem. Specifically, suggested +O2 rather than +O1 for the
1697 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1699 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1700 record, the port was not correctly logged when the outgoing_port log
1701 selector was set (it logged the transort's default port).
1703 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1704 fallback_hosts, and "hosts" in the smtp transport.
1706 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1707 host errors such as "Connection refused".
1709 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1710 authentication with radiusclient 0.4.9:
1712 - Error returned from rc_read_config was caught wrongly
1713 - Username/password not passed on to radius server due to wrong length.
1715 The presumption is that some radiusclient API changes for 4.51/PH/17
1716 were not taken care of correctly. The code is still untested by me (my
1717 Linux distribution still has 0.3.2 of radiusclient), but it was
1718 contributed by a Radius user.
1720 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1721 expanding the "port" option of the smtp transport.
1723 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1724 while reading a MIME header. Thanks to Tom Hughes for a patch.
1726 PH/24 Include config.h inside local_scan.h so that configuration settings are
1729 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1730 that in an ACL for RCPT (for example), you can examine exactly what was
1733 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1734 commands, but it was not correctly comparing the address with the actual
1735 client host address. Thus, it would show the EHLO address in Received:
1736 header lines when this was not necessary.
1738 PH/27 Added the % operator to ${eval:}.
1740 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1741 it should be ignoring failures (because with -C, for example, it has lost
1742 privilege). It wasn't ignoring creation failures other than "already
1745 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1748 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1749 because an input line was too long, either on its own, or by virtue of
1750 too many continuations, the temporary file was not being removed, and the
1751 return code was incorrect.
1753 PH/31 Missing "BOOL" in function definition in filtertest.c.
1755 PH/32 Applied Sieve patches from the maintainer.
1757 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1759 PH/33 Added "verify = not_blind".
1761 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1762 Local/Makefile (with some defaults set). These are used in built scripts
1763 such as exicyclog, but they have never been used in the exim_install
1764 script (though there are many overriding facilities there). I have
1765 arranged that the exim_install script now takes note of these two
1768 PH/35 Installed configuration files for Dragonfly.
1770 PH/36 When a locally submitted message by a trusted user did not contain a
1771 From: header, and the sender address was obtained from -f or from an SMTP
1772 MAIL command, and the trusted user did not use -F to supply a sender
1773 name, $originator_name was incorrectly used when constructing a From:
1774 header. Furthermore, $originator_name was used for submission mode
1775 messages from external hosts without From: headers in a similar way,
1776 which is clearly wrong.
1778 PH/37 Added control=suppress_local_fixups.
1780 PH/38 When log_selector = +received_sender was set, and the addition of the
1781 sender made the log line's construction buffer exactly full, or one byte
1782 less than full, an overflow happened when the terminating "\n" was
1785 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1786 when the result of a list match is failure because a DNS lookup failed.
1788 PH/40 RM_COMMAND is now used in the building process.
1790 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1791 the "build-* directories that it finds.
1793 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1794 address in a domain literal was a prefix of an interface address.
1796 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1797 when verifying a sender address, unless rewrite_headers is false.
1799 PH/44 Wrote a long comment about why errors_to addresses are verified as
1800 recipients, not senders.
1802 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1803 the ratelimit ACL was added.
1805 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1807 PH/46 Added extra information about PostgreSQL errors to the error string.
1809 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1810 avoiding the need to install special zones in a real server. This is
1811 backwards compatible; if it can't find the fake resolver, it drops back.
1812 Thus, both old and new test suites can be run.
1814 TF/02 Added util/ratelimit.pl
1816 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1817 clock is set back in time.
1819 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1820 Candler <B.Candler@pobox.com>.
1822 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1823 OK for addresses that are the result of redirections.
1825 PH/48 A number of further additions for the benefit of the new test suite,
1826 including a fake gethostbyname() that interfaces to the fake DNS resolver
1829 TF/06 The fix for widen_domains has also been applied to qualify_single and
1830 search_parents which are the other dnslookup options that can cause
1833 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1836 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1838 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1839 service" if it encountered an MX record with an empty target hostname.
1840 The message is now "an MX or SRV record indicated no SMTP service".
1842 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1843 if the DNS of the sending site is misconfigured. This is quite a
1844 common situation. This change restores the behaviour of treating a
1845 helo verification defer as a failure.
1847 PH/51 If self=fail was set on a router, the bounce message did not include the
1848 actual error message.
1854 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1856 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1857 command itself ended in error, the underlying message about the transport
1858 filter timeout was being overwritten with the pipe command error. Now the
1859 underlying error message should be appended to the second error message.
1861 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1863 PH/02 Reduce the amount of output that "make" produces by default. Full output
1864 can still be requested.
1866 PH/03 The warning log line about a condition test deferring for a "warn" verb
1867 was being output only once per connection, rather than after each
1868 occurrence (because it was using the same function as for successful
1869 "warn" verbs). This seems wrong, so I have changed it.
1871 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1872 it should not have, which might have caused a crash in the right
1873 circumstances, but probably never did.
1875 PH/04 Installed a modified version of Tony Finch's patch to make submission
1876 mode fix the return path as well as the Sender: header line, and to
1877 add a /name= option so that you can make the user's friendly name appear
1880 TF/03 Added the control = fakedefer ACL modifier.
1882 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1883 Mark Lowes for thorough testing.
1885 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1887 TK/03 Merged latest SRS patch from Miles Wilton.
1889 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1890 with the definition in sysexits.h (which is #included earlier).
1891 Fortunately, Exim does not actually use EX_OK. The code used to try to
1892 preserve the sysexits.h value, by assumimg that macro definitions were
1893 scanned for macro replacements. I have been disabused of this notion,
1894 so now the code just undefines EX_OK before #including unistd.h.
1896 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1897 in the smtp transport. When a block could not be written in a single
1898 write() function, the timeout was being re-applied to each part-write.
1899 This seems wrong - if the receiver was accepting one byte at a time it
1900 would take for ever. The timeout is now adjusted when this happens. It
1901 doesn't have to be particularly precise.
1903 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1904 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1906 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1907 without a domain if the check to <postmaster@domain> fails.
1909 SC/01 Eximstats: added -xls and the ability to specify output files
1910 (patch written by Frank Heydlauf).
1912 SC/02 Eximstats: use FileHandles for outputing results.
1914 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1916 SC/04 Eximstats: fixed display of large numbers with -nvr option
1918 SC/05 Eximstats: fixed merging of reports with empty tables.
1920 SC/06 Eximstats: added the -include_original_destination flag
1922 SC/07 Eximstats: removed tabs and trailing whitespace.
1924 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1926 TK/06 MBOX spool code: Add real "From " MBOX separator line
1927 so the .eml file is really in mbox format (even though
1928 most programs do not really care). Patch from Alex Miller.
1930 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1931 The latter is generated from $received_to and is only set if the
1932 message has one envelope recipient. SA can use these headers,
1933 obviously out-of-the-box. Patch from Alex Miller.
1935 PH/08 The ${def test on a variable was returning false if the variable's
1936 value was "0", contrary to what the specification has always said!
1937 The result should be true unless the variable is empty.
1939 PH/09 The syntax error of a character other than { following "${if
1940 def:variable_name" (after optional whitespace) was not being diagnosed.
1941 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1942 accidental colon was present, for example, could give incorrect results.
1944 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1945 result is used (not including appendfile, where other changes are about
1948 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1949 This involved changing a lot of size variables from int to off_t. It
1950 should work with maildirs and everything.
1952 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1953 spamd dying while we are connected to it.
1955 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1958 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1959 function for BDB 4.3.
1961 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1963 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1964 the sources for PCRE have been re-arranged and more files are now
1967 PH/15 The code I had for printing potentially long long variables in PH/11
1968 above was not the best (it lost precision). The length of off_t variables
1969 is now inspected at build time, and an appropriate printing format (%ld
1970 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1971 to be "long long int" or "long int". This is needed for the internal
1972 formatting function string_vformat().
1974 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1975 the configuration file to be ":syslog", then the script "guesses" where
1976 the logs files are, rather than using the compiled in default. In our
1977 case the guess is not the same as the compiled default, so the script
1978 suddenly stopped working when I started to use syslog. The patch checks
1979 to see if log_file_path is "". If so, it attempts to read it from exim
1980 with no configuration file to get the compiled in version, before it
1981 falls back to the previous guessing code."
1983 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1984 implementing BATV in an Exim configuration. See NewStuff for the gory
1987 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1988 Makefile that are specific to HP-UX.
1990 PH/18 If the "use_postmaster" option was set for a recipient callout together
1991 with the "random" option, the postmaster address was used as the MAIL
1992 FROM address for the random test, but not for the subsequent recipient
1993 test. It is now used for both.
1995 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1996 patch removes a few documentation additions to RFC 3028, because the
1997 latest draft now contains them. It adds the new en;ascii-case comparator
1998 and a new error check for 8bit text in MIME parts. Comparator and
1999 require names are now matched exactly. I enabled the subaddress
2000 extension, but it is not well tested yet (read: it works for me)."
2002 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2003 rework some of the code of TK/09 above to avoid the hardwired use of
2004 "%lld" and "long long". Replaced the call to snprintf() with a call to
2007 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2008 records point to non-existent hosts", "retry timeout exceeded", and
2009 "retry time not reached for any host after a long failure period".
2011 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2012 experimental DomainKeys support:
2014 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2015 (2) On an error such as an illegally used "control", the wrong name for
2016 the control was given.
2018 These problems did NOT occur unless DomainKeys support was compiled.
2020 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2022 PH/24 Added ${if match_ip condition.
2024 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2025 databases so that it will be absolutely obvious if a crash occurs in the
2026 DB library. This is a regular occurrence (often caused by mis-matched
2029 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2030 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2031 picked up on a user's system that detects such things. There doesn't seem
2032 to be a gcc warning option for this - only an attribute that has to be
2033 put on the function's prototype. It seems that in Fedora Core 4 they have
2034 set this on a number of new functions. No doubt there will be more in due
2037 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2038 specify a transport. However, if an address that was verified by such a
2039 router was the subject of a callout, Exim crashed because it tried to
2040 read the rcpt_include_affixes from the non-existent transport. Now it
2041 just assumes that the setting of that option is false. This bug was
2042 introduced by 4.51/PH/31.
2044 PH/28 Changed -d+all to exclude +memory, because that information is very
2045 rarely of interest, but it makes the output a lot bigger. People tend to
2046 do -d+all out of habit.
2048 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2049 code in os-type was giving problems when libc.so lives in lib64, like on
2052 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2053 aren't the modern standard, and it seems that some systems' include files
2054 don't always have them. Exim was already checking for some of the newer
2055 ones like T_AAAA, and defining it itself. I've added checks for all the
2056 record types that Exim uses.
2058 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2059 not automatically generating a new one, as it is supposed to. This
2060 prevented TLS from working. If the file did exist, but contained invalid
2061 data, a new version was generated, as expected. It was only the case of a
2062 non-existent file that was broken.
2064 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2065 with a change in libdomainkeys > 0.64.
2067 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2068 from DNS. If the selector record carries the flag, it now has
2069 precedence over the domain-wide flag.
2071 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2073 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2074 the use of a port name that isn't defined in /etc/services) occurred, the
2075 message was deferred as in a normal delivery, and thus remained on the
2076 spool, instead of being failed because of the mua_wrapper setting. This
2077 is now fixed, and I tidied up some of the mua_wrapper messages at the
2080 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2081 the messages in a hash of arrays rather than using individual hashes.
2082 This is a bit cleaner and results in dramatic memory savings, albeit
2083 at a slight CPU cost.
2085 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2086 as requested by Marc Sherman.
2088 SC/10 Eximstats: added histograms for user specified patterns as requested
2091 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2093 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2094 fopen() in the content-scanning modules that did not already have it.
2100 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2101 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2103 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2105 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2107 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2108 to test Sieve filters that use "vacation".
2110 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2111 that changes the way the GnuTLS parameters are stored in the cache file.
2112 The new format can be generated externally. For backward compatibility,
2113 if the data in the cache doesn't make sense, Exim assumes it has read an
2114 old-format file, and it generates new data and writes a new file. This
2115 means that you can't go back to an older release without removing the
2118 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2119 work if there are any delivery delays because "one_time" forces the
2120 parent to be marked "delivered", so its unseen clone is never tried
2121 again. For this reason, Exim now forbids the simultaneous setting of
2124 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2125 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2126 entry if you want to know the details. The fix, however, neglected to
2127 consider the case when local delivery batching is involved. The test for
2128 "previously delivered" was not happening when checking to see if an
2129 address could be batched with a previous (undelivered) one; under
2130 certain circumstances this could lead to multiple deliveries to the same
2133 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2134 in its include files, and this causes problems building Exim.
2136 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2137 header_syntax) but Exim was just ignoring anything given after a slash.
2138 In particular, this caused confusion with an attempt to use "verify =
2139 reverse_host_lookup/defer_ok". An error is now given when options are
2140 supplied for verify items that do not have them. (Maybe reverse_host_
2141 lookup should have a defer_ok option, but that's a different point.)
2143 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2144 defined by RFC 821) to 2048, because there were problems with some AUTH
2145 commands, and RFC 1869 says the size should be increased for extended
2146 SMTP commands that take arguments.
2148 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2151 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2152 "unknown" error; now it says that the functionality isn't in the binary.
2154 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2155 an address' error message when a string expansion fails (syntax or
2156 whatever). Otherwise the password may appear in the log. Following change
2157 PH/42 below, there is no longer a chance of it appearing in a bounce
2160 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2162 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2163 of its addresses was ever tried. (Bugzilla bug #2.)
2165 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2166 the result incorrectly in the debug output. (It correctly added a newline
2167 to what was transported.)
2169 TF/01 Added $received_time.
2171 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2172 commented out examples of how to interface to a virus scanner and to
2173 SpamAssassin. Also added commented examples of av_scanner and
2174 spamd_address settings.
2176 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2177 and controls are allowed in which ACLs. There were a couple of minor
2178 errors. Some of the entries in the conditions table (which is a table of
2179 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2180 negation of where the condition IS allowed.
2182 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2184 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2185 header file does not have a version number, so I've had to invent a new
2186 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2187 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2188 radiusclient), but it was contributed by a Radius user.
2190 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2191 files or directories, for OpenSSL.
2193 PH/19 When an Exim process that is running as root has to create an Exim log
2194 file, it does so in a subprocess that runs as exim:exim so as to get the
2195 ownership right at creation (otherwise, other Exim processes might see
2196 the file with the wrong ownership). There was no test for failure of this
2197 fork() call, which would lead to the process getting stuck as it waited
2198 for a non-existent subprocess. Forks do occasionally fail when resources
2199 run out. I reviewed all the other calls to fork(); they all seem to check
2202 PH/20 When checking for unexpected SMTP input at connect time (before writing
2203 the banner), Exim was not dealing correctly with a non-positive return
2204 from the read() function. If the client had disconnected by this time,
2205 the result was a log entry for a synchronization error with an empty
2206 string after "input=" when read() returned zero. If read() returned -1
2207 (an event I could not check), uninitialized data bytes were printed.
2208 There were reports of junk text (parts of files, etc) appearing after
2211 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2213 PH/22 Added support for macro redefinition, and (re)definition in between
2214 driver and ACL definitions.
2216 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2217 forgetting to use the resulting value; it was using the unexpanded value.
2219 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2220 hadn't been configured. The fix is from Juergen Kreileder, who
2221 understands it better than I do:
2223 "Here's what I see happening with three configured cyrus_sasl
2224 authenticators configured (plain, login, cram-md5):
2226 On startup auth_cyrus_sasl_init() gets called for each of these.
2227 This means three calls to sasl_listmech() without a specified mech_list.
2228 => SASL tests which mechs of all available mechs actually work
2229 => three warnings about OTP not working
2230 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2232 With the patch, sasl_listmech() also gets called three times. But now
2233 SASL's mech_list option is set to the server_mech specified in the the
2234 authenticator. Or in other words, the answer from sasl_listmech()
2235 gets limited to just the mech you're testing for (which is different
2237 => the return list contains just 'plain' or 'login', 'cram-md5' or
2238 nothing depending on the value of ob->server_mech.
2240 I've just tested the patch: Authentication still works fine,
2241 unavailable mechs specified in the exim configuration are still
2242 caught, and the auth.log warnings about OTP are gone."
2244 PH/25 When debugging is enabled, the contents of the command line are added
2245 to the debugging output, even when log_selector=+arguments is not
2248 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2249 answer is "GNU", and only if the return is "GNU/something" is the answer
2252 PH/27 $acl_verify_message is now set immediately after the failure of a
2253 verification in an ACL, and so is available in subsequent modifiers. In
2254 particular, the message can be preserved by coding like this:
2256 warn !verify = sender
2257 set acl_m0 = $acl_verify_message
2259 Previously, $acl_verify_message was set only while expanding "message"
2260 and "log_message" when a very denied access.
2262 PH/28 Modified OS/os.c-Linux with
2264 -#ifndef OS_LOAD_AVERAGE
2265 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2267 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2268 nomenclature these days.)
2270 PH/29 Installed patch from the Sieve maintainer that adds the options
2271 sieve_useraddress and sieve_subaddress to the redirect router.
2273 PH/30 In these circumstances:
2274 . Two addresses routed to the same list of hosts;
2275 . First host does not offer TLS;
2276 . First host accepts first address;
2277 . First host gives temporary error to second address;
2278 . Second host offers TLS and a TLS session is established;
2279 . Second host accepts second address.
2280 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2281 and peerdn, if requested) that were in fact used only for the second
2284 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2285 attention to any local part prefix or suffix that was matched by the
2286 router that accepted the address. It now behaves in the same way as it
2287 does for delivery: the affixes are removed from the local part unless
2288 rcpt_include_affixes is set on the transport.
2290 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2291 timeout during the DATA phase of an incoming message.
2293 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2294 applied a patch sanctioned by the Sieve maintainer.
2296 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2297 the uid or gid is negative. A case of a negative gid caused this to be
2298 noticed. The fix allows for either to be negative.
2300 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2301 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2304 PH/36 The change PH/12 above was broken. Fixed it.
2306 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2307 the grounds that routing the same address twice would always produce the
2308 same answer. This might have been true once, but it is certainly no
2309 longer true now. Routing a child address may depend on the previous
2310 routing that produced that child. Some complicated redirection strategies
2311 went wrong when messages had multiple recipients, and made Exim's
2312 behaviour dependent on the order in which the addresses were given.
2314 I have moved the duplicate checking until after the routing is complete.
2315 Exim scans the addresses that are assigned to local and remote
2316 transports, and removes any duplicates. This means that more work will be
2317 done, as duplicates will always all be routed, but duplicates are
2318 presumably rare, so I don't expect this is of any significance.
2320 For deliveries to pipes, files, and autoreplies, the duplicate checking
2321 still happens during the routing process, since they are not going to be
2324 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2325 It corrects a timeout issue with spamd. This is Ian's comment: "The
2326 background is that sometimes spamd either never reads data from a
2327 connection it has accepted, or it never writes response data. The exiscan
2328 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2329 blindly assumes that writes won't block so it may never time out."
2331 PH/39 Allow G after quota size as well as K and M.
2333 PH/40 The value set for $authenticated_id in an authenticator may not contain
2334 binary zeroes or newlines because the value is written to log lines and
2335 to spool files. There was no check on this. Now the value is run through
2336 the string_printing() function so that such characters are converted to
2337 printable escape sequences.
2339 PH/41 $message_linecount is a new variable that contains the total number of
2340 lines in the message. Compare $body_linecount, which is the count for the
2343 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2344 bounce and delay warning messages, except in certain special cases, which
2347 (a) An SMTP error message from a remote host;
2348 (b) A message specified in a :fail: redirection;
2349 (c) A message specified in a "fail" command in a system filter;
2350 (d) A message specified in a FAIL return from the queryprogram router;
2351 (e) A message specified by the cannot_route_message router option.
2353 In these cases only, Exim does include the error details in bounce and
2354 warning messages. There are also a few cases where bland messages such
2355 as "unrouteable address" or "local delivery error" are given.
2357 PH/43 $value is now also set for the "else" part of a ${run expansion.
2359 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2360 being worked on, but at least Exim now implements the latest version to
2363 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2364 process to complete was treated as a delivery failure, a timeout while
2365 writing the message to the pipe was logged, but erroneously treated as a
2366 successful delivery. Such timeouts include transport filter timeouts. For
2367 consistency with the overall process timeout, these timeouts are now
2368 treated as errors, giving rise to delivery failures by default. However,
2369 there is now a new Boolean option for the pipe transport called
2370 timeout_defer, which, if set TRUE, converts the failures into defers for
2371 both kinds of timeout. A transport filter timeout is now identified in
2374 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2375 systems where "make" and "gmake" are different, calling "gmake" at top
2376 level broke things. I've arranged for the value of $(MAKE) to be passed
2377 from the Makefile to this script so that it can call the same version of
2381 A note about Exim versions 4.44 and 4.50
2382 ----------------------------------------
2384 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2385 changes of various kinds. As a consequence, a big documentation update was
2386 needed. This delayed the release for rather longer than seemed good, especially
2387 in the light of a couple of (minor) security issues. Therefore, the changes
2388 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2389 release. So 4.44 and 4.50 are in effect two different branches that both start
2392 I have left the 4.50 change log unchanged; it contains all the changes since
2393 4.43. The change log for 4.44 is below; many of its items are identical to
2394 those for 4.50. This seems to be the most sensible way to preserve the
2395 historical information.
2401 1. Minor wording change to the doc/README.SIEVE file.
2403 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2404 computation of the current number of files was incorrect.
2406 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2407 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2410 4. Give more explanation in the error message when the command for a transport
2411 filter fails to execute.
2413 5. There are several places where Exim runs a non-Exim command in a
2414 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2415 was being done only for the command run by the queryprogram router. It is
2416 now done for all such subprocesses. The other cases are: ${run, transport
2417 filters, and the commands run by the lmtp and pipe transports.
2419 6. Added CONFIGURE_GROUP build-time option.
2421 7. Some older OS have a limit of 256 on the maximum number of file
2422 descriptors. Exim was using setrlimit() to set 1000 as a large value
2423 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2424 systems. I've change it so that if it can't get 1000, it tries for 256.
2426 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2427 was an oversight, and furthermore, ever since the addition of extra
2428 controls (e.g. 4.43/32), the checks on when to allow different forms of
2429 "control" were broken. There should now be diagnostics for all cases when a
2430 control that does not make sense is encountered.
2432 9. Added the /retain_sender option to "control=submission".
2434 10. $recipients is now available in the predata ACL (oversight).
2436 11. Tidy the search cache before the fork to do a delivery from a message
2437 received from the command line. Otherwise the child will trigger a lookup
2438 failure and thereby defer the delivery if it tries to use (for example) a
2439 cached ldap connection that the parent has called unbind on.
2441 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2442 of $address_data from the recipient verification was clobbered by the
2443 sender verification.
2445 13. The value of address_data from a sender verification is now available in
2446 $sender_address_data in subsequent conditions in the ACL statement.
2448 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2450 15. Added a new option "connect=<time>" to callout options, to set a different
2453 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2454 was its contents. (It was OK if the option was not defined at all.)
2456 17. A "Completed" log line is now written for messages that are removed from
2457 the spool by the -Mrm option.
2459 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2460 information about exactly what failed.
2462 19. Added -dd to debug only the daemon process.
2464 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2465 handles timeouts, both on the server side and network timeouts. Renamed the
2466 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2468 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2469 It is now set to "smtps".
2471 22. $host_address is now set to the target address during the checking of
2472 ignore_target_hosts.
2474 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2475 being passed; this would have caused $sender_host_name to have been used if
2476 matching the list had actually called for a host name (not very likely,
2477 since this list is usually IP addresses). A host name is now passed as
2480 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2481 code that specifies a non-restarting handler (typically sigaction() in
2482 modern systems) in an attempt to fix a rare and obscure crash bug.
2484 25. Narrowed the window for a race in the daemon that could cause it to ignore
2485 SIGCHLD signals. This is not a major problem, because they are used only to
2486 wake it up if nothing else does.
2488 26. A malformed maildirsize file could cause Exim to calculate negative values
2489 for the mailbox size or file count. Odd effects could occur as a result.
2490 The maildirsize information is now recalculated if the size or filecount
2493 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2494 support for a long time. Removed HAVE_SYS_VFS_H.
2496 28. Installed the latest version of exipick from John Jetmore.
2498 29. In an address list, if the pattern was not a regular expression, an empty
2499 subject address (from a bounce message) matched only if the pattern was an
2500 empty string. Non-empty patterns were not even tested. This was the wrong
2501 because it is perfectly reasonable to use an empty address as part of a
2502 database query. An empty address is now tested by patterns that are
2503 lookups. However, all the other forms of pattern expect the subject to
2504 contain a local part and a domain, and therefore, for them, an empty
2505 address still always fails if the pattern is not itself empty.
2507 30. Exim went into a mad DNS loop when attempting to do a callout where the
2508 host was specified on an smtp transport, and looking it up yielded more
2509 than one IP address.
2511 31. Re-factored the code for checking spool and log partition space into a
2512 function that finds that data and another that does the check. The former
2513 is then used to implement four new variables: $spool_space, $log_space,
2514 $spool_inodes, and $log_inodes.
2516 32. The RFC2047 encoding function was originally intended for short strings
2517 such as real names; it was not keeping to the 75-character limit for
2518 encoded words that the RFC imposes. It now respects the limit, and
2519 generates multiple encoded words if necessary. To be on the safe side, I
2520 have increased the buffer size for the ${rfc2047: expansion operator from
2523 33. It is now permitted to omit both strings after an "if" condition; if the
2524 condition is true, the result is "true". As before, when the second string
2525 is omitted, a false condition yields an empty string. This makes it less
2526 cumbersome to write custom ACL and router conditions.
2528 34. Failure to deliver a bounce message always caused it to be frozen, even if
2529 there was an errors_to setting on the router. The errors_to setting is now
2532 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2533 canonical form (fully expanded) before being placed in
2534 $sender_host_address.
2536 36. The table in the code that translates DNS record types into text (T_A to
2537 "A" for instance) was missing entries for NS and CNAME. It is just possible
2538 that this could have caused confusion if both these types were looked up
2539 for the same domain, because the text type is used as part of Exim's
2540 per-process caching. But the chance of anyone hitting this buglet seems
2543 37. The dnsdb lookup has been extended in a number of ways.
2545 (1) There is a new type, "zns", which walks up the domain tree until it
2546 finds some nameserver records. It should be used with care.
2548 (2) There is a new type, "mxh", which is like "mx" except that it returns
2549 just the host names, not the priorities.
2551 (3) It is now possible to give a list of domains (or IP addresses) to be
2552 looked up. The behaviour when one of the lookups defers can be
2553 controlled by a keyword.
2555 (4) It is now possible to specify the separator character for use when
2556 multiple records are returned.
2558 38. The dnslists ACL condition has been extended: it is now possible to supply
2559 a list of IP addresses and/or domains to be looked up in a particular DNS
2562 39. Added log_selector=+queue_time_overall.
2564 40. When running the queue in the test harness, wait just a tad after forking a
2565 delivery process, to get repeatability of debugging output.
2567 41. Include certificate and key file names in error message when GnuTLS fails
2568 to set them up, because the GnuTLS error message doesn't include the name
2569 of the failing file when there is a problem reading it.
2571 42. Allow both -bf and -bF in the same test run.
2573 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2575 44. The "Exiscan patch" is now merged into the mainline Exim source.
2577 45. Sometimes the final signoff response after QUIT could fail to get
2578 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2579 < 0 before doing a fflush(). This bug looks as though it goes back to the
2580 introduction of TLS in release 3.20, but "sometimes" must have been rare
2581 because the tests only now provoked it.
2583 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2584 (this can affect the format of dates).
2586 47. exim_tidydb, when checking for the continued existence of a message for
2587 which it has found a message-specific retry record, was not finding
2588 messages that were in split spool directories. Consequently, it was
2589 deleting retry records that should have stayed in existence.
2591 48. Steve fixed some bugs in eximstats.
2593 49. The SPA authentication driver was not abandoning authentication and moving
2594 on to the next authenticator when an expansion was forced to fail,
2595 contradicting the general specification for all authenticators. Instead it
2596 was generating a temporary error. It now behaves as specified.
2598 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2599 (the order specifies the preference for clients). The order is now AES256,
2600 AES128, 3DES, ARCFOUR128.
2602 51. Small patch to Sieve code - explicitly set From: when generating an
2605 52. Exim crashed if a remote delivery caused a very long error message to be
2606 recorded - for instance if somebody sent an entire SpamAssassin report back
2607 as a large number of 550 error lines. This bug was coincidentally fixed by
2608 increasing the size of one of Exim's internal buffers (big_buffer) that
2609 happened as part of the Exiscan merge. However, to be on the safe side, I
2610 have made the code more robust (and fixed the comments that describe what
2613 53. Now that there can be additional text after "Completed" in log lines (if
2614 the queue_time_overall log selector is set), a one-byte patch to exigrep
2615 was needed to allow it to recognize "Completed" as not the last thing in
2618 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2619 patch that reportedly fixes this has been added. I am not expert enough to
2620 create a test for it. This is what the patch creator wrote:
2622 "I found a little strange behaviour of ldap code when working with
2623 Windows 2003 AD Domain, where users was placed in more than one
2624 Organization Units. When I tried to give exim partial DN, the exit code
2625 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2626 But simultaneously result of request was absolutely normal ldap result,
2627 so I produce this patch..."
2629 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2630 REFERENCE, so I have modified the code to exclude the patch when that macro
2633 55. Some experimental protocols are using DNS PTR records for new purposes. The
2634 keys for these records are domain names, not reversed IP addresses. The
2635 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2636 leaves it alone. Component reversal etc. now happens only for IP addresses.
2639 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2641 57. Double the size of the debug message buffer (to 2048) so that more of very
2642 long debug lines gets shown.
2644 58. The exicyclog utility now does better if the number of log files to keep
2645 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2647 59. Two changes related to the smtp_active_hostname option:
2649 (1) $smtp_active_hostname is now available as a variable.
2650 (2) The default for smtp_banner uses $smtp_active_hostname instead
2651 of $primary_hostname.
2653 60. The host_aton() function is supposed to be passed a string that is known
2654 to be a valid IP address. However, in the case of IPv6 addresses, it was
2655 not checking this. This is a hostage to fortune. Exim now panics and dies
2656 if the condition is not met. A case was found where this could be provoked
2657 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2658 components; fortuitously, this particular loophole had already been fixed
2659 by change 4.50/55 above.
2661 If there are any other similar loopholes, the new check in host_aton()
2662 itself should stop them being exploited. The report I received stated that
2663 data on the command line could provoke the exploit when Exim was running as
2664 exim, but did not say which command line option was involved. All I could
2665 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2666 running as the user.
2669 61. There was a buffer overflow vulnerability in the SPA authentication code
2670 (which came originally from the Samba project). I have added a test to the
2671 spa_base64_to_bits() function which I hope fixes it.
2674 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2675 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2677 63. The daemon start-up calls getloadavg() while still root for those OS that
2678 need the first call to be done as root, but it missed one case: when
2679 deliver_queue_load_max is set with deliver_drop_privilege. This is
2680 necessary for the benefit of the queue runner, because there is no re-exec
2681 when deliver_drop_privilege is set.
2683 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2684 This has been fixed.
2686 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2687 was in use, was not putting the data itself into the right store pool;
2688 consequently, it could be overwritten for a subsequent message in the same
2689 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2692 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2694 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2695 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2696 the function were treating the return as a boolean value, which happened to
2697 work because 0=false and not-0=true, but is not correct code.
2699 68. The host_aton() function was not handling scoped IPv6 addresses (those
2700 with, for example, "%eth0" on the end) correctly.
2702 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2703 negated items (that is, ~something) in unsigned ints. Some compilers
2704 apparently mutter when there is no cast.
2706 70. If an address verification called from an ACL failed, and did not produce a
2707 user-specific message (i.e. there was only a "system" message), nothing was
2708 put in $acl_verify_message. In this situation, it now puts the system
2711 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2712 session; change 4.31/43 added the unwanted input to the log line - except
2713 that it did not do this in the start of session case. It now does.
2715 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2716 This is wrong and can cause the other end to generate a synchronization
2717 error if it is another Exim or anything else that does the synchronization
2718 check. A QUIT command is no longer sent after a timeout.
2720 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2721 during host lookups.
2723 74. The defer_ok option of callout verification was not working if it was used
2724 when verifying addresses in header lines, that is, for this case:
2726 verify = header_sender/callout=defer_ok
2728 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2729 those file descriptors could be used for SMTP connections. If anything
2730 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2731 could be sent to the SMTP client, causing chaos. The daemon now opens
2732 stdin, stdout, and stderr to /dev/null when it puts itself into the
2735 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2736 log by default. The user can override this with suitable Perl magic.
2738 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2739 the log message when discard triggers, was not working for the DATA ACL or
2740 for the non-SMTP ACL.
2742 78. Error message wording change in sieve.c.
2744 79. If smtp_accept_max_per_host was set, the number of connections could be
2745 restricted to fewer than expected, because the daemon was trying to set up
2746 a new connection before checking whether the processes handling previous
2747 connections had finished. The check for completed processes is now done
2748 earlier. On busy systems, this bug wouldn't be noticed because something
2749 else would have woken the daemon, and it would have reaped the completed
2752 80. If a message was submitted locally by a user whose login name contained one
2753 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2754 It caused a spool format error. I have fixed the spool reading code. A
2755 related problem was that the "from" clause in the Received: line became
2756 illegal because of the space(s). It is now covered by ${quote_local_part.
2758 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2761 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2763 83. Patch from Sieve maintainer to fix unterminated string problem in
2764 "vacation" handling.
2766 84. Some minor changes to the Linux configuration files to help with other
2767 OS variants using glibc.
2769 85. One more patch for Sieve to update vacation handling to latest spec.
2772 ----------------------------------------------------
2773 See the note above about the 4.44 and 4.50 releases.
2774 ----------------------------------------------------
2780 1. Change 4.43/35 introduced a bug that caused file counts to be
2781 incorrectly computed when quota_filecount was set in an appendfile
2784 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2785 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2788 3. Give more explanation in the error message when the command for a transport
2789 filter fails to execute.
2791 4. There are several places where Exim runs a non-Exim command in a
2792 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2793 was being done only for the command run by the queryprogram router. It is
2794 now done for all such subprocesses. The other cases are: ${run, transport
2795 filters, and the commands run by the lmtp and pipe transports.
2797 5. Some older OS have a limit of 256 on the maximum number of file
2798 descriptors. Exim was using setrlimit() to set 1000 as a large value
2799 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2800 systems. I've change it so that if it can't get 1000, it tries for 256.
2802 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2803 was an oversight, and furthermore, ever since the addition of extra
2804 controls (e.g. 4.43/32), the checks on when to allow different forms of
2805 "control" were broken. There should now be diagnostics for all cases when a
2806 control that does not make sense is encountered.
2808 7. $recipients is now available in the predata ACL (oversight).
2810 8. Tidy the search cache before the fork to do a delivery from a message
2811 received from the command line. Otherwise the child will trigger a lookup
2812 failure and thereby defer the delivery if it tries to use (for example) a
2813 cached ldap connection that the parent has called unbind on.
2815 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2816 of $address_data from the recipient verification was clobbered by the
2817 sender verification.
2819 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2820 was its contents. (It was OK if the option was not defined at all.)
2822 11. A "Completed" log line is now written for messages that are removed from
2823 the spool by the -Mrm option.
2825 12. $host_address is now set to the target address during the checking of
2826 ignore_target_hosts.
2828 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2829 being passed; this would have caused $sender_host_name to have been used if
2830 matching the list had actually called for a host name (not very likely,
2831 since this list is usually IP addresses). A host name is now passed as
2834 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2835 code that specifies a non-restarting handler (typically sigaction() in
2836 modern systems) in an attempt to fix a rare and obscure crash bug.
2838 15. Narrowed the window for a race in the daemon that could cause it to ignore
2839 SIGCHLD signals. This is not a major problem, because they are used only to
2840 wake it up if nothing else does.
2842 16. A malformed maildirsize file could cause Exim to calculate negative values
2843 for the mailbox size or file count. Odd effects could occur as a result.
2844 The maildirsize information is now recalculated if the size or filecount
2847 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2848 support for a long time. Removed HAVE_SYS_VFS_H.
2850 18. Updated exipick to current release from John Jetmore.
2852 19. Allow an empty sender to be matched against a lookup in an address list.
2853 Previously the only cases considered were a regular expression, or an
2856 20. Exim went into a mad DNS lookup loop when doing a callout where the
2857 host was specified on the transport, if the DNS lookup yielded more than
2860 21. The RFC2047 encoding function was originally intended for short strings
2861 such as real names; it was not keeping to the 75-character limit for
2862 encoded words that the RFC imposes. It now respects the limit, and
2863 generates multiple encoded words if necessary. To be on the safe side, I
2864 have increased the buffer size for the ${rfc2047: expansion operator from
2867 22. Failure to deliver a bounce message always caused it to be frozen, even if
2868 there was an errors_to setting on the router. The errors_to setting is now
2871 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2872 canonical form (fully expanded) before being placed in
2873 $sender_host_address.
2875 24. Updated eximstats to version 1.33
2877 25. Include certificate and key file names in error message when GnuTLS fails
2878 to set them up, because the GnuTLS error message doesn't include the name
2879 of the failing file when there is a problem reading it.
2881 26. Expand error message when OpenSSL has problems setting up cert/key files.
2884 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2885 (this can affect the format of dates).
2887 28. exim_tidydb, when checking for the continued existence of a message for
2888 which it has found a message-specific retry record, was not finding
2889 messages that were in split spool directories. Consequently, it was
2890 deleting retry records that should have stayed in existence.
2892 29. eximstats updated to version 1.35
2893 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2894 1.35 - bugfix such that pie charts by volume are generated correctly
2896 30. The SPA authentication driver was not abandoning authentication and moving
2897 on to the next authenticator when an expansion was forced to fail,
2898 contradicting the general specification for all authenticators. Instead it
2899 was generating a temporary error. It now behaves as specified.
2901 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2902 (the order specifies the preference for clients). The order is now AES256,
2903 AES128, 3DES, ARCFOUR128.
2905 31. Small patch to Sieve code - explicitly set From: when generating an
2908 32. Exim crashed if a remote delivery caused a very long error message to be
2909 recorded - for instance if somebody sent an entire SpamAssassin report back
2910 as a large number of 550 error lines. This bug was coincidentally fixed by
2911 increasing the size of one of Exim's internal buffers (big_buffer) that
2912 happened as part of the Exiscan merge. However, to be on the safe side, I
2913 have made the code more robust (and fixed the comments that describe what
2916 33. Some experimental protocols are using DNS PTR records for new purposes. The
2917 keys for these records are domain names, not reversed IP addresses. The
2918 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2919 leaves it alone. Component reversal etc. now happens only for IP addresses.
2922 34. The host_aton() function is supposed to be passed a string that is known
2923 to be a valid IP address. However, in the case of IPv6 addresses, it was
2924 not checking this. This is a hostage to fortune. Exim now panics and dies
2925 if the condition is not met. A case was found where this could be provoked
2926 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2927 components; fortuitously, this particular loophole had already been fixed
2928 by change 4.50/55 or 4.44/33 above.
2930 If there are any other similar loopholes, the new check in host_aton()
2931 itself should stop them being exploited. The report I received stated that
2932 data on the command line could provoke the exploit when Exim was running as
2933 exim, but did not say which command line option was involved. All I could
2934 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2935 running as the user.
2938 35. There was a buffer overflow vulnerability in the SPA authentication code
2939 (which came originally from the Samba project). I have added a test to the
2940 spa_base64_to_bits() function which I hope fixes it.
2943 36. The daemon start-up calls getloadavg() while still root for those OS that
2944 need the first call to be done as root, but it missed one case: when
2945 deliver_queue_load_max is set with deliver_drop_privilege. This is
2946 necessary for the benefit of the queue runner, because there is no re-exec
2947 when deliver_drop_privilege is set.
2949 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2950 was in use, was not putting the data itself into the right store pool;
2951 consequently, it could be overwritten for a subsequent message in the same
2952 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2955 38. Sometimes the final signoff response after QUIT could fail to get
2956 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2957 < 0 before doing a fflush(). This bug looks as though it goes back to the
2958 introduction of TLS in release 3.20, but "sometimes" must have been rare
2959 because the tests only now provoked it.
2965 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2966 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2967 Now it returns an integer. A number of calls were still expecting a T/F
2968 return. Fortuitously, in all cases, the tests worked in OK situations,
2969 which is the norm. However, things would have gone wrong on any write
2970 failures on the smtp file descriptor. This function is used when sending
2971 messages over SMTP and also when doing verify callouts.
2973 2. When Exim is called to do synchronous delivery of a locally submitted
2974 message (the -odf or -odi options), it no longer closes stderr before doing
2977 3. Implemented the mua_wrapper option.
2979 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2981 5. Implemented the functions header_remove(), header_testname(),
2982 header_add_at_position(), and receive_remove_recipient(), and exported them
2985 6. If an ACL "warn" statement specified the addition of headers, Exim already
2986 inserted X-ACL-Warn: at the start if there was no header name. However, it
2987 was not making this test for the second and subsequent header lines if
2988 there were newlines in the string. This meant that an invalid header could
2989 be inserted if Exim was badly configured.
2991 7. Allow an ACL "warn" statement to add header lines at the start or after all
2992 the Received: headers, as well as at the end.
2994 8. Added the rcpt_4xx retry error code.
2996 9. Added postmaster_mailfrom=xxx to callout verification option.
2998 10. Added mailfrom=xxxx to the callout verification option, for verify=
3001 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3002 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3004 12. Inserted some casts to stop certain compilers warning when using pointer
3005 differences as field lengths or precisions in printf-type calls (mostly
3006 affecting debugging statements).
3008 13. Added optional readline() support for -be (dynamically loaded).
3010 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3011 same clock tick as a message's arrival, so that its received time was the
3012 same as the "first fail" time on the retry record, and that message
3013 remained on the queue past the ultimate address timeout, every queue runner
3014 would try a delivery (because it was past the ultimate address timeout) but
3015 after another failure, the ultimate address timeout, which should have then
3016 bounced the address, did not kick in. This was a "< instead of <=" error;
3017 in most cases the first failure would have been in the next clock tick
3018 after the received time, and all would be well.
3020 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3021 being recognized when the domain list was tested by the match_domain
3022 condition in an expansion string.
3024 16. Added the ${str2b64: operator.
3026 17. Exim was always calling setrlimit() to set a large limit for the number of
3027 processes, without checking whether the existing limit was already
3028 adequate. (It did check for the limit on file descriptors.) Furthermore,
3029 errors from getrlimit() and setrlimit() were being ignored. Now they are
3030 logged to the main and panic logs, but Exim does carry on, to try to do its
3031 job under whatever limits there are.
3033 18. Imported PCRE 5.0.
3035 19. Trivial typo in log message " temporarily refused connection" (the leading
3038 20. If the log selector return_path_on_delivery was set and an address was
3039 redirected to /dev/null, the delivery process crashed because it assumed
3040 that a return path would always be set for a "successful" delivery. In this
3041 case, the whole delivery is bypassed as an optimization, and therefore no
3044 21. Internal re-arrangement: the function for sending a challenge and reading
3045 a response while authentication was assuming a zero-terminated challenge
3046 string. It's now changed to take a pointer and a length, to allow for
3047 binary data in such strings.
3049 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3051 23. Exim was not respecting finduser_retries when seeking the login of the
3052 uid under which it was called; it was always trying 10 times. (The default
3053 setting of finduser_retries is zero.) Also, it was sleeping after the final
3054 failure, which is pointless.
3056 24. Implemented tls_on_connect_ports.
3058 25. Implemented acl_smtp_predata.
3060 26. If the domain in control=submission is set empty, Exim assumes that the
3061 authenticated id is a complete email address when it generates From: or
3062 Sender: header lines.
3064 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3065 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3066 chown and chgrp in /bin and hostname in /usr/bin.
3068 28. Exim was keeping the "process log" file open after each use, just as it
3069 does for the main log. This opens the possibility of it remaining open for
3070 long periods when the USR1 signal hits a daemon. Occasional processlog
3071 errors were reported, that could have been caused by this. Anyway, it seems
3072 much more sensible not to leave this file open at all, so that is what now
3075 29. The long-running daemon process does not normally write to the log once it
3076 has entered its main loop, and it closes the log before doing so. This is
3077 so that log files can straightforwardly be renamed and moved. However,
3078 there are a couple of unusual error situations where the daemon does write
3079 log entries, and I had neglected to close the log afterwards.
3081 30. The text of an SMTP error response that was received during a remote
3082 delivery was being truncated at 512 bytes. This is too short for some of
3083 the long messages that one sometimes sees. I've increased the limit to
3086 31. It is now possible to make retry rules that apply only when a message has a
3087 specific sender, in particular, an empty sender.
3089 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3090 it possible to be selective about when SMTP synchronization is enforced.
3092 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3094 32. Implemented hosts_connection_nolog.
3096 33. Added an ACL for QUIT.
3098 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3101 35. Added mailbox_size and mailbox_filecount to appendfile.
3103 36. Added control = no_multiline_responses to ACLs.
3105 37. There was a bug in the logic of the code that waits for the clock to tick
3106 in the case where the clock went backwards by a substantial amount such
3107 that the microsecond fraction of "now" was more than the microsecond
3108 fraction of "then" (but the whole seconds number was less).
3110 38. Added support for the libradius Radius client library this is found on
3111 FreeBSD (previously only the radiusclient library was supported).
3117 1. When certain lookups returned multiple values in the form name=value, the
3118 quoting of the values was not always being done properly. Specifically:
3119 (a) If the value started with a double quote, but contained no whitespace,
3121 (b) If the value contained whitespace other than a space character (i.e.
3122 tabs or newlines or carriage returns) it was not quoted.
3123 This fix has been applied to the mysql and pgsql lookups by writing a
3124 separate quoting function and calling it from the lookup code. The fix
3125 should probably also be applied to nisplus, ibase and oracle lookups, but
3126 since I cannot test any of those, I have not disturbed their existing code.
3128 2. A hit in the callout cache for a specific address caused a log line with no
3129 reason for rejecting RCPT. Now it says "Previous (cached) callout
3130 verification failure".
3132 3. There was an off-by-one bug in the queryprogram router. An over-long
3133 return line was truncated at 256 instead of 255 characters, thereby
3134 overflowing its buffer with the terminating zero. As well as fixing this, I
3135 have increased the buffer size to 1024 (and made a note to document this).
3137 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3138 when Exim is waiting for an SMTP response from a remote server, Exim
3139 restarts its select() call on the socket, thereby resetting its timeout.
3140 This is not a problem when such interrupts are rare. Somebody set up a cron
3141 job to run exiwhat every 2 minutes, which is less than the normal select()
3142 timeout (5 or 10 minutes). This meant that the select() timeout never
3143 kicked in because it was always reset. I have fixed this by comparing the
3144 time when an interrupt arrives with the time at the start of the first call
3145 to select(). If more time than the timeout has elapsed, the interrupt is
3146 treated as a timeout.
3148 5. Some internal re-factoring in preparation for the addition of Sieve
3149 extensions (by MH). In particular, the "personal" test is moved to a
3150 separate function, and given an option for scanning Cc: and Bcc: (which is
3151 not set for Exim filters).
3153 6. When Exim created an email address using the login of the caller as the
3154 local part (e.g. when creating a From: or Sender: header line), it was not
3155 quoting the local part when it contained special characters such as @.
3157 7. Installed new OpenBSD configuration files.
3159 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3160 try to make them clearer.
3162 9. Callout options, other than the timeout value, were being ignored when
3163 verifying sender addresses in header lines. For example, when using
3165 verify = header_sender/callout=no_cache
3167 the cache was (incorrectly) being used.
3169 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3170 only the Cygwin environment.
3172 11. When return_path_on_delivery was set as a log selector, if different remote
3173 addresses in the same message used different return paths and parallel
3174 remote delivery occurred, the wrong values would sometimes be logged.
3175 (Whenever a remote delivery process finished, the return path value from
3176 the most recently started remote delivery process was logged.)
3178 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3179 lines when AUTH and/or TLS are in use. This is the "received protocol"
3180 field. Exim used to use "asmtp" for authenticated SMTP, without any
3181 indication (in the protocol name) for TLS use. Now it follows the RFC and
3182 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3183 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3184 names appear in log lines as well as in Received: header lines.
3186 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3187 extensions, and comparison tests, and to fix some bugs.
3189 14. Changes to the "personal" filter test:
3191 (1) The test was buggy in that it was just doing the equivalent of
3192 "contains" tests on header lines. For example, if a user's address was
3193 anne@some.where, the "personal" test would incorrectly be true for
3195 To: susanne@some.where
3197 This test is now done by extracting each address from the header in turn,
3198 and checking the entire address. Other tests that are part of "personal"
3199 are now done using regular expressions (for example, to check local parts
3200 of addresses in From: header lines).
3202 (2) The list of non-personal local parts in From: addresses has been
3203 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3204 taken from the Sieve specification recommendations.
3206 (3) If the message contains any header line starting with "List-" it is
3207 treated as non-personal.
3209 (4) The test for "circular" in the Subject: header line has been removed
3210 because it now seems ill-conceived.
3212 15. Minor typos in src/EDITME comments corrected.
3214 16. Installed latest exipick from John Jetmore.
3216 17. If headers_add on a router specified a text string that was too long for
3217 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3218 of string_sprintf() is now avoided.
3220 18. $message_body_size was not set (it was always zero) when running the DATA
3221 ACL and the local_scan() function.
3223 19. For the "mail" command in an Exim filter, no default was being set for
3224 the once_repeat time, causing a random time value to be used if "once" was
3225 specified. (If the value happened to be <= 0, no repeat happened.) The
3226 default is now 0s, meaning "never repeat". The "vacation" command was OK
3227 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3228 (I found it when inspecting the code).
3230 20. There is now an overall timeout for performing a callout verification. It
3231 defaults to 4 times the callout timeout, which applies to individual SMTP
3232 commands during the callout. The overall timeout applies when there is more
3233 than one host that can be tried. The timeout is checked before trying the
3234 next host. This prevents very long delays if there are a large number of
3235 hosts and all are timing out (e.g. when the network connections are timing
3236 out). The value of the overall timeout can be changed by specifying an
3237 additional sub-option for "callout", called "maxwait". For example:
3239 verify = sender/callout=5s,maxwait=20s
3241 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3242 to the end before writing, but this should make it even safer).
3244 22. Exim was forgetting that it had advertised PIPELINING for the second and
3245 subsequent messages on an SMTP connection. It was also not resetting its
3246 memory on STARTTLS and an internal HELO.
3248 23. When Exim logs an SMTP synchronization error within a session, it now
3249 records whether PIPELINING has been advertised or not.
3251 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3252 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3253 rather than long int.
3255 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3257 26. Added the never_mail option to autoreply.
3263 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3264 crash if the getsockname() call failed; this can happen if a connection is
3265 closed very soon after it is established. The problem was simply in the
3266 order in which certain operations were done, causing Exim to try to write
3267 to the SMTP stream before it had set up the file descriptor. The bug has
3268 been fixed by making things happen in the correct order.
3274 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3275 before the connection was closed, thus losing the rejection response.
3277 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3278 some early Solaris releases, but causes trouble in current releases where
3279 socklen_t is defined.
3281 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3284 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3287 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3288 that releasing the top of it at the end releases what was used for sub-
3289 expansions (unless the block got too big). However, discard this block if
3290 the first thing is a variable or header, so that we can use its block when
3291 it is dynamic (useful for very large $message_headers, for example).
3293 6. Lookups now cache *every* query, not just the most recent. A new, separate
3294 store pool is used for this. It can be recovered when all lookup caches are
3295 flushed. Lookups now release memory at the end of their result strings.
3296 This has involved some general refactoring of the lookup sources.
3298 7. Some code has been added to the store_xxx() functions to reduce the amount
3299 of flapping under certain conditions.
3301 8. log_incoming_interface used to affect only the <= reception log lines. Now
3302 it causes the local interface and port to be added to several more SMTP log
3303 lines, for example "SMTP connection from", and rejection lines.
3305 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3307 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3309 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3310 could be overwritten at the end of the current message (or the start of a
3311 new message if it was set in a HELO ACL). The value is now preserved for
3312 the duration of the SMTP connection.
3314 12. If a transport had a headers_rewrite setting, and a matching header line
3315 contained an unqualified address, that address was qualified, even if it
3316 did not match any rewriting rules. The underlying bug was that the values
3317 of the flags that permit the existence of unqualified sender and recipient
3318 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3319 non-local messages, and by -bnq for local messages) were not being
3320 preserved with the message after it was received.
3322 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3323 "next input=" as part of the text comprising the host identity instead of
3324 the correct text. The code was using the same buffer for two different
3325 strings. However, depending on which order the printing function evaluated
3326 its arguments, the bug did not always show up. Under Linux, for example, my
3327 test suite worked just fine.
3329 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3330 doesn't work with some older versions of Perl. It has been changed to "my",
3331 which in any case is probably the better facility to use.
3333 15. A really picky compiler found some instances of statements for creating
3334 error messages that either had too many or two few arguments for the format
3337 16. The size of the buffer for calls to the DNS resolver has been increased
3338 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3339 for addresses that have a lot of PTR records. This alleviates a problem; it
3340 does not fully solve it.
3342 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3343 buffer now truncates the list and logs the incident, which is the same
3344 action as happens when Exim is looking up a host name and its aliases.
3345 Previously in this situation something unpredictable would happen;
3346 sometimes it was "internal error: store_reset failed".
3348 18. If a server dropped the connection unexpectedly when an Exim client was
3349 using GnuTLS and trying to read a response, the client delivery process
3350 crashed while trying to generate an error log message.
3352 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3353 string, for example:
3355 warn message = H1: something\nH2: something
3357 the text was added as a single header line from Exim's point of view
3358 though it ended up OK in the delivered message. However, searching for the
3359 second and subsequent header lines using $h_h2: did not work. This has been
3360 fixed. Similarly, if a system filter added multiple headers in this way,
3361 the routers could not see them.
3363 20. Expanded the error message when iplsearch is called with an invalid key to
3364 suggest using net-iplsearch in a host list.
3366 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3367 ACLs are no longer actually imposed (and a message to that effect is
3370 22. If a "gecos" field in a passwd entry contained escaped characters, in
3371 particular, if it contained a \" sequence, Exim got it wrong when building
3372 a From: or a Sender: header from that name. A second bug also caused
3373 incorrect handling when an unquoted " was present following a character
3374 that needed quoting.
3376 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3377 was not being matched caselessly.
3379 24. Arranged for all hyphens in the exim.8 source to be escaped with
3382 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3383 properly. Recipient callout cache records were still being keyed to include
3384 the sender, even when use_sender was set false. This led to far more
3385 callouts that were necessary. The sender is no longer included in the key
3386 when use_sender is false.
3388 26. Added "control = submission" modifier to ACLs.
3390 27. Added the ${base62d: operator to decode base 62 numbers.
3392 28. dnsdb lookups can now access SRV records.
3394 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3395 the configuration file.
3397 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3398 (-v) mode. This makes the output for a verbose queue run more intelligible.
3400 31. Added a use_postmaster feature to recipient callouts.
3402 32. Added the $body_zerocount variable, containing the number of binary zero
3403 bytes in the message body.
3405 33. The time of last modification of the "new" subdirectory is now used as the
3406 "mailbox time last read" when there is a quota error for a maildir
3409 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3411 35. Added +ignore_unknown as a special item in host lists.
3413 36. Code for decoding IPv6 addresses in host lists is now included, even if
3414 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3415 address was recognized as an IP address, but was then not correctly decoded
3416 into binary, causing unexpected and incorrect effects when compared with
3423 1. Very minor rewording of debugging text in manualroute to say "list of
3424 hosts" instead of "hostlist".
3426 2. If verify=header_syntax was set, and a header line with an unqualified
3427 address (no domain) and a large number of spaces between the end of the
3428 name and the colon was received, the reception process suffered a buffer
3429 overflow, and (when I tested it) crashed. This was caused by some obsolete
3430 code that should have been removed. The fix is to remove it!
3432 3. When running in the test harness, delay a bit after writing a bounce
3433 message to get a bit more predictability in the log output.
3435 4. Added a call to search_tidyup() just before forking a reception process. In
3436 theory, someone could use a lookup in the expansion of smtp_accept_max_
3437 per_host which, without the tidyup, could leave open a database connection.
3439 5. Added the variables $recipient_data and $sender_data which get set from a
3440 lookup success in an ACL "recipients" or "senders" condition, or a router
3441 "senders" option, similar to $domain_data and $local_part_data.
3443 6. Moved the writing of debug_print from before to after the "senders" test
3446 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3447 problems for message scanning, either using a data ACL, or using
3448 local_scan() because the Received: header was not generated till after they
3449 were called (in order to set the time as the time of reception completion).
3450 I have revised the way this works. The header is now generated after the
3451 body is received, but before the ACL or local_scan() are called. After they
3452 are run, the timestamp in the header is updated.
3458 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3459 before starting a queue runner without re-exec. This happened only when
3460 deliver_drop_privilege was set or when the Exim user was set to root. The
3461 effect of the bug was that timeouts during subsequent deliveries caused
3462 crashes instead of being properly handled. The handler is now left at its
3463 default (and expected) setting.
3465 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3466 message, again when deliver_drop_privilege is set or Exim is run as root.
3467 The bug described in (1) was not present in this case, but the tidying up
3468 of the other signals was missing. I have made the two cases consistent.
3470 3. The ignore_target_hosts setting on a manualroute router was being ignored
3471 for hosts that were looked up using the /MX notation.
3473 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3476 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3477 operated on the sender address. After changing the $sender_address to <>
3478 for the sender address verify, Exim was re-instated it as the original
3479 (before rewriting) address, but remembering that it had rewritten it, so it
3480 wasn't rewriting it again. This bug also had the effect of breaking the
3481 sender address verification caching when the sender address was rewritten.
3483 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3484 This has been changed so that if the ip literal address matches
3485 ignore_target_hosts, the router declines.
3487 7. Added expansion conditions match_domain, match_address, and match_local_
3488 part (NOT match_host).
3490 8. The placeholder for the Received: header didn't have a length field set.
3492 9. Added code to Exim itself and to exim_lock to test for a specific race
3493 condition that could lead to file corruption when using MBX delivery. The
3494 issue is with the lockfile that is created in /tmp. If this file is removed
3495 after a process has opened it but before that process has acquired a lock,
3496 there is the potential for a second process to recreate the file and also
3497 acquire a lock. This could lead to two Exim processes writing to the file
3498 at the same time. The added code performs the same test as UW imapd; it
3499 checks after acquiring the lock that its file descriptor still refers to
3500 the same named file.
3502 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3503 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3506 11. Added the smtp_active_hostname option. If used, this will typically be made
3507 to depend on the incoming interface address. Because $interface_address is
3508 not set up until the daemon has forked a reception process, error responses
3509 that can happen earlier (such as "too many connections") no longer contain
3512 12. If an expansion in a condition on a "warn" statement fails because a lookup
3513 defers, the "warn" statement is abandoned, and the next ACL statement is
3514 processed. Previously this caused the whole ACL to be aborted.
3516 13. Added the iplsearch lookup type.
3518 14. Added ident_timeout as a log selector.
3520 15. Added tls_certificate_verified as a log selector.
3522 16. Added a global option tls_require_ciphers (compare the smtp transport
3523 option of the same name). This controls incoming TLS connections.
3525 17. I finally figured out how to make tls_require_ciphers do a similar thing
3526 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3527 before starting the TLS session.
3529 18. Tabs are now shown as \t in -bP output.
3531 19. If the log selector return_path_on_delivery was set, Exim crashed when
3532 bouncing a message because it had too many Received: header lines.
3534 20. If two routers both had headers_remove settings, and the first one included
3535 a superfluous trailing colon, the final name in the first list and the
3536 first name in the second list were incorrectly joined into one item (with a
3537 colon in the middle).
3543 1. Added -C and -D options to the exinext utility, mainly to make it easier
3544 to include in the automated testing, but these could be helpful when
3545 multiple configurations are in use.
3547 2. The exinext utility was not formatting the output nicely when there was
3548 an alternate port involved in the retry record key, nor when there was a
3549 message id as well (for retries that were specific to a specific message
3550 and a specific host). It was also confused by IPv6 addresses, because of
3551 the additional colons they contain. I have fixed the IPv4 problem, and
3552 patched it up to do a reasonable job for IPv6.
3554 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3555 delivery, the log line now contains "pipelined" if PIPELINING was used.
3557 4. An SMTP transport process used to panic and die if the bind() call to set
3558 an explicit outgoing interface failed. This has been changed; it is now
3559 treated in the same way as a connect() failure.
3561 5. A reference to $sender_host_name in the part of a conditional expansion
3562 that was being skipped was still causing a DNS lookup. This no longer
3565 6. The def: expansion condition was not recognizing references to header lines
3566 that used bh_ and bheader_.
3568 7. Added the _cache feature to named lists.
3570 8. The code for checking quota_filecount in the appendfile transport was
3571 allowing one more file than it should have been.
3573 9. For compatibility with Sendmail, the command line option
3581 and sets the incoming protocol and host name (for trusted callers). The
3582 host name and its colon can be omitted when only the protocol is to be set.
3583 Note the Exim already has two private options, -pd and -ps, that refer to
3584 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3585 "s", but I don't think that's a major issue.
3587 10. A number of refactoring changes to the code, none of which should affect
3590 (a) The number of logging options was getting close to filling up the
3591 32-bit word that was used as a bit map. I have split them into two classes:
3592 those that are passed in the argument to log_write(), and those that are
3593 only ever tested independently outside of that function. These are now in
3594 separate 32-bit words, so there is plenty of room for expansion again.
3595 There is no change in the user interface or the logging behaviour.
3597 (b) When building, for example, log lines, the code previously used a
3598 macro that called string_cat() twice, in order to add two strings. This is
3599 not really sufficiently general. Furthermore, there was one instance where
3600 it was actually wrong because one of the argument was used twice, and in
3601 one call a function was used. (As it happened, calling the function twice
3602 did not affect the overall behaviour.) The macro has been replaced by a
3603 function that can join an arbitrary number of extra strings onto a growing
3606 (c) The code for expansion conditions now uses a table and a binary chop
3607 instead of a serial search (which was left over from when there were very
3608 few conditions). Also, it now recognizes conditions like "pam" even when
3609 the relevant support is not compiled in: a suitably worded error message is
3610 given if an attempt is made to use such a condition.
3612 11. Added ${time_interval:xxxxx}.
3614 12. A bug was causing one of the ddress fields not to be passed back correctly
3615 from remote delivery subprocesses. The field in question was not being
3616 subsequently used, so this caused to problems in practice.
3618 13. Added new log selectors queue_time and deliver_time.
3620 14. Might have fixed a bug in maildirsizefile handling that threw up
3621 "unexpected character" debug warnings, and recalculated the data
3622 unnecessarily. In any case, I expanded the warning message to give more
3625 15. Added the message "Restricted characters in address" to the statements in
3626 the default ACL that block characters like @ and % in local parts.
3628 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3629 Three changes have been made:
3631 (a) There was a serious bug; a negative response to MAIL caused the whole
3632 recipient domain to be cached as invalid, thereby blocking all messages
3633 to all local parts at the same domain, from all senders. This bug has
3634 been fixed. The domain is no longer cached after a negative response to
3635 MAIL if the sender used is not empty.
3637 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3640 (c) A new callout option, "use_sender" has been added for people who want
3641 the modified behaviour.
3647 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3650 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3651 indeed breaks things for older releases.
3653 3. Added additional logging to the case where there is a problem reading data
3654 from a filter that is running in a subprocess using a pipe, in order to
3655 try to track down a specific problem.
3657 4. Testing facility fudge: when running in the test harness and attempting
3658 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3659 getting "No route to host". Convert this to a timeout.
3661 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3664 6. Some OS don't have socklen_t but use size_t instead. This affects the
3665 fifth argument of getsockopt() amongst other things. This is now
3666 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3667 can be set for individual OS. I have set it for SunOS5, OSF1, and
3668 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3669 some earlier ones do not.
3671 7. Change 4.30/15 was not doing the test caselessly.
3673 8. The standard form for an IPv6 address literal was being rejected by address
3674 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3675 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3676 this, as well as the form without the "IPv6" on the front (but only when
3677 address literals are enabled, of course).
3679 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3681 10. Exim crashed if a message with an empty sender address specified by -f
3682 encountered a router with an errors_to setting. This could be provoked only
3683 by a command such as
3687 where an empty string was supplied; "<>" did not hit this bug.
3689 11. Installed PCRE release 4.5.
3691 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3692 remained set. It is now erased.
3694 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3695 times from message ids (which are base 36 rather than the normal 62).
3697 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3698 were being counted as actual protocol errors, and logged if the log
3699 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3700 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3701 and DATA following a set of rejected RCPTs do not count as protocol errors.
3702 In other words, Exim assumes they were pipelined, though this may not
3703 actually be the case. Of course, in all cases the client gets an
3704 appropriate error code.
3706 15. If a lookup fails in an ACL condition, a message about the failure may
3707 be available; it is used if testing the ACL cannot continue, because most
3708 such messages specify what the cause of the deferral is. However, some
3709 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3710 that caused an old message to be retained and used if a later statement
3711 caused a defer, replacing the real cause of the deferral.
3713 16. If an IP address had so many PTR records that the DNS lookup buffer
3714 was not large enough to hold them, Exim could crash while trying to process
3715 the truncated data. It now detects and logs this case.
3717 17. Further to 4.21/58, another change has been made: if (and only if) the
3718 first line of a message (the first header line) ends with CRLF, a bare LF
3719 in a subsequent header line has a space inserted after it, so as not to
3720 terminate the header.
3722 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3723 unnecessarily, used atoi() instead of strtol(), and didn't check the
3724 termination when getting file sizes from file names by regex.
3726 19. Completely re-implemented the support for maildirsize files, in the light
3727 of a number of problems with the previous contributed implementation
3728 (4.30/29). In particular:
3730 . If the quota is zero, the maildirsize file is maintained, but no quota is
3733 . If the maildir directory does not exist, it is created before any attempt
3734 to write a maildirsize file.
3736 . The quota value in the file is just a cache; if the quota is changed in
3737 the transport, the new value overrides.
3739 . A regular expression is available for excluding directories from the
3742 20. The autoreply transport checks the characters in options that define the
3743 message's headers; it allows continued headers, but it was checking with
3744 isspace() after an embedded newline instead of explicitly looking for a
3747 21. If all the "regular" hosts to which an address was routed had passed their
3748 expiry times, and had not reached their retry times, the address was
3749 bounced, even if fallback hosts were defined. Now Exim should go on to try
3752 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3753 equivalent code in the SMTP transport. Some hosts send humungous responses
3754 to HELO/EHLO, more than 1024 it seems.
3756 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3757 gives compiler warnings in some environments. I've now done it "properly",
3760 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3761 (because of problems with the built-in one) was declared to return uschar *
3762 instead of char *, causing compiler failure.
3764 25. Fixed a file descriptor leak when processing alias/forward files.
3766 26. Fixed a minor format string issue in dbfn.c.
3768 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3770 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3771 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3774 29. When a sender address is verified, it is cached, to save repeating the test
3775 when there is more than one recipient in a message. However, when the
3776 verification involves a callout, it is possible for different callout
3777 options to be set for different recipients. It is too complicated to keep
3778 track of this in the cache, so now Exim always runs a verification when a
3779 callout is required, relying on the callout cache for the optimization.
3780 The overhead is duplication of the address routing, but this should not be
3783 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3784 to be verified with callout=postmaster, and the main callout worked but the
3785 postmaster check failed, the verification correctly failed. However, if a
3786 subsequent RCPT command asked for sender verification *without* the
3787 postmaster check, incorrect caching caused this verification also to fail,
3790 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3791 it was not caching the DNS options (qualify_single, search_parents) that
3792 were used when the lookup failed. A subsequent lookup with different
3793 options therefore always gave the same answer, though there were cases
3794 where it should not have. (Example: a "domains = !$mx_any" option on a
3795 dnslookup router: the "domains" option is always processed without any
3796 widening, but the router might have qualify_single set.) Now Exim uses the
3797 cached value only when the same options are set.
3799 32. Added John Jetmore's "exipick" utility to the distribution.
3801 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3802 than a timeout (e.g. a certificate is required, and is not provided), an
3803 Exim server now closes the connection immediately. Previously it waited for
3804 the client to close - but if the client is SSL, it seems that they each
3805 wait for each other, leading to a delay before one of them times out.
3807 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3808 maintained 0.8.x compatibility because I don't think many are using it, and
3809 it is clearly obsolete.
3811 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3814 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3815 client certificate was expired. A simple patch fixes this, though I don't
3816 understand the full logic of why the verify callback is called multiple
3819 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3820 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3821 which causes problems with some clients (such as the Certicom SSL Plus
3822 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3823 disables the coutermeasure allowing Eudora to connect."
3825 38. Exim was not checking that a write() to a log file succeeded. This could
3826 lead to Bad Things if a log got too big, in particular if it hit a file
3827 size limit. Exim now panics and dies if it cannot write to a log file, just
3828 as it does if it cannot open a log file.
3830 39. Modified OS/Makefile-Linux so that it now contains
3832 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3834 The two -D definitions ensure that Exim is compiled with large file
3835 support, which makes it possible to handle log files that are bigger than
3838 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3839 instance) a domain was checked against a named list that involved a lookup,
3840 causing $domain_data to be set, then another domain was checked against the
3841 same list, then the first domain was re-checked, the value of $domain_data
3842 after the final check could be wrong. In particular, if the second check
3843 failed, it could be set empty. This bug probably also applied to
3846 41. The strip_trailing_dot option was not being applied to the address given
3847 with the -f command-line option.
3849 42. The code for reading a message's header from the spool was incrementing
3850 $received_count, but never initializing it. This meant that the value was
3851 incorrect (doubled) while delivering a message in the same process in which
3852 it was received. In the most common configuration of Exim, this never
3853 happens - a fresh exec is done - but it can happen when
3854 deliver_drop_privilege is set.
3856 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3857 it now includes up to 150 characters of the unexpected data in the log
3860 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3861 and building data strings. The size of both of these buffers was 10 000
3862 bytes - far larger than anybody would *ever* want, thought I. Needless to
3863 say, somebody hit the limit. I have increased the maximum line length to
3864 20 000 and the maximum data length of concatenated lines to 100 000. I have
3865 also fixed two bugs, because there was no checking on these buffers. Tsk,
3866 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3867 buffer is too small.
3869 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3870 lsearch lookups. Now it does.
3872 46. When parsing a route_list item in a manualroute router, a fixed-length
3873 buffer was used for the list of hosts. I made this 1024 bytes long,
3874 thinking that nobody would ever have a list of hosts that long. Wrong.
3875 Somebody had a whole pile of complicated expansion conditions, and the
3876 string was silently truncated, leading to an expansion error. It turns out
3877 that it is easier to change to an unlimited length (owing to other changes
3878 that have happened since this code was originally written) than to build
3879 structure for giving a limitation error. The length of the item that
3880 expands into the list of hosts is now unlimited.
3882 47. The lsearch lookup could not handle data where the length of text line was
3883 more than 4095 characters. Such lines were truncated, leading to shortened
3884 data being returned. It should now handle lines of any length.
3886 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3887 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3890 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3891 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3892 no longer happens in comment lines. A list of replacements is now placed
3893 at the head of all of the source files, except those whose only change is
3894 to replace PERL_COMMAND in the very first #! line.
3896 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3897 messages on the queue, with a bottom-up merge sort, using code contributed
3898 by Michael Haardt. This should make operations like -bp somewhat faster on
3899 large queues. It won't affect queue runners, except when queue_run_in_order
3902 51. Installed eximstats 1.31 in the distribution.
3904 52. Added support for SRV lookups to the dnslookup router.
3906 53. If an ACL referred to $message_body or $message_body_end, the value was not
3907 reset for any messages that followed in the same SMTP session.
3909 54. The store-handling optimization for building very long strings was not
3910 differentiating between the different store pools. I don't think this
3911 actually made any difference in practice, but I've tidied it.
3913 55. While running the routers to verify a sender address, $sender_address
3914 was still set to the sender address. This is wrong, because when routing to
3915 send a bounce to the sender, it would be empty. Therefore, I have changed
3916 it so that, while verifying a sender address, $sender_address is set to <>.
3917 (There is no change to what happens when verifying a recipient address.)
3919 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3920 target A or AAAA records (if not already returned) without resetting the
3921 qualify_single or search_parents options of the DNS resolver. These are
3922 inappropriate in this case because the targets of MX and SRV records must
3923 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3924 target that, when qualified, matched something in the local domain. These
3925 two options are now turned off when doing these lookups.
3927 57. It seems that at least some releases of Reiserfs (which does not have the
3928 concept of a fixed number of inodes) returns zero and not -1 for the
3929 number of available inodes. This interacted badly with check_spool_inodes,
3930 which assumed that -1 was the "no such thing" setting. What I have done is
3931 to check that the total number of inodes is greater than zero before doing
3932 the test of how many are available.
3934 58. When a "warn" ACL statement has a log_message modifier, the message is
3935 remembered, and not repeated. This is to avoid a lot of repetition when a
3936 message has many recipients that cause the same warning to be written.
3937 Howewer, Exim was preserving the list of already written lines for an
3938 entire SMTP session, which doesn't seem right. The memory is now reset if a
3939 new message is started.
3941 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3942 debugging output unless log_rewrite was also set.
3944 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3945 of (int)(handle) when we know that handle contains (void *)(-1).
3947 61. The Exim daemon panic-logs an error return when it closes the incoming
3948 connection. However "connection reset by peer" seems to be common, and
3949 isn't really an error worthy of noting specially, so that particular error
3952 62. When Exim is trying to find all the local interfaces, it used to panic and
3953 die if the ioctl to get the interface flags failed. However, it seems that
3954 on at least one OS (Solaris 9) it is possible to have an interface that is
3955 included in the list of interfaces, but for which you get a failure error
3956 for this call. This happens when the interface is not "plumbed" into a
3957 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3958 failure of the "get flags" call assumes that the interface is down.
3960 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3961 makes life easier for people who are doing arithmetic on fields extracted
3962 from dates, where you often get leading zeros that should not be
3963 interpreted as octal.
3965 64. Added qualify_domain to the redirect router, to override the global
3968 65. If a pathologically long header line contained very many addresses (the
3969 report of this problem mentioned 10 000) and each of them was rewritten,
3970 Exim could use up a very large amount of memory. (It kept on making new
3971 copies of the header line as it rewrote, and never released the old ones.)
3972 At the expense of a bit more processing, the header rewriting function has
3973 been changed so that it no longer eats memory in this way.
3975 66. The generation of the Received: header has been moved from the time that a
3976 message starts to be received, to the time that it finishes. The timestamp
3977 in the Received: header should now be very close to that of the <= log
3978 line. There are two side-effects of this change:
3980 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3981 logged header lines no longer include the local Received: line, because
3982 it has not yet been created. The same applies to a copy of the message
3983 that is returned to a non-SMTP sender when a message is rejected.
3985 (b) When a filter file is tested using -bf, no additional Received: header
3986 is added to the test message. After some thought, I decided that this
3989 This change does not affect the value of $received_for. It is still set
3990 after address rewriting, but before local_scan() is called.
3992 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3994 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3995 gave an unhelpful panic error message, and a defer error. I have managed to
3996 change this behaviour so that it now rejects any supplied certificate,
3997 which seems right, as the list of acceptable certificates is empty.
3999 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4000 gave an unhelpful defer error. I have not managed to make this reject any
4001 supplied certificates, but the error message it gives is "no certificate
4002 supplied", which is not helpful.
4004 70. exigrep's output now also includes lines that are not associated with any
4005 message, but which match the given pattern. Implemented by a patch from
4006 Martin Sluka, which also tidied up the Perl a bit.
4008 71. Recipient callout verification, like sender verification, was using <> in
4009 the MAIL FROM command. This isn't really the right thing, since the actual
4010 sender may affect whether the remote host accepts the recipient or not. I
4011 have changed it to use the actual sender in the callout; this means that
4012 the cache record is now keyed on a recipient/sender pair, not just the
4013 recipient address. There doesn't seem to be a real danger of callout loops,
4014 since a callout by the remote host to check the sender would use <>.
4015 [SEE ABOVE: changed after hitting problems.]
4017 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4018 temporary errors. However, in the case of such a code being given after
4019 the end of a data transmission (i.e. after ".") Exim was failing to write
4020 a retry record for the message. (Yes, there was some broken host that was
4021 actually sending 8xx at this point.)
4023 73. An unknown lookup type in a host list could cause Exim to panic-die when
4024 the list was checked. (An example that provoked this was putting <; in the
4025 middle of a list instead of at the start.) If this happened during a DATA
4026 ACL check, a -D file could be left lying around. This kind of configuration
4027 error no longer causes Exim to die; instead it causes a defer errror. The
4028 incident is still logged to the main and panic logs.
4030 74. Buglet left over from Exim 3 conversion. The message "too many messages
4031 in one connection" was written to the rejectlog but not the mainlog, except
4032 when address rewriting (yes!) was being logged.
4034 75. Added write_rejectlog option.
4036 76. When a system filter was run not as root (that is, when system_filter_user
4037 was set), the values of the $n variables were not being returned to the
4038 main process; thus, they were not subsequently available in the $sn
4041 77. Added +return_path_on_delivery log selector.
4043 78. A connection timeout was being treated differently from recipients deferred
4044 when testing hosts_max_try with a message that was older than the host's
4045 retry timeout. (The host should not be counted, thus allowing all hosts to
4046 be tried at least once before bouncing.) This may have been the cause of an
4047 occasionally reported bug whereby a message would remain on the queue
4048 longer than the retry timeout, but would be bounced if a delivery was
4049 forced. I say "may" because I never totally pinned down the problem;
4050 setting up timeout/retry tests is difficult. See also the next item.
4052 79. The ultimate address timeout was not being applied to errors that involved
4053 a combination of host plus message (for example, a timeout on a MAIL
4054 command). When an address resolved to a number of possible hosts, and they
4055 were not all tried for each delivery (e.g. because of hosts_max_try), a
4056 message could remain on the queue longer than the retry timeout.
4058 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4061 81. Fixed an obscure SMTP outgoing bug which required at least the following
4062 conditions: (a) there was another message waiting for the same server;
4063 (b) the server returned 5xx to all RCPT commands in the first message so
4064 that the message was not completed; (c) the server dropped the connection
4065 or gave a negative response to the RSET that Exim sends to abort the
4066 transaction. The observed case was a dropped connection after DATA that had
4067 been sent in pipelining mode. That is, the server had advertised PIPELINING
4068 but was not implementing it correctly. The effect of the bug was incorrect
4069 behaviour, such as trying another host, and this could lead to a crash.
4075 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4076 and daemon.c were passed as pointers to ints; they should have been
4077 pointers to socklen_t variables (which are typically unsigned ints).
4079 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4082 3. Fixed a really odd bug that affected only the testing scheme; patching a
4083 certain fixed string in the binary changed the value of another string that
4084 happened to be identical to the end of the original first string.
4086 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4087 name", it returns that address as the IP address. On some operating
4088 systems (e.g. Solaris), it also passes back the IP address string as the
4089 "host name". However, on others (e.g. Linux), it passes back an empty
4090 string. Exim wasn't checking for this, and was changing the host name to an
4091 empty string, assuming it had been canonicized.
4093 5. Although rare, it is permitted to have more than one PTR record for a given
4094 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4095 all the names associated with an address, because they do in Solaris.
4096 However, it seems that they do not in Linux for data that comes from the
4097 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4098 I found this out when I moved to a new Linux workstation and tried to run
4099 the Exim test suite.
4101 To get round this problem I have changed the code so that it now does its
4102 own call to the DNS to look up PTR records when searching for a host name.
4103 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4104 addresses that are only in /etc/hosts are still found.
4106 This behaviour is, however, controlled by an option called host_lookup_
4107 order, which defaults to "bydns:byaddr". If people want to use the other
4108 order, or indeed, just use one or the other means of lookup, they can
4109 specify it in this variable.
4111 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4112 some operating systems, this comes back from gethostbyaddr() as an empty
4113 string, and this is what Exim used to test for. However, it seems that in
4114 other systems, "." is yielded. Exim now tests for this case too.
4116 7. The values of check_spool_space and check_log_space are now held internally
4117 as a number of kilobytes instead of an absolute number of bytes. If a
4118 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4119 kilobyte. This means that much larger values can be stored.
4121 8. Exim monitor: an attempt to get the action menu when not actually pointing
4122 at a message produces an empty menu entitled "No message selected". This
4123 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4124 no entries in it ("Shell widget menu has zero width and/or height"). So I
4125 have added a single, blank menu entry in this case.
4127 9. Added ${quote_local_part.
4129 10. MIME decoding is now applied to the contents of Subject: header lines when
4132 11. Now that a reference to $sender_host_address automatically causes a reverse
4133 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4134 host lookup before query-style lookups in lists that might use this
4135 variable. This has therefore been abolished, and the "net-" prefix is no
4136 longer necessary for query-style lookups.
4138 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4139 have been a typo for LFLAGS, so it has been changed.
4141 13. The install script calls Exim with "-C /dev/null" in order to find the
4142 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4143 to be output. Howeve, since Exim outputs its version number before the
4144 error, it didn't break the script. It just looked ugly. I fixed this by
4145 always allowing "-C /dev/null" if the caller is root.
4147 14. Ignore overlarge ACL variable number when reading spool file - insurance
4148 against a later release with more variables having written the file.
4150 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4151 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4152 this, as well as the form without the "IPv6" on the front.
4154 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4155 OS/Makefile-Darwin file.
4157 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4158 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4160 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4161 variable. However, some OS use ints for inodes. I've added cast to long int
4162 to get rid of the compiler warning.
4164 19. I had forgotten to lock out "/../" in configuration file names when
4165 ALT_CONFIG_PREFIX was set.
4167 20. Routers used for verification do not need to specify transports. However,
4168 if such a router generated a host list, and callout was configured, Exim
4169 crashed, because it could not find a port number from the (non-existent)
4170 transport. It now assumes port 25 in this circumstance.
4172 21. Added the -t option to exigrep.
4174 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4175 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4176 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4177 them is set, LOOKUP_LSEARCH is forced.
4179 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4181 24. Added sender and host information to the "rejected by local_scan()" log
4182 line; previously there was no indication of these.
4184 25. Added .include_if_exists.
4186 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4187 Linux. It turns out that not all file systems support this. Apparently some
4188 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4189 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4190 not supported on the file descriptor, is now ignored when Exim is trying to
4191 sync a directory. This applies only to Linux.
4193 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4195 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4196 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4199 29. Added support for maildirsize files from supplied patch (modified a bit).
4201 30. The use of :fail: followed by an empty string could lead Exim to respond to
4202 sender verification failures with (e.g.):
4204 550 Verification failed for <xxx>
4205 550 Sender verify failed
4207 where the first response line was missing the '-' that indicates it is not
4208 the final line of the response.
4210 31. The loop for finding the name of the user that called Exim had a hardwired
4211 limit of 10; it now uses the value of finduser_retries, which is used for
4212 all other user lookups.
4214 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4217 33. Exim was neglecting to zero errno before one call of strtol() when
4218 expanding a string and expecting an integer value. On some systems this
4219 resulted in spurious "integer overflow" errors. Also, it was casting the
4220 result into an int without checking.
4222 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4223 did not work. The code looks as if it has *never* worked, though it appears
4224 to have been documented since at least releast 1.62. I have made it work.
4226 35. The "timeout_DNS" error in retry rules, also documented since at least
4227 1.62, also never worked. As it isn't clear exactly what this means, and
4228 clearly it isn't a major issue, I have abolished the feature by treating it
4229 as "timeout", and writing a warning to the main and panic logs.
4231 36. The display of retry rules for -brt wasn't always showing the error code
4234 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4235 timeout_connect_A, timeout_connect_MX.
4237 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4238 to the empty sender.
4240 39. The daemon was not analysing the content of -oX till after it had closed
4241 stderr and disconnected from the controlling terminal. This meant that any
4242 syntax errors were only noted on the panic log, and the return code from
4243 the command was 0. By re-arranging the code a little, I've made the
4244 decoding happen first, so such errors now appear on stderr, and the return
4245 code is 1. However, the actual setting up of the sockets still happens in
4246 the disconnected process, so errors there are still only recorded on the
4249 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4250 connections (as happens on some IP stacks) was logged at start up time as
4251 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4252 it from "IPv6 and IPv4", which means that two separate sockets are being
4255 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4256 says whether AF_INET or AF_INET6 was passed as an argument.
4258 42. Exiwhat output was messed up when time zones were included in log
4261 43. Exiwhat now gives more information about the daemon's listening ports,
4262 and whether -tls-on-connect was used.
4264 44. The "port" option of the smtp transport is now expanded.
4266 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4267 silently ignored. Now an error message is written to the main and panic
4270 46. There's a new ACL modifier called "logwrite" which writes to a log file
4271 as soon as it is encountered.
4273 47. Added $local_user_uid and $local_user_gid at routing time.
4275 48. Exim crashed when trying to verify a sender address that was being
4278 49. Exim was recognizing only a space character after ".include". It now also
4279 recognizes a tab character.
4281 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4282 extracting the relevant information from the specification. The man page no
4283 longer contains scrambled data for the -d option, and I've added a section
4284 at the front about calling Exim under different names.
4286 51. Added "extra_headers" argument to the "mail" command in filter files.
4288 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4291 53. Installed eximstats 1.29.
4293 54. Added transport_filter_timeout as a generic transport option.
4295 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4296 Cc: header lines. This was required by RFC 822, but it not required by RFC
4299 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4300 incoming messages that did not have them. Now it does so only if the
4301 message originates locally, that is, if there is no associated remote host
4302 address. When Resent- header lines are present, this applies to the Resent-
4303 lines rather than the non-Resent- lines.
4305 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4306 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4308 58. Messages for configuration errors now include the name of the main
4309 configuration file - useful now that there may be more than one file in a
4310 list (.included file names were always shown).
4312 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4313 for those rare installations that do not start the daemon as root or run it
4314 setuid root. I've cut out the call to initgroups() if the daemon is not
4317 60. The Exim user and group can now be bound into the binary as text strings
4318 that are looked up at the start of Exim's processing.
4320 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4322 62. Added $mailstore_basename variable.
4324 63. Installed patch to sieve.c from Michael Haardt.
4326 64. When Exim failed to open the panic log after failing to open the main log,
4327 the original message it was trying to log was written to stderr and debug
4328 output, but if they were not available (the usual case in production), it
4329 was lost. Now it is written to syslog before the two lines that record the
4330 failures to open the logs.
4332 65. Users' Exim filters run in subprocesses under the user's uid. It is
4333 possible for a "deliver" command or an alias in a "personal" command to
4334 provoke an address rewrite. If logging of address rewriting is configured,
4335 this fails because the process is not running as root or exim. There may be
4336 a better way of dealing with this, but for the moment (because 4.30 needs
4337 to be released), I have disabled address rewrite logging when running a
4338 filter in a non-root, non-exim process.
4344 1. The buildconfig auxiliary program wasn't quoting the value set for
4345 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4346 not defined. This bug was masked in 4.22 by the effect that was fixed in
4349 2. Some messages that were rejected after a message id was allocated were
4350 shown as "incomplete" by exigrep. It no longer does this for messages that
4351 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4353 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4354 have allow_domain_literals set, the ID did not get logged in the <= line.
4355 Domain literals are now always recognized in Message-ID: header lines.
4357 4. The first argument for a ${extract expansion item is the key name or field
4358 number. Leading and trailing spaces in this item were not being ignored,
4359 causing some misleading effects.
4361 5. When deliver_drop_privilege was set, single queue runner processes started
4362 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4363 same command in the process it spins off) were not dropping privilege.
4365 6. When the daemon running as "exim" started a queue runner, it always
4366 re-executed Exim in the spun-off process. This is a waste of effort when
4367 deliver_drop_privilege is set. The new process now just calls the
4368 queue-runner function directly.
4374 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4377 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4378 ignored. Though the use of -oP was forcing the writing of a pid file, it
4379 was always written to the default place.
4381 3. If the message "no IP address found for host xxxx" is generated during
4382 incoming verification, it is now followed by identification of the incoming
4383 connection (so you can more easily find what provoked it).
4385 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4387 5. Added some features to "harden" Exim a bit more against certain attacks:
4389 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4390 be put in Local/Makefile. This is like the never_users runtime option,
4391 but it cannot be overridden. The default setting is "root".
4393 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4394 prefix string with which any file named in a -C command line option
4397 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4398 is retained for -C and -D only if the caller of Exim is root. Without
4399 it, the exim user may also use -C and -D and retain privilege.
4401 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4402 command line option is disabled.
4404 6. Macro names set by the -D option must start with an upper case letter, just
4405 like macro names defined in the configuration file.
4407 7. Added "dereference=" facility to LDAP.
4409 8. Two instances of the typo "uknown" in the source files are fixed.
4411 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4412 the Configure-Makefile script screwed up while processing it.
4414 10. Incorporated PCRE 4.4.
4416 11. The SMTP synchronization check was not operating right at the start of an
4417 SMTP session. For example, it could not catch a HELO sent before the client
4418 waited for the greeting. There is now a check for outstanding input at the
4419 point when the greeting is written. Because of the duplex, asynchronous
4420 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4421 way, but not yet received, when the check is performed.
4423 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4424 on TCP/IP sockets, because this apparently causes some broken clients to
4427 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4428 unchanged) from the Cygwin maintainer.
4430 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4431 was supported instead of testing for mbx. Effectively a typo.
4433 15. The spa authenticator server code was not checking that the input it
4434 received was valid base64.
4436 16. The debug output line for the "set" modifier in ACLs was not showing the
4437 name of the variable that was being set.
4439 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4441 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4442 lookup on its own. Something else was needed to trigger the lookup. For
4443 example, a match in host_lookup or the need for a host name in a host list.
4444 Now, if $sender_host_name is referenced and the host name has not yet been
4445 looked up, a lookup is performed. If the lookup fails, the variable remains
4446 empty, and $host_lookup_failed is set to "1".
4448 19. Added "eqi" as a case-independent comparison operator.
4450 20. The saslauthd authentication condition could segfault if neither service
4451 nor realm was specified.
4453 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4454 error message that was logged was misleading, and incoming SMTP
4455 connections were dropped. The message is now more accurate, and temporary
4456 errors are given to SMTP connections.
4458 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4459 (except RSET) with a 421 error, until QUIT is received. However, it was
4460 failing to send a response to QUIT.
4462 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4463 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4464 ACL is now run after the helo_try_verify_hosts code.
4466 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4467 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4468 case-independent, so other case variants are also recognized). Apparently
4469 some systems use these upper case variants.
4471 25. If more than two messages were waiting for the same host, and a transport
4472 filter was specified for the transport, Exim sent two messages over the
4473 same TCP/IP connection, and then failed with "socket operation on non-
4474 socket" when it tried to send the third.
4476 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4478 27. The extern definition of crypt16() in expand.c was not being excluded when
4479 the OS had its own crypt16() function.
4481 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4482 as a preferred synonym for return_size_limit, both as an option and as an
4485 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4487 30. Changed the default configuration ACL to relax the local part checking rule
4488 for addresses that are not in any local domains. For these addresses,
4489 slashes and pipe symbols are allowed within local parts, but the sequence
4490 /../ is explicitly forbidden.
4492 31. SPA server authentication was not clearing the challenge buffer before
4495 32. log_message in a "warn" ACL statement was writing to the reject log as
4496 well as to the main log, which contradicts the documentation and doesn't
4497 seem right (because no rejection is happening). So I have stopped it.
4499 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4500 However, I am unable to do any testing of this.
4502 34. Fixed an infelicity in the appendfile transport. When checking directories
4503 for a mailbox, to see if any needed to be created, it was accidentally
4504 using path names with one or more superfluous leading slashes; tracing
4505 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4507 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4508 contents are added to the log line that is written for every discarded
4509 recipient. (Previously a log_message setting was ignored.)
4511 36. The ${quote: operator now quotes the string if it is empty.
4513 37. The install script runs exim in order to find its version number. If for
4514 some reason other than non-existence or emptiness, which it checks, it
4515 could not run './exim', it was installing it with an empty version number,
4516 i.e. as "exim-". This error state is now caught, and the installation is
4519 38. An argument was missing from the function that creates an error message
4520 when Exim fails to connect to the socket for saslauthd authentication.
4521 This could cause Exim to crash, or give a corrupted message.
4523 39. Added isip, isip4, and isip6 to ${if conditions.
4525 40. The ACL variables $acl_xx are now saved with the message, and can be
4526 accessed later in routers, transports, and filters.
4528 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4529 strings in the file are not string-expanded.
4531 42. If a MAIL command specified a SIZE value that was too large to fit into an
4532 int variable, the check against message_size_limit failed. Such values are
4533 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4534 day this will have to be increased, but I don't think I want to be around
4535 when emails are that large.
4542 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4543 iconv() is not standard in FreeBSD.
4545 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4546 IPv6 enabled. The observed symptom was a segmentation fault on return from
4547 the function os_common_find_running_interfaces() in src/os.c.
4549 3. In the check_special_case() function in daemon.c I had used "errno" as an
4550 argument name, which causes warnings on some systems. This was basically a
4551 typo, since it was named "eno" in the comments!
4553 4. The code that waits for the clock to tick (at a resolution of some fraction
4554 of a second) so as to ensure message-id uniqueness was always waiting for
4555 at least one whole tick, when it could have waited for less. [This is
4556 almost certainly not relevant at current processor speeds, where it is
4557 unlikely to ever wait at all. But we try to future-proof.]
4559 5. The function that sleeps for a time interval that includes fractions of a
4560 second contained a race. It did not block SIGALRM between setting the
4561 timer, and suspending (a couple of lines later). If the interval was short
4562 and the sigsuspend() was delayed until after it had expired, the suspension
4563 never ended. On busy systems this could lead to processes getting stuck for
4566 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4567 process, before it forks any delivery processes. The open lookup caching
4568 mechanism meant that the open file or database connection was passed into
4569 the delivery process. The problem was that delivery processes always tidy
4570 up cached lookup data. This could cause a problem for the next delivery
4571 process started by the queue runner, because the external queue runner
4572 process does not know about the closure. So the next delivery process
4573 still has data in the lookup cache. In the case of a file lookup, there was
4574 no problem because closing a file descriptor in a subprocess doesn't affect
4575 the parent. However, if the lookup was caching a connection to a database,
4576 the connection was closed, and the second delivery process was likely to
4577 see errors such as "PGSQL: query failed: server closed the connection
4578 unexpectedly". The problem has been fixed by closing all cached lookups
4579 in a queue runner before running a delivery process.
4581 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4582 seem to have the "const" qualifier which it has on other OS. I've
4585 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4586 *of the same type* (client or server) with the same public name that an
4587 error should be diagnosed.
4589 9. When Exim looked up a host name for an IP address, but failed to find the
4590 original IP address when looking up the host name (a safety check), it
4591 output the message "<ip address> does not match any IP for NULL", which was
4592 confusing, to say the least. The bug was that the host name should have
4593 appeared instead of "NULL".
4595 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4596 user that is built into the binary, and the -C or -D options is used, root
4597 privilege is dropped before the configuration file is read. In addition,
4598 logging is switched to stderr instead of the normal log files. If the
4599 configuration then re-defines the Exim user, the unprivileged environment
4600 is probably not what is expected, so Exim logs a panic warning message (but
4603 However, if deliver_drop_privilege is set, the unprivileged state may well
4604 be exactly what is intended, so the warning has been cut out in that case,
4605 and Exim is allowed to try to write to its normal log files.
4611 1. smtp_return_error_details was not giving details for temporary sender
4612 or receiver verification errors.
4614 2. Diagnose a configuration error if two authenticators have the same public
4617 3. Exim used not to create the message log file for a message until the first
4618 delivery attempt. This could be confusing when incoming messages were held
4619 for policy or load reasons. The message log file is now created at the time
4620 the message is received, and an initial "Received" line is written to it.
4622 4. The automatically generated man page for command line options had a minor
4623 bug that caused no ill effects; however, a more serious problem was that
4624 the procedure for building the man page automatically didn't always
4625 operate. Consequently, release 4.20 contains an out-of-date version. This
4626 shouldn't happen again.
4628 5. When building Exim with embedded Perl support, the script that builds the
4629 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4630 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4632 6. The freeze_tell option was not being used for messages that were frozen on
4633 arrival, either by an ACL or by local_scan().
4635 7. Added the smtp_incomplete_transaction log selector.
4637 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4638 was accepting AUTH without a new EHLO.
4640 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4641 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4644 10. Logging of TCP/IP connections (when configured) now happens in the main
4645 daemon process instead of the child process, so that the TCP/IP connection
4646 count is more accurate (but it can never be perfect).
4648 11. The use of "drop" in a nested ACL was not being handled correctly in the
4649 outer ACL. Now, if condition failure induced by the nested "drop" causes
4650 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4651 or "require"), the connection is dropped.
4653 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4654 that yield "discard" can now be used with an "accept" or a "discard" verb,
4655 but an error is generated for any others (because I can't see a useful way
4656 to define what should happen).
4658 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4659 are now processed in the same way as lines in the Exim configuration file.
4660 In particular, continuation lines are supported.
4662 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4664 15. Added -ti meaning -t -i.
4666 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4667 domains, and warn by logging if others are found.
4669 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4670 returned from GIFCONF when Exim is trying to find the list of interfaces on
4671 a host. The code in os.c has been modified to copy each ifreq to an aligned
4672 structure in all cases.
4674 Also, in some cases, the returned ifreq's were being copied to a 'struct
4675 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4676 means the last couple of bytes of an IPv6 address could be chopped if the
4677 ifreq contained only a normal sockaddr (14 bytes storage).
4679 18. Named domain lists were not supported in the hosts_treat_as_local option.
4680 An entry such as +xxxx was not recognized, and was treated as a literal
4683 19. Ensure that header lines added by a DATA ACL are included in the reject log
4684 if the ACL subsequently rejects the message.
4686 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4687 MD5 (which is deprecated).
4689 21. When testing a filter file using -bf, Exim was writing a message when it
4690 took the sender from a "From " line in the message, but it was not doing so
4691 when it took $return_path from a Return-Path: header line. It now does.
4693 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4694 with a valid header line field name (a series of printing characters
4695 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4697 23. Changed "disc" in the source to "disk" to conform to the documentation and
4698 the book and for uniformity.
4700 24. Ignore Sendmail's -Ooption=value command line item.
4702 25. When execve() failed while trying to run a command in a pipe transport,
4703 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4704 could be confused with a return value of 69 from the command itself. This
4705 has been changed to 127, the value the shell returns if it is asked to run
4706 a non-existent command. The wording for the related log line suggests a
4707 non-existent command as the problem.
4709 26. If received_header_text expands to an empty string, do not add a Received:
4710 header line to the message. (Well, it adds a token one on the spool, but
4711 marks it "old" so that it doesn't get used or transmitted.)
4713 27. Installed eximstats 1.28 (addition of -nt option).
4715 28. There was no check for failure on the call to getsockname() in the daemon
4716 code. This can fail if there is a shortage of resources on the system, with
4717 ENOMEM, for example. A temporary error is now given on failure.
4719 29. Contrary to the C standard, it seems that in some environments, the
4720 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4721 program. Exim now does this explicitly; it affects the formatting of
4722 timestamps using strftime().
4724 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4725 complaints. I've now initialized all the variables, to avoid this.
4727 32. Header lines added by a system filter were not being "seen" during
4728 transport-time rewrites.
4730 33. The info_callback() function passed to OpenSSL is set up with type void
4731 (*)(SSL *, int, int), as described somewhere. However, when calling the
4732 function (actually a macro) that sets it up, the type void(*)() is
4733 expected. I've put in a cast to prevent warnings from picky compilers.
4735 34. If a DNS black list lookup found a CNAME record, but there were no A
4736 records associated with the domain it pointed at, Exim crashed.
4738 35. If a DNS black list lookup returned more than one A record, Exim ignored
4739 all but the first. It now scans all returned addresses if a particular IP
4740 value is being sought. In this situation, the contents of the
4741 $dnslist_value variable are a list of all the addresses, separated by a
4744 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4745 to accept a host name and all its aliases if the forward lookup for any of
4746 them yielded the IP address of the incoming connection. Now it accepts only
4747 those names whose forward lookup yields the correct IP address. Any other
4748 names are discarded. This closes a loophole whereby a rogue DNS
4749 administrator could create reverse DNS records to break through a
4750 wildcarded host restriction in an ACL.
4752 37. If a user filter or a system filter that ran in a subprocess used any of
4753 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4754 the wrong values were passed to the pipe command ($thisaddress had the
4755 value of $0, $0 had the value of $1, etc). This bug was introduced by
4756 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4758 38. Improved the line breaking for long SMTP error messages from ACLs.
4759 Previously, if there was no break point between 40 and 75 characters, Exim
4760 left the rest of the message alone. Two changes have been made: (a) I've
4761 reduced the minimum length to 35 characters; (b) if it can't find a break
4762 point between 35 and 75 characters, it looks ahead and uses the first one
4763 that it finds. This may give the occasional overlong line, but at least the
4764 remaining text gets split now.
4766 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4767 file descriptors might be low, and that setting 1000 would always raise it.
4768 It turns out that in some environments, the limit is already over 1000 and
4769 that lowering it causes trouble. So now Exim takes care not to decrease it.
4771 40. When delivering a message, the value of $return_path is set to $sender_
4772 address at the start of routing (routers may change the value). By an
4773 oversight, this default was not being set up when an address was tested by
4774 -bt or -bv, which affected the outcome if any router or filter referred to
4777 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4778 log only when "message" or "log_message" are set. However, if one of the
4779 conditions was an address verification, or a call to a nested ACL, the
4780 messages generated by the underlying test were being passed through. This
4781 no longer happens. The underlying message is available in $acl_verify_
4782 message for both "message" and "log_message" expansions, so it can be
4783 passed through if needed.
4785 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4786 new expansion $bh_ to give the encoded byte string without charset
4787 translation. Translation happens only if iconv() is available; HAVE_ICONV
4788 indicates this at build time. HEADERS_CHARSET gives the charset to
4789 translate to; headers_charset can change it in the configuration, and
4790 "headers charset" can change it in an individual filter file.
4792 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4793 that creates RFC 2047 encoded "words" labels them as that charset instead
4794 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4795 expansion operator; (ii) when Exim creates a From: line for a local
4796 message; (iii) when a header line is rewritten to include a "phrase" part.
4798 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4799 buggy, causing it to skip the first lines of messages whose message ID
4800 ended in 'D'. This would not have bitten before Exim release 4.14, because
4801 message IDs were unlikely to end in 'D' before then. The effect was to have
4802 incorrect size information for certain domains.
4804 45. #include "config.h" was missing at the start of the crypt16.c module. This
4805 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4808 46. If there was a timeout during a "random" callout check, Exim treated it as
4809 a failure of the random address, and carried on sending RSET and the real
4810 address. If the delay was just some slowness somewhere, the response to the
4811 original RCPT would be taken as a response to RSET and so on, causing
4812 mayhem of various kinds.
4814 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4815 when I implemented it. It didn't allow for the fact that some option values
4816 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4817 the right test for positive values.
4819 48. Domain names in DNS records are case-independent. Exim always looks them up
4820 in lower case. Some resolvers return domain names in exactly the case they
4821 appear in the zone file, that is, they may contain uppercase letters. Not
4822 all resolvers do this - some return always lower case. Exim was treating a
4823 change of case by a resolver as a change of domain, similar to a widening
4824 of a domain abbreviation. This triggered its re-routing code and so it was
4825 trying to route what was effectively the same domain again. This normally
4826 caused routing to fail (because the router wouldn't handle the domain
4827 twice). Now Exim checks for this case specially, and just changes the
4828 casing of the domain that it ultimately uses when it transmits the message
4831 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4834 50. If a filter generated a file delivery with a non-absolute name (possible if
4835 no home directory exists for the router), the forbid_file option was not
4838 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4839 the existing equality matching.
4841 52. Exim was using ints instead of ino_t variables in some places where it was
4842 dealing with inode numbers.
4844 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4845 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4846 called TMPDIR, and if it finds it is different, it changes its value.
4848 54. The smtp_printf() function is now made available to local_scan() so
4849 additional output lines can be written before returning. There is also an
4850 smtp_fflush() function to enable the detection of a dropped connection.
4851 The variables smtp_input and smtp_batched_input are exported to
4854 55. Changed the default runtime configuration: the message "Unknown user"
4855 has been removed from the ACL, and instead placed on the localuser router,
4856 using the cannot_route_message feature. This means that any verification
4857 failures that generate their own messages won't get overridden. Similarly,
4858 the "Unrouteable address" message that was in the ACL for unverifiable
4859 relay addresses has also been removed.
4861 56. Added hosts_avoid_esmtp to the smtp transport.
4863 57. The exicyclog script was not checking for the esoteric option
4864 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4865 will work only if exicyclog is run under the appropriate euid.
4867 58. Following a discussion on the list, the rules by which Exim recognises line
4868 endings on incoming messages have been changed. The -dropcr and drop_cr
4869 options are now no-ops, retained only for backwards compatibility. The
4870 following line terminators are recognized: LF CRLF CR. However, special
4871 processing applies to CR:
4873 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4874 nor a local message in the state where . is a terminator.
4876 (ii) If a bare CR is encountered in a header line, an extra space is added
4877 after the line terminator so as not to end the header. The reasoning
4878 behind this is that bare CRs in header lines are most likely either
4879 to be mistakes, or people trying to play silly games.
4881 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4882 was being incorrectly given as 18 bytes larger than it should have been.
4883 This is a VOB (very old bug).
4885 60. This may never have affected anything current, but just in case it has:
4886 When the local host is found other than at the start of a list of hosts,
4887 the local host, those with the same MX, and any that follow, are discarded.
4888 When the list in question was part of a longer list of hosts, the following
4889 hosts (not currently being processed) were also being discarded. This no
4890 longer happens. I'm not sure if this situation could ever has previously
4893 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4896 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4897 header. This is something that is recommended in a new Internet Draft, and
4898 is something that is documented as being done by Sendmail. There are two
4899 possible values. For messages generated by the autoreply transport, Exim
4902 Auto-Submitted: auto-replied
4904 whereas for all other generated messages (e.g. bounces) it adds
4906 Auto-Submitted: auto-generated
4908 63. The "personal" condition in filters now includes a test for the
4909 Auto-Submitted: header. If it contains the string "auto-" the message it
4910 not considered personal.
4912 64. Added rcpt_include_affixes as a generic transport option.
4914 65. Added queue_only_override (default true).
4916 66. Added the syslog_duplication option.
4918 67. If what should have been the first header line of a message consisted of
4919 a space followed by a colon, Exim was mis-interpreting it as a header line.
4920 It isn't of course - it is syntactically invalid and should therefore be
4921 treated as the start of the message body. The misbehaviour could have
4922 caused a number of strange effects, including loss of data in subsequent
4923 header lines, and spool format errors.
4925 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4926 client host had authenticated. This control can now be exercised by an ACL
4927 for more flexibility.
4929 69. By default, callouts do not happen when testing with -bh. There is now a
4930 variant, -bhc, which does actually run the callout code, including
4931 consulting and updating the callout cache.
4933 70. Added support for saslauthd authentication, courtesy of Alexander
4936 71. If statvfs() failed on the spool or log directories while checking their
4937 size for availability, Exim confusingly gave the error "space shortage".
4938 Furthermore, in debugging mode it crashed with a floating point exception.
4939 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4940 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4941 serious problem, Exim now writes to the main and panic logs when this
4942 happens, with details of the failure. It then refuses to accept the
4943 incoming message, giving the message "spool directory problem" or "log
4944 directory problem" with a 421 code for SMTP messages.
4946 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4947 0, 1, and 2 exist, because some OS complain for execs without them (see
4948 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4949 descriptors. However, the code omitted to check that the open succeeded,
4950 causing mysterious errors if for some reason the permissions on /dev/null
4951 got screwed. Now Exim writes a message to the main and panic logs, and
4952 bombs out if it can't open /dev/null.
4954 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4955 interact so that it is all more flexible. It is supposed to remain
4956 backwards compatible. Also added extra_local_interfaces.
4958 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4959 to bomb out with an assertion failure - to the client this appears as a
4960 connection drop. This problem occurs in the part of the code that was taken
4961 from the Samba project. Fortunately, the assertion is in a very simple
4962 function, so I have fixed this by reproducing the function inline in the
4963 one place where it is called, and arranging for authentication to fail
4964 instead of killing the process with assert().
4966 75. The SPA client code was not working when the server requested OEM rather
4967 than Unicode encoding.
4969 76. Added code to make require_files with a specific uid setting more usable in
4970 the case where statting the file as root fails - usually a non-root-mounted
4971 NFS file system. When this happens and the failure is EACCES, Exim now
4972 forks a subprocess and does the per-uid checking as the relevant uid.
4974 77. Added process_log_path.
4976 78. If log_file_path was not explicitly set, a setting of check_log_space or
4977 check_log_inodes was ignored.
4979 79. If a space check for the spool or log partitions fails, the incident is now
4980 logged. Of course, in the latter case the data may get lost...
4982 80. Added the %p formatting code to string_format() so that it can be used to
4983 print addresses in debug_print(). Adjusted all the address printing in the
4984 debugging in store.c to use %p rather than %d.
4986 81. There was a concern that a line of code in smtp_in.c could overflow a
4987 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4988 initially expressed, the concern was not well-founded, because trailing
4989 spaces are removed early. However, if the trailing spaces were followed by
4990 a NULL, they did not get removed, so the overflow was possible. Two fixes
4993 (a) I re-wrote the offending code in a cleaner fashion.
4994 (b) If an incoming SMTP command contains a NULL character, it is rejected
4997 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4998 runs initgroups(), so that if the Exim user is in any additional groups,
4999 they will be used during message reception.
5005 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob