1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.324 2006/03/08 09:43:10 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
17 PH/02 The ipliteral router was not recognizing addresses of the form user@
18 [ipv6:....] because it didn't know about the "ipv6:" prefix.
20 PH/03 Added disable_ipv6.
22 PH/04 Changed $reply_address to use the raw form of the headers instead of the
23 decoded form, because it is most often used to construct To: headers
24 lines in autoreplies, and the decoded form may well be syntactically
25 invalid. However, $reply_address has leading white space removed, and all
26 newlines turned into spaces so that the autoreply transport does not
29 PH/05 If group was specified without a user on a router, and no group or user
30 was specified on a transport, the group from the router was ignored.
32 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
33 for visible compile-time settings that can be used to change these
34 numbers, for those that want even more. Backwards compatibility with old
35 spool files has been maintained. However, going back to a previous Exim
36 release will lost any variables that are in spool files.
38 PH/07 Two small changes when running in the test harness: increase delay when
39 passing a TCP/IP connection to a new process, in case the original
40 process has to generate a bounce, and remove special handling of
41 127.0.0.2 (sic), which is no longer necessary.
43 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
44 be the same on different OS.
46 PH/09 Moved a debug statement in filter processing to avoid a race problem when
49 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
50 whether --show-vars was specified or not
52 JJ/02 exipick: Added support for new ACL variable spool format introduced
55 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
56 syntactically invalid From: or Reply-to: line, and a filter used this to
57 generate an autoreply, and therefore failed to obtain an address for the
58 autoreply, Exim could try to deliver to a non-existent relative file
59 name, causing unrelated and misleading errors. What now happens is that
60 it logs this as a hard delivery error, but does not attempt to create a
63 PH/11 The exinext utility has a -C option for testing purposes, but although
64 the given file was scanned by exinext itself; it wasn't being passed on
67 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
68 an end-of-file indication when reading a command response.
70 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
71 compiled. In many other places in Exim, IPv6 addresses are always
72 recognized, so I have changed this. It also means that IPv4 domain
73 literals of the form [IPV4:n.n.n.n] are now always recognized.
75 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
76 used if the router is not running as root, for example, when verifying at
77 ACL time, or when using -bh. The debugging output from this situation was
78 non-existent - all you got was a failure to exec. I have made two
81 (a) Failures to set uid/gid, the current directory, or a process leader
82 in a subprocess such as that created by queryprogram now generate
83 suitable debugging ouput when -d is set.
85 (b) The queryprogram router detects when it is not running as root,
86 outputs suitable debugging information if -d is set, and then runs
87 the subprocess without attempting to change uid/gid.
89 PH/15 Minor change to Makefile for building test_host (undocumented testing
92 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
93 additional section of a DNS packet that returns MX or SRV records.
94 Instead, it always explicitly searches for A/AAAA records. This avoids
95 major problems that occur when a DNS server includes only records of one
96 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
97 fixed another bug: if SRV records were looked up and the corresponding
98 address records were *not* found in the additional section, the port
99 values from the SRV records were lost.
101 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
102 using the correct key (the original address) when searching the retry
103 rules in order to find which one to use for generating the retry hint.
105 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
106 adding the default one. Similarly, if it contains a Reply-To: header, the
107 errors_reply_to option, if set, is not used.
109 PH/19 When calculating a retry time, Exim used to measure the "time since
110 failure" by looking at the "first failed" field in the retry record. Now
111 it does not use this if it is later than than the arrival time of the
112 message. Instead it uses the arrival time. This makes for better
113 behaviour in cases where some deliveries succeed, thus re-setting the
114 "first failed" field. An example is a quota failure for a huge message
115 when small messages continue to be delivered. Without this change, the
116 "time since failure" will always be short, possible causing more frequent
117 delivery attempts for the huge message than are intended.
119 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
120 $1, $2, $3) because the numerical variables can be reset during some
121 expansion items (e.g. "match"), thereby losing the authentication data.
123 PH/21 Make -bV show the size of off_t variables so that the test suite can
124 decide whether to run tests for quotas > 2G.
126 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
127 mailbox_size, and mailbox_filecount in the appendfile transport. If a
128 filecount value is greater than 2G or if a quota value is greater than 2G
129 on a system where the size of off_t is not greater than 4, a panic error
132 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
133 never match. The debug and -bh output now contains an explicit error
134 message indicating a malformed IPv4 address or mask.
136 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
137 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
140 PH/25 Do not write to syslog when running in the test harness. The only
141 occasion when this arises is a failure to open the main or panic logs
142 (for which there is an explicit test).
144 PH/26 Added the /no_tell option to "control=freeze".
146 PH/27 If a host name lookup failed very early in a connection, for example, if
147 the IP address matched host_lookup and the reverse lookup yielded a name
148 that did not have a forward lookup, an error message of the form "no IP
149 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
150 could be logged. Now it outputs the IP address instead of "NULL".
152 PH/28 An enabling patch from MH: add new function child_open_exim2() which
153 allows the sender and the authenticated sender to be set when
154 submitting a message from within Exim. Since child_open_exim() is
155 documented for local_scan(), the new function should be too.
157 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
158 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
159 results in an empty string is now treated as unset.
161 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
163 PH/31 Added sender_verify_fail logging option.
165 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
166 needed by RFC 822 but not by RFC 2822 was commented out. I have now
167 tidied the source and removed it altogether.
169 PH/33 When a queue run was abandoned because the load average was too high, a
170 log line was always written; now it is written only if the queue_run log
171 selector is set. In addition, the log line for abandonment now contains
172 information about the queue run such as the pid. This is always present
173 in "start" and "stop" lines but was omitted from the "abandon" line.
175 PH/34 Omit spaces between a header name and the colon in the error message that
176 is given when verify = headers_syntax fails (if there are lots of them,
177 the message gets confusing).
179 PH/35 Change the default for dns_check_names_pattern to allow slashes within
180 names, as there are now some PTR records that contain slashes. This check
181 is only to protect against broken name servers that fall over on strange
182 characters, so the fact that it applies to all lookups doesn't matter.
184 PH/36 Now that the new test suite is complete, we can remove some of the
185 special code in Exim that was needed for the old test suite. For example,
186 sorting DNS records because real resolvers return them in an arbitrary
187 order. The new test suite's fake resolver always returns records in the
190 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
191 bounces) except when queue_only is set, to avoid logging races between
192 the different processes.
194 PH/38 Panic-die if .include specifies a non-absolute path.
196 PH/39 A tweak to the "H" retry rule from its user.
198 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
199 a label. They prevented compilation on older perls.
201 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
202 a warning to be raised on newish perls.
204 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
205 on queue. Changes to match documented behaviour of showing count of
206 messages matching specified criteria.
208 PH/40 Changed the default ident timeout from 30s to 5s.
210 PH/41 Added support for the use of login_cap features, on those BSD systems
211 that have them, for controlling the resources used by pipe deliveries.
213 PH/42 The content-scanning code uses fopen() to create files in which to put
214 message data. Previously it was not paying any attention to the mode of
215 the files. Exim runs with umask(0) because the rest of the code creates
216 files with open(), and sets the required mode explicitly. Thus, these
217 files were ending up world-writeable. This was not a big issue, because,
218 being within the spool directory, they were not world-accessible. I have
219 created a function called modefopen, which takes an additional mode
220 argument. It sets umask(777), creates the file, chmods it to the required
221 mode, then resets the umask. All the relevant calls to fopen() in the
222 content scanning code have been changed to use this function.
224 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
225 to 24 hours. This avoids potential overflow problems when processing G
226 and H retry rules. I suspect nobody ever tinkers with this value.
228 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
230 PH/45 When the plaintext authenticator is running as a client, the server's
231 challenges are checked to ensure they are valid base64 strings. By
232 default, the authentication attempt is cancelled if an invalid string is
233 received. Setting client_ignore_invalid_base64 true ignores these errors.
234 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
235 they are received. Thus, the responses can be made to depend on the
236 challenges. If an invalid string is ignored, an empty string is placed in
239 PH/46 Messages that are created by the autoreply transport now contains a
240 References: header, in accordance with RFCs 2822 and 3834.
242 PH/47 Added authenticated_sender_force to the smtp transport.
244 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
246 PH/49 Installed latest patch from the Sieve maintainer.
248 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
249 was also set, the appendfile transport was unnecessarily scanning a
250 directory of message files (e.g. for maildir delivery) to find the count
251 of files (along with the size), even though it did not need this
252 information. It now does the scan only if it needs to find either the
253 size of the count of files.
255 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
257 PH/52 Two bugs concerned with error handling when the smtp transport is
260 (i) Exim was not creating retry information for temporary errors given
261 for individual recipients after the DATA command when the smtp transport
262 was used in LMTP mode. This meant that they could be retried too
263 frequently, and not timed out correctly.
265 (ii) Exim was setting the flag that allows error details to be returned
266 for LMTP errors on RCPT commands, but not for LMTP errors for individual
267 recipients that were returned after the DATA command.
269 PH/53 This is related to PH/52, but is more general: for any failing address,
270 when detailed error information was permitted to be returned to the
271 sender, but the error was temporary, then after the final timeout, only
272 "retry timeout exceeded" was returned. Now it returns the full error as
273 well as "retry timeout exceeded".
275 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
276 do this, and (what is worse) MTAs that accept it.
278 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
279 will now be deprecated.
281 PH/56 New os.c-cygwin from the Cygwin maintainer.
283 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
284 formats (previously only available in exim formats via -bpr, -bpru,
285 and -bpra. Now also available in native and exiqgrep formats)
287 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
288 with very large, slow to parse queues
290 JJ/08 exipick: added ! as generic prefix to negate any criteria format
292 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
298 PH/01 Two changes to the default runtime configuration:
300 (1) Move the checks for relay_from_hosts and authenticated clients from
301 after to before the (commented out) DNS black list checks.
303 (2) Add control=submission to the relay_from_hosts and authenticated
304 clients checks, on the grounds that messages accepted by these
305 statements are most likely to be submissions.
307 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
309 (1) Generate an error if the third argument for the ${prvs expansion is
312 (2) Treat a missing third argument of ${prvscheck as if it were an empty
315 (3) Reset the variables that are obtained from the first argument of
316 ${prvscheck and used in the second argument before leaving the code,
317 because their memory is reclaimed, so using them afterwards may do
320 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
321 one (it's much easier than Tom thought :-).
323 (5) Because of (4), we can now allow for the use of $prvscheck_result
324 inside the third argument.
326 PH/03 For some reason, the default setting of PATH when running a command from
327 a pipe transport was just "/usr/bin". I have changed it to
330 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
331 anything to be listed in the output from -bV.
333 PH/05 When a filter generated an autoreply, the entire To: header line was
334 quoted in the delivery log line, like this:
336 => >A.N.Other <ano@some.domain> <original@ddress> ...
338 This has been changed so that it extracts the operative address. There
339 may be more than one such address. If so, they are comma-separated, like
342 => >ano@some.domain,ona@other.domain <original@ddress> ...
344 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
345 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
346 not being looked up in the rDNS to get a host name, Exim was showing the
347 IP address twice in Received: lines, even though the IP addresses were
348 identical. For example:
350 Received: from [1.2.3.4] (helo=[1.2.3.4])
352 However, if the real host name was known, it was omitting the HELO data
353 if it matched the actual IP address. This has been tidied up so that it
354 doesn't show the same IP address twice.
356 PH/07 When both +timestamp and +memory debugging was on, the value given by
357 $tod_xxx expansions could be wrong, because the tod_stamp() function was
358 called by the debug printing, thereby overwriting the timestamp buffer.
359 Debugging no longer uses the tod_stamp() function when +timestamp is set.
361 PH/08 When the original message was included in an autoreply transport, it
362 always said "this is a copy of the message, including all the headers",
363 even if body_only or headers_only was set. It now gives an appropriate
366 PH/09 Applied a patch from the Sieve maintainer which:
368 o fixes some comments
369 o adds the (disabled) notify extension core
370 o adds some debug output for the result of if/elsif tests
371 o points to the current vacation draft in the documentation
372 and documents the missing references header update
376 o fixes a bug in processing the envelope test (when testing
377 multiple envelope elements, the last element determinted the
380 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
381 Electronic Mail") by including:
383 Auto-submitted: auto-generated
385 in the messages that it generates (bounce messages and others, such as
386 warnings). In the case of bounce messages for non-SMTP mesages, there was
387 also a typo: it was using "Auto_submitted" (underscore instead of
388 hyphen). Since every message generated by Exim is necessarily in response
389 to another message, thes have all been changed to:
391 Auto-Submitted: auto-replied
393 in accordance with these statements in the RFC:
395 The auto-replied keyword:
397 - SHOULD be used on messages sent in direct response to another
398 message by an automatic process,
400 - MUST NOT be used on manually-generated messages,
402 - MAY be used on Delivery Status Notifications (DSNs) and Message
403 Disposition Notifications (MDNs),
405 - MUST NOT be used on messages generated by automatic or periodic
406 processes, except for messages which are automatic responses to
409 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
410 to the default Received: header definition.
412 PH/12 Added log selector acl_warn_skipped (default on).
414 PH/13 After a successful wildlsearch lookup, discard the values of numeric
415 variables because (a) they are in the wrong storage pool and (b) even if
416 they were copied, it wouldn't work properly because of the caching.
418 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
419 checking when decoding. Apparently there are clients that generate
420 overlong encoded strings. Why am I not surprised?
422 PH/15 If the first argument of "${if match_address" was not empty, but did not
423 contain an "@" character, Exim crashed. Now it writes a panic log message
424 and treats the condition as false.
426 PH/16 In autoreply, treat an empty string for "once" the same as unset.
428 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
429 extension "envelope-auth". The code is finished and in agreement with
430 other implementations, but there is no documentation so far and in fact,
431 nobody wrote the draft yet. This extension is currently #undef'ed, thus
432 not changing the active code.
434 Print executed "if" and "elsif" statements when debugging is used. This
435 helps a great deal to understand what a filter does.
437 Document more things not specified clearly in RFC3028. I had all this
438 sorted out, when out of a sudden new issues came to my mind. Oops."
440 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
443 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
444 canonical form (as documented). However, after a host name lookup from
445 the IP address, check_host() was doing a simple string comparison with
446 addresses acquired from the DNS when checking that the found name did
447 have the original IP as one of its addresses. Since any found IPv6
448 addresses are likely to be in abbreviated form, the comparison could
449 fail. Luckily, there already exists a function for doing the comparison
450 by converting both addresses to binary, so now that is used instead of
453 PH/20 There was another similar case to PH/19, when a complete host name was
454 given in a host list; looking up its IP address could give an abbreviated
455 form, whereas the current host's name might or might not be abbreviated.
456 The same fix has been applied.
462 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
463 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
466 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
467 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
469 PH/03 Typo: missing ".o" in src/pcre/Makefile.
471 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
472 header line, restrict the check to what is listed in RFCs 2369 and 2929.
473 Also, for "Auto-Submitted", treat anything other than "no" as
474 non-personal, in accordance with RFC 3834. (Previously it treated
475 anything starting "auto-" as non-personal.)
477 TF/01 The control=submission/name=... option had a problem with syntax
478 errors if the name included a slash character. The /name= option
479 now slurps the rest of the string, so it can include any characters
480 but it must come last in the list of options (after /sender_retain
483 PH/05 Some modifications to the interface to the fake nameserver for the new
491 TK/01 Added the "success_on_redirect" address verification option. See
492 NewStuff for rationale and an example.
494 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
496 PH/02 Patch to exigrep to allow it to work on syslog lines.
498 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
499 fread() to skip over the body file's header line, because in Cygwin the
500 header line is locked and is inaccessible.
502 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
503 co-exist for some time) to make it clear that it is the Exim ID that is
504 referenced, not the Message-ID: header line.
506 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
507 string_format() function, because snprintf() does not exist on all
510 PH/06 The use of forbid_filter_existstest now also locks out the use of the
511 ${stat: expansion item.
513 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
514 protocol synchronization error", to keep the pedants happy.
516 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
517 well as for IRIX systems, when gcc is being used. See the host.c source
520 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
522 PH/10 Named domain lists were not working if used in a queue_smtp_domains
525 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
526 transport and to the smtp transport in LMTP mode.
528 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
530 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
531 run a filter in a subprocess. This could lead to confusion in subsequent
532 lookups in the parent process. There should also be a search_tidyup() at
533 the end of the subprocess.
535 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
536 only if the host matched helo_try_verify_hosts, which caused the
537 verification to occur when the EHLO/HELO command was issued. The ACL just
538 tested the remembered result. Now, if a previous verification attempt has
539 not happened, "verify = helo" does it there and then.
541 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
543 TK/03 Fix log output including CR from clamd.
545 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
546 exist provoked a memory error which could cause a segfault.
548 PH/15 Installed PCRE 6.2
550 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
552 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
553 of the problem. Specifically, suggested +O2 rather than +O1 for the
556 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
558 PH/20 If a delivery was routed to a non-standard port by means of an SRV
559 record, the port was not correctly logged when the outgoing_port log
560 selector was set (it logged the transort's default port).
562 PH/21 Added support for host-specific ports to manualroute, queryprogram,
563 fallback_hosts, and "hosts" in the smtp transport.
565 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
566 host errors such as "Connection refused".
568 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
569 authentication with radiusclient 0.4.9:
571 - Error returned from rc_read_config was caught wrongly
572 - Username/password not passed on to radius server due to wrong length.
574 The presumption is that some radiusclient API changes for 4.51/PH/17
575 were not taken care of correctly. The code is still untested by me (my
576 Linux distribution still has 0.3.2 of radiusclient), but it was
577 contributed by a Radius user.
579 PH/24 When doing a callout, the value of $domain wasn't set correctly when
580 expanding the "port" option of the smtp transport.
582 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
583 while reading a MIME header. Thanks to Tom Hughes for a patch.
585 PH/24 Include config.h inside local_scan.h so that configuration settings are
588 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
589 that in an ACL for RCPT (for example), you can examine exactly what was
592 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
593 commands, but it was not correctly comparing the address with the actual
594 client host address. Thus, it would show the EHLO address in Received:
595 header lines when this was not necessary.
597 PH/27 Added the % operator to ${eval:}.
599 PH/28 Exim tries to create and chdir to its spool directory when it starts;
600 it should be ignoring failures (because with -C, for example, it has lost
601 privilege). It wasn't ignoring creation failures other than "already
604 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
607 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
608 because an input line was too long, either on its own, or by virtue of
609 too many continuations, the temporary file was not being removed, and the
610 return code was incorrect.
612 PH/31 Missing "BOOL" in function definition in filtertest.c.
614 PH/32 Applied Sieve patches from the maintainer.
616 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
618 PH/33 Added "verify = not_blind".
620 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
621 Local/Makefile (with some defaults set). These are used in built scripts
622 such as exicyclog, but they have never been used in the exim_install
623 script (though there are many overriding facilities there). I have
624 arranged that the exim_install script now takes note of these two
627 PH/35 Installed configuration files for Dragonfly.
629 PH/36 When a locally submitted message by a trusted user did not contain a
630 From: header, and the sender address was obtained from -f or from an SMTP
631 MAIL command, and the trusted user did not use -F to supply a sender
632 name, $originator_name was incorrectly used when constructing a From:
633 header. Furthermore, $originator_name was used for submission mode
634 messages from external hosts without From: headers in a similar way,
635 which is clearly wrong.
637 PH/37 Added control=suppress_local_fixups.
639 PH/38 When log_selector = +received_sender was set, and the addition of the
640 sender made the log line's construction buffer exactly full, or one byte
641 less than full, an overflow happened when the terminating "\n" was
644 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
645 when the result of a list match is failure because a DNS lookup failed.
647 PH/40 RM_COMMAND is now used in the building process.
649 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
650 the "build-* directories that it finds.
652 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
653 address in a domain literal was a prefix of an interface address.
655 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
656 when verifying a sender address, unless rewrite_headers is false.
658 PH/44 Wrote a long comment about why errors_to addresses are verified as
659 recipients, not senders.
661 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
662 the ratelimit ACL was added.
664 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
666 PH/46 Added extra information about PostgreSQL errors to the error string.
668 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
669 avoiding the need to install special zones in a real server. This is
670 backwards compatible; if it can't find the fake resolver, it drops back.
671 Thus, both old and new test suites can be run.
673 TF/02 Added util/ratelimit.pl
675 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
676 clock is set back in time.
678 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
679 Candler <B.Candler@pobox.com>.
681 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
682 OK for addresses that are the result of redirections.
684 PH/48 A number of further additions for the benefit of the new test suite,
685 including a fake gethostbyname() that interfaces to the fake DNS resolver
688 TF/06 The fix for widen_domains has also been applied to qualify_single and
689 search_parents which are the other dnslookup options that can cause
692 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
695 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
697 TF/07 Exim produced the error message "an SRV record indicated no SMTP
698 service" if it encountered an MX record with an empty target hostname.
699 The message is now "an MX or SRV record indicated no SMTP service".
701 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
702 if the DNS of the sending site is misconfigured. This is quite a
703 common situation. This change restores the behaviour of treating a
704 helo verification defer as a failure.
706 PH/51 If self=fail was set on a router, the bounce message did not include the
707 actual error message.
713 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
715 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
716 command itself ended in error, the underlying message about the transport
717 filter timeout was being overwritten with the pipe command error. Now the
718 underlying error message should be appended to the second error message.
720 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
722 PH/02 Reduce the amount of output that "make" produces by default. Full output
723 can still be requested.
725 PH/03 The warning log line about a condition test deferring for a "warn" verb
726 was being output only once per connection, rather than after each
727 occurrence (because it was using the same function as for successful
728 "warn" verbs). This seems wrong, so I have changed it.
730 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
731 it should not have, which might have caused a crash in the right
732 circumstances, but probably never did.
734 PH/04 Installed a modified version of Tony Finch's patch to make submission
735 mode fix the return path as well as the Sender: header line, and to
736 add a /name= option so that you can make the user's friendly name appear
739 TF/03 Added the control = fakedefer ACL modifier.
741 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
742 Mark Lowes for thorough testing.
744 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
746 TK/03 Merged latest SRS patch from Miles Wilton.
748 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
749 with the definition in sysexits.h (which is #included earlier).
750 Fortunately, Exim does not actually use EX_OK. The code used to try to
751 preserve the sysexits.h value, by assumimg that macro definitions were
752 scanned for macro replacements. I have been disabused of this notion,
753 so now the code just undefines EX_OK before #including unistd.h.
755 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
756 in the smtp transport. When a block could not be written in a single
757 write() function, the timeout was being re-applied to each part-write.
758 This seems wrong - if the receiver was accepting one byte at a time it
759 would take for ever. The timeout is now adjusted when this happens. It
760 doesn't have to be particularly precise.
762 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
763 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
765 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
766 without a domain if the check to <postmaster@domain> fails.
768 SC/01 Eximstats: added -xls and the ability to specify output files
769 (patch written by Frank Heydlauf).
771 SC/02 Eximstats: use FileHandles for outputing results.
773 SC/03 Eximstats: allow any combination of xls, txt, and html output.
775 SC/04 Eximstats: fixed display of large numbers with -nvr option
777 SC/05 Eximstats: fixed merging of reports with empty tables.
779 SC/06 Eximstats: added the -include_original_destination flag
781 SC/07 Eximstats: removed tabs and trailing whitespace.
783 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
785 TK/06 MBOX spool code: Add real "From " MBOX separator line
786 so the .eml file is really in mbox format (even though
787 most programs do not really care). Patch from Alex Miller.
789 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
790 The latter is generated from $received_to and is only set if the
791 message has one envelope recipient. SA can use these headers,
792 obviously out-of-the-box. Patch from Alex Miller.
794 PH/08 The ${def test on a variable was returning false if the variable's
795 value was "0", contrary to what the specification has always said!
796 The result should be true unless the variable is empty.
798 PH/09 The syntax error of a character other than { following "${if
799 def:variable_name" (after optional whitespace) was not being diagnosed.
800 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
801 accidental colon was present, for example, could give incorrect results.
803 PH/10 Tidied the code in a number of places where the st_size field of a stat()
804 result is used (not including appendfile, where other changes are about
807 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
808 This involved changing a lot of size variables from int to off_t. It
809 should work with maildirs and everything.
811 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
812 spamd dying while we are connected to it.
814 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
817 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
818 function for BDB 4.3.
820 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
822 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
823 the sources for PCRE have been re-arranged and more files are now
826 PH/15 The code I had for printing potentially long long variables in PH/11
827 above was not the best (it lost precision). The length of off_t variables
828 is now inspected at build time, and an appropriate printing format (%ld
829 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
830 to be "long long int" or "long int". This is needed for the internal
831 formatting function string_vformat().
833 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
834 the configuration file to be ":syslog", then the script "guesses" where
835 the logs files are, rather than using the compiled in default. In our
836 case the guess is not the same as the compiled default, so the script
837 suddenly stopped working when I started to use syslog. The patch checks
838 to see if log_file_path is "". If so, it attempts to read it from exim
839 with no configuration file to get the compiled in version, before it
840 falls back to the previous guessing code."
842 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
843 implementing BATV in an Exim configuration. See NewStuff for the gory
846 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
847 Makefile that are specific to HP-UX.
849 PH/18 If the "use_postmaster" option was set for a recipient callout together
850 with the "random" option, the postmaster address was used as the MAIL
851 FROM address for the random test, but not for the subsequent recipient
852 test. It is now used for both.
854 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
855 patch removes a few documentation additions to RFC 3028, because the
856 latest draft now contains them. It adds the new en;ascii-case comparator
857 and a new error check for 8bit text in MIME parts. Comparator and
858 require names are now matched exactly. I enabled the subaddress
859 extension, but it is not well tested yet (read: it works for me)."
861 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
862 rework some of the code of TK/09 above to avoid the hardwired use of
863 "%lld" and "long long". Replaced the call to snprintf() with a call to
866 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
867 records point to non-existent hosts", "retry timeout exceeded", and
868 "retry time not reached for any host after a long failure period".
870 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
871 experimental DomainKeys support:
873 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
874 (2) On an error such as an illegally used "control", the wrong name for
875 the control was given.
877 These problems did NOT occur unless DomainKeys support was compiled.
879 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
881 PH/24 Added ${if match_ip condition.
883 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
884 databases so that it will be absolutely obvious if a crash occurs in the
885 DB library. This is a regular occurrence (often caused by mis-matched
888 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
889 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
890 picked up on a user's system that detects such things. There doesn't seem
891 to be a gcc warning option for this - only an attribute that has to be
892 put on the function's prototype. It seems that in Fedora Core 4 they have
893 set this on a number of new functions. No doubt there will be more in due
896 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
897 specify a transport. However, if an address that was verified by such a
898 router was the subject of a callout, Exim crashed because it tried to
899 read the rcpt_include_affixes from the non-existent transport. Now it
900 just assumes that the setting of that option is false. This bug was
901 introduced by 4.51/PH/31.
903 PH/28 Changed -d+all to exclude +memory, because that information is very
904 rarely of interest, but it makes the output a lot bigger. People tend to
905 do -d+all out of habit.
907 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
908 code in os-type was giving problems when libc.so lives in lib64, like on
911 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
912 aren't the modern standard, and it seems that some systems' include files
913 don't always have them. Exim was already checking for some of the newer
914 ones like T_AAAA, and defining it itself. I've added checks for all the
915 record types that Exim uses.
917 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
918 not automatically generating a new one, as it is supposed to. This
919 prevented TLS from working. If the file did exist, but contained invalid
920 data, a new version was generated, as expected. It was only the case of a
921 non-existent file that was broken.
923 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
924 with a change in libdomainkeys > 0.64.
926 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
927 from DNS. If the selector record carries the flag, it now has
928 precedence over the domain-wide flag.
930 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
932 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
933 the use of a port name that isn't defined in /etc/services) occurred, the
934 message was deferred as in a normal delivery, and thus remained on the
935 spool, instead of being failed because of the mua_wrapper setting. This
936 is now fixed, and I tidied up some of the mua_wrapper messages at the
939 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
940 the messages in a hash of arrays rather than using individual hashes.
941 This is a bit cleaner and results in dramatic memory savings, albeit
942 at a slight CPU cost.
944 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
945 as requested by Marc Sherman.
947 SC/10 Eximstats: added histograms for user specified patterns as requested
950 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
952 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
953 fopen() in the content-scanning modules that did not already have it.
959 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
960 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
962 TK/02 Fix ACL "control" statement not being available in MIME ACL.
964 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
966 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
967 to test Sieve filters that use "vacation".
969 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
970 that changes the way the GnuTLS parameters are stored in the cache file.
971 The new format can be generated externally. For backward compatibility,
972 if the data in the cache doesn't make sense, Exim assumes it has read an
973 old-format file, and it generates new data and writes a new file. This
974 means that you can't go back to an older release without removing the
977 PH/03 A redirect router that has both "unseen" and "one_time" set does not
978 work if there are any delivery delays because "one_time" forces the
979 parent to be marked "delivered", so its unseen clone is never tried
980 again. For this reason, Exim now forbids the simultaneous setting of
983 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
984 redirected to themselves ("homonym" addresses). Read the long ChangeLog
985 entry if you want to know the details. The fix, however, neglected to
986 consider the case when local delivery batching is involved. The test for
987 "previously delivered" was not happening when checking to see if an
988 address could be batched with a previous (undelivered) one; under
989 certain circumstances this could lead to multiple deliveries to the same
992 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
993 in its include files, and this causes problems building Exim.
995 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
996 header_syntax) but Exim was just ignoring anything given after a slash.
997 In particular, this caused confusion with an attempt to use "verify =
998 reverse_host_lookup/defer_ok". An error is now given when options are
999 supplied for verify items that do not have them. (Maybe reverse_host_
1000 lookup should have a defer_ok option, but that's a different point.)
1002 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1003 defined by RFC 821) to 2048, because there were problems with some AUTH
1004 commands, and RFC 1869 says the size should be increased for extended
1005 SMTP commands that take arguments.
1007 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1010 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1011 "unknown" error; now it says that the functionality isn't in the binary.
1013 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1014 an address' error message when a string expansion fails (syntax or
1015 whatever). Otherwise the password may appear in the log. Following change
1016 PH/42 below, there is no longer a chance of it appearing in a bounce
1019 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1021 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1022 of its addresses was ever tried. (Bugzilla bug #2.)
1024 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1025 the result incorrectly in the debug output. (It correctly added a newline
1026 to what was transported.)
1028 TF/01 Added $received_time.
1030 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1031 commented out examples of how to interface to a virus scanner and to
1032 SpamAssassin. Also added commented examples of av_scanner and
1033 spamd_address settings.
1035 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1036 and controls are allowed in which ACLs. There were a couple of minor
1037 errors. Some of the entries in the conditions table (which is a table of
1038 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1039 negation of where the condition IS allowed.
1041 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1043 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1044 header file does not have a version number, so I've had to invent a new
1045 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1046 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1047 radiusclient), but it was contributed by a Radius user.
1049 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1050 files or directories, for OpenSSL.
1052 PH/19 When an Exim process that is running as root has to create an Exim log
1053 file, it does so in a subprocess that runs as exim:exim so as to get the
1054 ownership right at creation (otherwise, other Exim processes might see
1055 the file with the wrong ownership). There was no test for failure of this
1056 fork() call, which would lead to the process getting stuck as it waited
1057 for a non-existent subprocess. Forks do occasionally fail when resources
1058 run out. I reviewed all the other calls to fork(); they all seem to check
1061 PH/20 When checking for unexpected SMTP input at connect time (before writing
1062 the banner), Exim was not dealing correctly with a non-positive return
1063 from the read() function. If the client had disconnected by this time,
1064 the result was a log entry for a synchronization error with an empty
1065 string after "input=" when read() returned zero. If read() returned -1
1066 (an event I could not check), uninitialized data bytes were printed.
1067 There were reports of junk text (parts of files, etc) appearing after
1070 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1072 PH/22 Added support for macro redefinition, and (re)definition in between
1073 driver and ACL definitions.
1075 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1076 forgetting to use the resulting value; it was using the unexpanded value.
1078 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1079 hadn't been configured. The fix is from Juergen Kreileder, who
1080 understands it better than I do:
1082 "Here's what I see happening with three configured cyrus_sasl
1083 authenticators configured (plain, login, cram-md5):
1085 On startup auth_cyrus_sasl_init() gets called for each of these.
1086 This means three calls to sasl_listmech() without a specified mech_list.
1087 => SASL tests which mechs of all available mechs actually work
1088 => three warnings about OTP not working
1089 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1091 With the patch, sasl_listmech() also gets called three times. But now
1092 SASL's mech_list option is set to the server_mech specified in the the
1093 authenticator. Or in other words, the answer from sasl_listmech()
1094 gets limited to just the mech you're testing for (which is different
1096 => the return list contains just 'plain' or 'login', 'cram-md5' or
1097 nothing depending on the value of ob->server_mech.
1099 I've just tested the patch: Authentication still works fine,
1100 unavailable mechs specified in the exim configuration are still
1101 caught, and the auth.log warnings about OTP are gone."
1103 PH/25 When debugging is enabled, the contents of the command line are added
1104 to the debugging output, even when log_selector=+arguments is not
1107 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1108 answer is "GNU", and only if the return is "GNU/something" is the answer
1111 PH/27 $acl_verify_message is now set immediately after the failure of a
1112 verification in an ACL, and so is available in subsequent modifiers. In
1113 particular, the message can be preserved by coding like this:
1115 warn !verify = sender
1116 set acl_m0 = $acl_verify_message
1118 Previously, $acl_verify_message was set only while expanding "message"
1119 and "log_message" when a very denied access.
1121 PH/28 Modified OS/os.c-Linux with
1123 -#ifndef OS_LOAD_AVERAGE
1124 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1126 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1127 nomenclature these days.)
1129 PH/29 Installed patch from the Sieve maintainer that adds the options
1130 sieve_useraddress and sieve_subaddress to the redirect router.
1132 PH/30 In these circumstances:
1133 . Two addresses routed to the same list of hosts;
1134 . First host does not offer TLS;
1135 . First host accepts first address;
1136 . First host gives temporary error to second address;
1137 . Second host offers TLS and a TLS session is established;
1138 . Second host accepts second address.
1139 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1140 and peerdn, if requested) that were in fact used only for the second
1143 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1144 attention to any local part prefix or suffix that was matched by the
1145 router that accepted the address. It now behaves in the same way as it
1146 does for delivery: the affixes are removed from the local part unless
1147 rcpt_include_affixes is set on the transport.
1149 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1150 timeout during the DATA phase of an incoming message.
1152 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1153 applied a patch sanctioned by the Sieve maintainer.
1155 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1156 the uid or gid is negative. A case of a negative gid caused this to be
1157 noticed. The fix allows for either to be negative.
1159 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1160 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1163 PH/36 The change PH/12 above was broken. Fixed it.
1165 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1166 the grounds that routing the same address twice would always produce the
1167 same answer. This might have been true once, but it is certainly no
1168 longer true now. Routing a child address may depend on the previous
1169 routing that produced that child. Some complicated redirection strategies
1170 went wrong when messages had multiple recipients, and made Exim's
1171 behaviour dependent on the order in which the addresses were given.
1173 I have moved the duplicate checking until after the routing is complete.
1174 Exim scans the addresses that are assigned to local and remote
1175 transports, and removes any duplicates. This means that more work will be
1176 done, as duplicates will always all be routed, but duplicates are
1177 presumably rare, so I don't expect this is of any significance.
1179 For deliveries to pipes, files, and autoreplies, the duplicate checking
1180 still happens during the routing process, since they are not going to be
1183 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1184 It corrects a timeout issue with spamd. This is Ian's comment: "The
1185 background is that sometimes spamd either never reads data from a
1186 connection it has accepted, or it never writes response data. The exiscan
1187 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1188 blindly assumes that writes won't block so it may never time out."
1190 PH/39 Allow G after quota size as well as K and M.
1192 PH/40 The value set for $authenticated_id in an authenticator may not contain
1193 binary zeroes or newlines because the value is written to log lines and
1194 to spool files. There was no check on this. Now the value is run through
1195 the string_printing() function so that such characters are converted to
1196 printable escape sequences.
1198 PH/41 $message_linecount is a new variable that contains the total number of
1199 lines in the message. Compare $body_linecount, which is the count for the
1202 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1203 bounce and delay warning messages, except in certain special cases, which
1206 (a) An SMTP error message from a remote host;
1207 (b) A message specified in a :fail: redirection;
1208 (c) A message specified in a "fail" command in a system filter;
1209 (d) A message specified in a FAIL return from the queryprogram router;
1210 (e) A message specified by the cannot_route_message router option.
1212 In these cases only, Exim does include the error details in bounce and
1213 warning messages. There are also a few cases where bland messages such
1214 as "unrouteable address" or "local delivery error" are given.
1216 PH/43 $value is now also set for the "else" part of a ${run expansion.
1218 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1219 being worked on, but at least Exim now implements the latest version to
1222 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1223 process to complete was treated as a delivery failure, a timeout while
1224 writing the message to the pipe was logged, but erroneously treated as a
1225 successful delivery. Such timeouts include transport filter timeouts. For
1226 consistency with the overall process timeout, these timeouts are now
1227 treated as errors, giving rise to delivery failures by default. However,
1228 there is now a new Boolean option for the pipe transport called
1229 timeout_defer, which, if set TRUE, converts the failures into defers for
1230 both kinds of timeout. A transport filter timeout is now identified in
1233 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1234 systems where "make" and "gmake" are different, calling "gmake" at top
1235 level broke things. I've arranged for the value of $(MAKE) to be passed
1236 from the Makefile to this script so that it can call the same version of
1240 A note about Exim versions 4.44 and 4.50
1241 ----------------------------------------
1243 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1244 changes of various kinds. As a consequence, a big documentation update was
1245 needed. This delayed the release for rather longer than seemed good, especially
1246 in the light of a couple of (minor) security issues. Therefore, the changes
1247 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1248 release. So 4.44 and 4.50 are in effect two different branches that both start
1251 I have left the 4.50 change log unchanged; it contains all the changes since
1252 4.43. The change log for 4.44 is below; many of its items are identical to
1253 those for 4.50. This seems to be the most sensible way to preserve the
1254 historical information.
1260 1. Minor wording change to the doc/README.SIEVE file.
1262 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1263 computation of the current number of files was incorrect.
1265 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1266 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1269 4. Give more explanation in the error message when the command for a transport
1270 filter fails to execute.
1272 5. There are several places where Exim runs a non-Exim command in a
1273 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1274 was being done only for the command run by the queryprogram router. It is
1275 now done for all such subprocesses. The other cases are: ${run, transport
1276 filters, and the commands run by the lmtp and pipe transports.
1278 6. Added CONFIGURE_GROUP build-time option.
1280 7. Some older OS have a limit of 256 on the maximum number of file
1281 descriptors. Exim was using setrlimit() to set 1000 as a large value
1282 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1283 systems. I've change it so that if it can't get 1000, it tries for 256.
1285 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1286 was an oversight, and furthermore, ever since the addition of extra
1287 controls (e.g. 4.43/32), the checks on when to allow different forms of
1288 "control" were broken. There should now be diagnostics for all cases when a
1289 control that does not make sense is encountered.
1291 9. Added the /retain_sender option to "control=submission".
1293 10. $recipients is now available in the predata ACL (oversight).
1295 11. Tidy the search cache before the fork to do a delivery from a message
1296 received from the command line. Otherwise the child will trigger a lookup
1297 failure and thereby defer the delivery if it tries to use (for example) a
1298 cached ldap connection that the parent has called unbind on.
1300 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1301 of $address_data from the recipient verification was clobbered by the
1302 sender verification.
1304 13. The value of address_data from a sender verification is now available in
1305 $sender_address_data in subsequent conditions in the ACL statement.
1307 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1309 15. Added a new option "connect=<time>" to callout options, to set a different
1312 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1313 was its contents. (It was OK if the option was not defined at all.)
1315 17. A "Completed" log line is now written for messages that are removed from
1316 the spool by the -Mrm option.
1318 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1319 information about exactly what failed.
1321 19. Added -dd to debug only the daemon process.
1323 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1324 handles timeouts, both on the server side and network timeouts. Renamed the
1325 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1327 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1328 It is now set to "smtps".
1330 22. $host_address is now set to the target address during the checking of
1331 ignore_target_hosts.
1333 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1334 being passed; this would have caused $sender_host_name to have been used if
1335 matching the list had actually called for a host name (not very likely,
1336 since this list is usually IP addresses). A host name is now passed as
1339 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1340 code that specifies a non-restarting handler (typically sigaction() in
1341 modern systems) in an attempt to fix a rare and obscure crash bug.
1343 25. Narrowed the window for a race in the daemon that could cause it to ignore
1344 SIGCHLD signals. This is not a major problem, because they are used only to
1345 wake it up if nothing else does.
1347 26. A malformed maildirsize file could cause Exim to calculate negative values
1348 for the mailbox size or file count. Odd effects could occur as a result.
1349 The maildirsize information is now recalculated if the size or filecount
1352 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1353 support for a long time. Removed HAVE_SYS_VFS_H.
1355 28. Installed the latest version of exipick from John Jetmore.
1357 29. In an address list, if the pattern was not a regular expression, an empty
1358 subject address (from a bounce message) matched only if the pattern was an
1359 empty string. Non-empty patterns were not even tested. This was the wrong
1360 because it is perfectly reasonable to use an empty address as part of a
1361 database query. An empty address is now tested by patterns that are
1362 lookups. However, all the other forms of pattern expect the subject to
1363 contain a local part and a domain, and therefore, for them, an empty
1364 address still always fails if the pattern is not itself empty.
1366 30. Exim went into a mad DNS loop when attempting to do a callout where the
1367 host was specified on an smtp transport, and looking it up yielded more
1368 than one IP address.
1370 31. Re-factored the code for checking spool and log partition space into a
1371 function that finds that data and another that does the check. The former
1372 is then used to implement four new variables: $spool_space, $log_space,
1373 $spool_inodes, and $log_inodes.
1375 32. The RFC2047 encoding function was originally intended for short strings
1376 such as real names; it was not keeping to the 75-character limit for
1377 encoded words that the RFC imposes. It now respects the limit, and
1378 generates multiple encoded words if necessary. To be on the safe side, I
1379 have increased the buffer size for the ${rfc2047: expansion operator from
1382 33. It is now permitted to omit both strings after an "if" condition; if the
1383 condition is true, the result is "true". As before, when the second string
1384 is omitted, a false condition yields an empty string. This makes it less
1385 cumbersome to write custom ACL and router conditions.
1387 34. Failure to deliver a bounce message always caused it to be frozen, even if
1388 there was an errors_to setting on the router. The errors_to setting is now
1391 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1392 canonical form (fully expanded) before being placed in
1393 $sender_host_address.
1395 36. The table in the code that translates DNS record types into text (T_A to
1396 "A" for instance) was missing entries for NS and CNAME. It is just possible
1397 that this could have caused confusion if both these types were looked up
1398 for the same domain, because the text type is used as part of Exim's
1399 per-process caching. But the chance of anyone hitting this buglet seems
1402 37. The dnsdb lookup has been extended in a number of ways.
1404 (1) There is a new type, "zns", which walks up the domain tree until it
1405 finds some nameserver records. It should be used with care.
1407 (2) There is a new type, "mxh", which is like "mx" except that it returns
1408 just the host names, not the priorities.
1410 (3) It is now possible to give a list of domains (or IP addresses) to be
1411 looked up. The behaviour when one of the lookups defers can be
1412 controlled by a keyword.
1414 (4) It is now possible to specify the separator character for use when
1415 multiple records are returned.
1417 38. The dnslists ACL condition has been extended: it is now possible to supply
1418 a list of IP addresses and/or domains to be looked up in a particular DNS
1421 39. Added log_selector=+queue_time_overall.
1423 40. When running the queue in the test harness, wait just a tad after forking a
1424 delivery process, to get repeatability of debugging output.
1426 41. Include certificate and key file names in error message when GnuTLS fails
1427 to set them up, because the GnuTLS error message doesn't include the name
1428 of the failing file when there is a problem reading it.
1430 42. Allow both -bf and -bF in the same test run.
1432 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1434 44. The "Exiscan patch" is now merged into the mainline Exim source.
1436 45. Sometimes the final signoff response after QUIT could fail to get
1437 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1438 < 0 before doing a fflush(). This bug looks as though it goes back to the
1439 introduction of TLS in release 3.20, but "sometimes" must have been rare
1440 because the tests only now provoked it.
1442 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1443 (this can affect the format of dates).
1445 47. exim_tidydb, when checking for the continued existence of a message for
1446 which it has found a message-specific retry record, was not finding
1447 messages that were in split spool directories. Consequently, it was
1448 deleting retry records that should have stayed in existence.
1450 48. Steve fixed some bugs in eximstats.
1452 49. The SPA authentication driver was not abandoning authentication and moving
1453 on to the next authenticator when an expansion was forced to fail,
1454 contradicting the general specification for all authenticators. Instead it
1455 was generating a temporary error. It now behaves as specified.
1457 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1458 (the order specifies the preference for clients). The order is now AES256,
1459 AES128, 3DES, ARCFOUR128.
1461 51. Small patch to Sieve code - explicitly set From: when generating an
1464 52. Exim crashed if a remote delivery caused a very long error message to be
1465 recorded - for instance if somebody sent an entire SpamAssassin report back
1466 as a large number of 550 error lines. This bug was coincidentally fixed by
1467 increasing the size of one of Exim's internal buffers (big_buffer) that
1468 happened as part of the Exiscan merge. However, to be on the safe side, I
1469 have made the code more robust (and fixed the comments that describe what
1472 53. Now that there can be additional text after "Completed" in log lines (if
1473 the queue_time_overall log selector is set), a one-byte patch to exigrep
1474 was needed to allow it to recognize "Completed" as not the last thing in
1477 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1478 patch that reportedly fixes this has been added. I am not expert enough to
1479 create a test for it. This is what the patch creator wrote:
1481 "I found a little strange behaviour of ldap code when working with
1482 Windows 2003 AD Domain, where users was placed in more than one
1483 Organization Units. When I tried to give exim partial DN, the exit code
1484 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1485 But simultaneously result of request was absolutely normal ldap result,
1486 so I produce this patch..."
1488 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1489 REFERENCE, so I have modified the code to exclude the patch when that macro
1492 55. Some experimental protocols are using DNS PTR records for new purposes. The
1493 keys for these records are domain names, not reversed IP addresses. The
1494 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1495 leaves it alone. Component reversal etc. now happens only for IP addresses.
1498 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1500 57. Double the size of the debug message buffer (to 2048) so that more of very
1501 long debug lines gets shown.
1503 58. The exicyclog utility now does better if the number of log files to keep
1504 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1506 59. Two changes related to the smtp_active_hostname option:
1508 (1) $smtp_active_hostname is now available as a variable.
1509 (2) The default for smtp_banner uses $smtp_active_hostname instead
1510 of $primary_hostname.
1512 60. The host_aton() function is supposed to be passed a string that is known
1513 to be a valid IP address. However, in the case of IPv6 addresses, it was
1514 not checking this. This is a hostage to fortune. Exim now panics and dies
1515 if the condition is not met. A case was found where this could be provoked
1516 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1517 components; fortuitously, this particular loophole had already been fixed
1518 by change 4.50/55 above.
1520 If there are any other similar loopholes, the new check in host_aton()
1521 itself should stop them being exploited. The report I received stated that
1522 data on the command line could provoke the exploit when Exim was running as
1523 exim, but did not say which command line option was involved. All I could
1524 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1525 running as the user.
1528 61. There was a buffer overflow vulnerability in the SPA authentication code
1529 (which came originally from the Samba project). I have added a test to the
1530 spa_base64_to_bits() function which I hope fixes it.
1533 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1534 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1536 63. The daemon start-up calls getloadavg() while still root for those OS that
1537 need the first call to be done as root, but it missed one case: when
1538 deliver_queue_load_max is set with deliver_drop_privilege. This is
1539 necessary for the benefit of the queue runner, because there is no re-exec
1540 when deliver_drop_privilege is set.
1542 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1543 This has been fixed.
1545 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1546 was in use, was not putting the data itself into the right store pool;
1547 consequently, it could be overwritten for a subsequent message in the same
1548 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1551 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1553 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1554 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1555 the function were treating the return as a boolean value, which happened to
1556 work because 0=false and not-0=true, but is not correct code.
1558 68. The host_aton() function was not handling scoped IPv6 addresses (those
1559 with, for example, "%eth0" on the end) correctly.
1561 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1562 negated items (that is, ~something) in unsigned ints. Some compilers
1563 apparently mutter when there is no cast.
1565 70. If an address verification called from an ACL failed, and did not produce a
1566 user-specific message (i.e. there was only a "system" message), nothing was
1567 put in $acl_verify_message. In this situation, it now puts the system
1570 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1571 session; change 4.31/43 added the unwanted input to the log line - except
1572 that it did not do this in the start of session case. It now does.
1574 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1575 This is wrong and can cause the other end to generate a synchronization
1576 error if it is another Exim or anything else that does the synchronization
1577 check. A QUIT command is no longer sent after a timeout.
1579 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1580 during host lookups.
1582 74. The defer_ok option of callout verification was not working if it was used
1583 when verifying addresses in header lines, that is, for this case:
1585 verify = header_sender/callout=defer_ok
1587 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1588 those file descriptors could be used for SMTP connections. If anything
1589 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1590 could be sent to the SMTP client, causing chaos. The daemon now opens
1591 stdin, stdout, and stderr to /dev/null when it puts itself into the
1594 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1595 log by default. The user can override this with suitable Perl magic.
1597 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1598 the log message when discard triggers, was not working for the DATA ACL or
1599 for the non-SMTP ACL.
1601 78. Error message wording change in sieve.c.
1603 79. If smtp_accept_max_per_host was set, the number of connections could be
1604 restricted to fewer than expected, because the daemon was trying to set up
1605 a new connection before checking whether the processes handling previous
1606 connections had finished. The check for completed processes is now done
1607 earlier. On busy systems, this bug wouldn't be noticed because something
1608 else would have woken the daemon, and it would have reaped the completed
1611 80. If a message was submitted locally by a user whose login name contained one
1612 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1613 It caused a spool format error. I have fixed the spool reading code. A
1614 related problem was that the "from" clause in the Received: line became
1615 illegal because of the space(s). It is now covered by ${quote_local_part.
1617 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1620 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1622 83. Patch from Sieve maintainer to fix unterminated string problem in
1623 "vacation" handling.
1625 84. Some minor changes to the Linux configuration files to help with other
1626 OS variants using glibc.
1628 85. One more patch for Sieve to update vacation handling to latest spec.
1631 ----------------------------------------------------
1632 See the note above about the 4.44 and 4.50 releases.
1633 ----------------------------------------------------
1639 1. Change 4.43/35 introduced a bug that caused file counts to be
1640 incorrectly computed when quota_filecount was set in an appendfile
1643 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1644 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1647 3. Give more explanation in the error message when the command for a transport
1648 filter fails to execute.
1650 4. There are several places where Exim runs a non-Exim command in a
1651 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1652 was being done only for the command run by the queryprogram router. It is
1653 now done for all such subprocesses. The other cases are: ${run, transport
1654 filters, and the commands run by the lmtp and pipe transports.
1656 5. Some older OS have a limit of 256 on the maximum number of file
1657 descriptors. Exim was using setrlimit() to set 1000 as a large value
1658 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1659 systems. I've change it so that if it can't get 1000, it tries for 256.
1661 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1662 was an oversight, and furthermore, ever since the addition of extra
1663 controls (e.g. 4.43/32), the checks on when to allow different forms of
1664 "control" were broken. There should now be diagnostics for all cases when a
1665 control that does not make sense is encountered.
1667 7. $recipients is now available in the predata ACL (oversight).
1669 8. Tidy the search cache before the fork to do a delivery from a message
1670 received from the command line. Otherwise the child will trigger a lookup
1671 failure and thereby defer the delivery if it tries to use (for example) a
1672 cached ldap connection that the parent has called unbind on.
1674 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1675 of $address_data from the recipient verification was clobbered by the
1676 sender verification.
1678 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1679 was its contents. (It was OK if the option was not defined at all.)
1681 11. A "Completed" log line is now written for messages that are removed from
1682 the spool by the -Mrm option.
1684 12. $host_address is now set to the target address during the checking of
1685 ignore_target_hosts.
1687 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1688 being passed; this would have caused $sender_host_name to have been used if
1689 matching the list had actually called for a host name (not very likely,
1690 since this list is usually IP addresses). A host name is now passed as
1693 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1694 code that specifies a non-restarting handler (typically sigaction() in
1695 modern systems) in an attempt to fix a rare and obscure crash bug.
1697 15. Narrowed the window for a race in the daemon that could cause it to ignore
1698 SIGCHLD signals. This is not a major problem, because they are used only to
1699 wake it up if nothing else does.
1701 16. A malformed maildirsize file could cause Exim to calculate negative values
1702 for the mailbox size or file count. Odd effects could occur as a result.
1703 The maildirsize information is now recalculated if the size or filecount
1706 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1707 support for a long time. Removed HAVE_SYS_VFS_H.
1709 18. Updated exipick to current release from John Jetmore.
1711 19. Allow an empty sender to be matched against a lookup in an address list.
1712 Previously the only cases considered were a regular expression, or an
1715 20. Exim went into a mad DNS lookup loop when doing a callout where the
1716 host was specified on the transport, if the DNS lookup yielded more than
1719 21. The RFC2047 encoding function was originally intended for short strings
1720 such as real names; it was not keeping to the 75-character limit for
1721 encoded words that the RFC imposes. It now respects the limit, and
1722 generates multiple encoded words if necessary. To be on the safe side, I
1723 have increased the buffer size for the ${rfc2047: expansion operator from
1726 22. Failure to deliver a bounce message always caused it to be frozen, even if
1727 there was an errors_to setting on the router. The errors_to setting is now
1730 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1731 canonical form (fully expanded) before being placed in
1732 $sender_host_address.
1734 24. Updated eximstats to version 1.33
1736 25. Include certificate and key file names in error message when GnuTLS fails
1737 to set them up, because the GnuTLS error message doesn't include the name
1738 of the failing file when there is a problem reading it.
1740 26. Expand error message when OpenSSL has problems setting up cert/key files.
1743 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1744 (this can affect the format of dates).
1746 28. exim_tidydb, when checking for the continued existence of a message for
1747 which it has found a message-specific retry record, was not finding
1748 messages that were in split spool directories. Consequently, it was
1749 deleting retry records that should have stayed in existence.
1751 29. eximstats updated to version 1.35
1752 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1753 1.35 - bugfix such that pie charts by volume are generated correctly
1755 30. The SPA authentication driver was not abandoning authentication and moving
1756 on to the next authenticator when an expansion was forced to fail,
1757 contradicting the general specification for all authenticators. Instead it
1758 was generating a temporary error. It now behaves as specified.
1760 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1761 (the order specifies the preference for clients). The order is now AES256,
1762 AES128, 3DES, ARCFOUR128.
1764 31. Small patch to Sieve code - explicitly set From: when generating an
1767 32. Exim crashed if a remote delivery caused a very long error message to be
1768 recorded - for instance if somebody sent an entire SpamAssassin report back
1769 as a large number of 550 error lines. This bug was coincidentally fixed by
1770 increasing the size of one of Exim's internal buffers (big_buffer) that
1771 happened as part of the Exiscan merge. However, to be on the safe side, I
1772 have made the code more robust (and fixed the comments that describe what
1775 33. Some experimental protocols are using DNS PTR records for new purposes. The
1776 keys for these records are domain names, not reversed IP addresses. The
1777 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1778 leaves it alone. Component reversal etc. now happens only for IP addresses.
1781 34. The host_aton() function is supposed to be passed a string that is known
1782 to be a valid IP address. However, in the case of IPv6 addresses, it was
1783 not checking this. This is a hostage to fortune. Exim now panics and dies
1784 if the condition is not met. A case was found where this could be provoked
1785 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1786 components; fortuitously, this particular loophole had already been fixed
1787 by change 4.50/55 or 4.44/33 above.
1789 If there are any other similar loopholes, the new check in host_aton()
1790 itself should stop them being exploited. The report I received stated that
1791 data on the command line could provoke the exploit when Exim was running as
1792 exim, but did not say which command line option was involved. All I could
1793 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1794 running as the user.
1797 35. There was a buffer overflow vulnerability in the SPA authentication code
1798 (which came originally from the Samba project). I have added a test to the
1799 spa_base64_to_bits() function which I hope fixes it.
1802 36. The daemon start-up calls getloadavg() while still root for those OS that
1803 need the first call to be done as root, but it missed one case: when
1804 deliver_queue_load_max is set with deliver_drop_privilege. This is
1805 necessary for the benefit of the queue runner, because there is no re-exec
1806 when deliver_drop_privilege is set.
1808 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1809 was in use, was not putting the data itself into the right store pool;
1810 consequently, it could be overwritten for a subsequent message in the same
1811 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1814 38. Sometimes the final signoff response after QUIT could fail to get
1815 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1816 < 0 before doing a fflush(). This bug looks as though it goes back to the
1817 introduction of TLS in release 3.20, but "sometimes" must have been rare
1818 because the tests only now provoked it.
1824 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1825 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1826 Now it returns an integer. A number of calls were still expecting a T/F
1827 return. Fortuitously, in all cases, the tests worked in OK situations,
1828 which is the norm. However, things would have gone wrong on any write
1829 failures on the smtp file descriptor. This function is used when sending
1830 messages over SMTP and also when doing verify callouts.
1832 2. When Exim is called to do synchronous delivery of a locally submitted
1833 message (the -odf or -odi options), it no longer closes stderr before doing
1836 3. Implemented the mua_wrapper option.
1838 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1840 5. Implemented the functions header_remove(), header_testname(),
1841 header_add_at_position(), and receive_remove_recipient(), and exported them
1844 6. If an ACL "warn" statement specified the addition of headers, Exim already
1845 inserted X-ACL-Warn: at the start if there was no header name. However, it
1846 was not making this test for the second and subsequent header lines if
1847 there were newlines in the string. This meant that an invalid header could
1848 be inserted if Exim was badly configured.
1850 7. Allow an ACL "warn" statement to add header lines at the start or after all
1851 the Received: headers, as well as at the end.
1853 8. Added the rcpt_4xx retry error code.
1855 9. Added postmaster_mailfrom=xxx to callout verification option.
1857 10. Added mailfrom=xxxx to the callout verification option, for verify=
1860 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1861 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1863 12. Inserted some casts to stop certain compilers warning when using pointer
1864 differences as field lengths or precisions in printf-type calls (mostly
1865 affecting debugging statements).
1867 13. Added optional readline() support for -be (dynamically loaded).
1869 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1870 same clock tick as a message's arrival, so that its received time was the
1871 same as the "first fail" time on the retry record, and that message
1872 remained on the queue past the ultimate address timeout, every queue runner
1873 would try a delivery (because it was past the ultimate address timeout) but
1874 after another failure, the ultimate address timeout, which should have then
1875 bounced the address, did not kick in. This was a "< instead of <=" error;
1876 in most cases the first failure would have been in the next clock tick
1877 after the received time, and all would be well.
1879 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1880 being recognized when the domain list was tested by the match_domain
1881 condition in an expansion string.
1883 16. Added the ${str2b64: operator.
1885 17. Exim was always calling setrlimit() to set a large limit for the number of
1886 processes, without checking whether the existing limit was already
1887 adequate. (It did check for the limit on file descriptors.) Furthermore,
1888 errors from getrlimit() and setrlimit() were being ignored. Now they are
1889 logged to the main and panic logs, but Exim does carry on, to try to do its
1890 job under whatever limits there are.
1892 18. Imported PCRE 5.0.
1894 19. Trivial typo in log message " temporarily refused connection" (the leading
1897 20. If the log selector return_path_on_delivery was set and an address was
1898 redirected to /dev/null, the delivery process crashed because it assumed
1899 that a return path would always be set for a "successful" delivery. In this
1900 case, the whole delivery is bypassed as an optimization, and therefore no
1903 21. Internal re-arrangement: the function for sending a challenge and reading
1904 a response while authentication was assuming a zero-terminated challenge
1905 string. It's now changed to take a pointer and a length, to allow for
1906 binary data in such strings.
1908 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1910 23. Exim was not respecting finduser_retries when seeking the login of the
1911 uid under which it was called; it was always trying 10 times. (The default
1912 setting of finduser_retries is zero.) Also, it was sleeping after the final
1913 failure, which is pointless.
1915 24. Implemented tls_on_connect_ports.
1917 25. Implemented acl_smtp_predata.
1919 26. If the domain in control=submission is set empty, Exim assumes that the
1920 authenticated id is a complete email address when it generates From: or
1921 Sender: header lines.
1923 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1924 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1925 chown and chgrp in /bin and hostname in /usr/bin.
1927 28. Exim was keeping the "process log" file open after each use, just as it
1928 does for the main log. This opens the possibility of it remaining open for
1929 long periods when the USR1 signal hits a daemon. Occasional processlog
1930 errors were reported, that could have been caused by this. Anyway, it seems
1931 much more sensible not to leave this file open at all, so that is what now
1934 29. The long-running daemon process does not normally write to the log once it
1935 has entered its main loop, and it closes the log before doing so. This is
1936 so that log files can straightforwardly be renamed and moved. However,
1937 there are a couple of unusual error situations where the daemon does write
1938 log entries, and I had neglected to close the log afterwards.
1940 30. The text of an SMTP error response that was received during a remote
1941 delivery was being truncated at 512 bytes. This is too short for some of
1942 the long messages that one sometimes sees. I've increased the limit to
1945 31. It is now possible to make retry rules that apply only when a message has a
1946 specific sender, in particular, an empty sender.
1948 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1949 it possible to be selective about when SMTP synchronization is enforced.
1951 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1953 32. Implemented hosts_connection_nolog.
1955 33. Added an ACL for QUIT.
1957 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1960 35. Added mailbox_size and mailbox_filecount to appendfile.
1962 36. Added control = no_multiline_responses to ACLs.
1964 37. There was a bug in the logic of the code that waits for the clock to tick
1965 in the case where the clock went backwards by a substantial amount such
1966 that the microsecond fraction of "now" was more than the microsecond
1967 fraction of "then" (but the whole seconds number was less).
1969 38. Added support for the libradius Radius client library this is found on
1970 FreeBSD (previously only the radiusclient library was supported).
1976 1. When certain lookups returned multiple values in the form name=value, the
1977 quoting of the values was not always being done properly. Specifically:
1978 (a) If the value started with a double quote, but contained no whitespace,
1980 (b) If the value contained whitespace other than a space character (i.e.
1981 tabs or newlines or carriage returns) it was not quoted.
1982 This fix has been applied to the mysql and pgsql lookups by writing a
1983 separate quoting function and calling it from the lookup code. The fix
1984 should probably also be applied to nisplus, ibase and oracle lookups, but
1985 since I cannot test any of those, I have not disturbed their existing code.
1987 2. A hit in the callout cache for a specific address caused a log line with no
1988 reason for rejecting RCPT. Now it says "Previous (cached) callout
1989 verification failure".
1991 3. There was an off-by-one bug in the queryprogram router. An over-long
1992 return line was truncated at 256 instead of 255 characters, thereby
1993 overflowing its buffer with the terminating zero. As well as fixing this, I
1994 have increased the buffer size to 1024 (and made a note to document this).
1996 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1997 when Exim is waiting for an SMTP response from a remote server, Exim
1998 restarts its select() call on the socket, thereby resetting its timeout.
1999 This is not a problem when such interrupts are rare. Somebody set up a cron
2000 job to run exiwhat every 2 minutes, which is less than the normal select()
2001 timeout (5 or 10 minutes). This meant that the select() timeout never
2002 kicked in because it was always reset. I have fixed this by comparing the
2003 time when an interrupt arrives with the time at the start of the first call
2004 to select(). If more time than the timeout has elapsed, the interrupt is
2005 treated as a timeout.
2007 5. Some internal re-factoring in preparation for the addition of Sieve
2008 extensions (by MH). In particular, the "personal" test is moved to a
2009 separate function, and given an option for scanning Cc: and Bcc: (which is
2010 not set for Exim filters).
2012 6. When Exim created an email address using the login of the caller as the
2013 local part (e.g. when creating a From: or Sender: header line), it was not
2014 quoting the local part when it contained special characters such as @.
2016 7. Installed new OpenBSD configuration files.
2018 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2019 try to make them clearer.
2021 9. Callout options, other than the timeout value, were being ignored when
2022 verifying sender addresses in header lines. For example, when using
2024 verify = header_sender/callout=no_cache
2026 the cache was (incorrectly) being used.
2028 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2029 only the Cygwin environment.
2031 11. When return_path_on_delivery was set as a log selector, if different remote
2032 addresses in the same message used different return paths and parallel
2033 remote delivery occurred, the wrong values would sometimes be logged.
2034 (Whenever a remote delivery process finished, the return path value from
2035 the most recently started remote delivery process was logged.)
2037 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2038 lines when AUTH and/or TLS are in use. This is the "received protocol"
2039 field. Exim used to use "asmtp" for authenticated SMTP, without any
2040 indication (in the protocol name) for TLS use. Now it follows the RFC and
2041 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2042 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2043 names appear in log lines as well as in Received: header lines.
2045 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2046 extensions, and comparison tests, and to fix some bugs.
2048 14. Changes to the "personal" filter test:
2050 (1) The test was buggy in that it was just doing the equivalent of
2051 "contains" tests on header lines. For example, if a user's address was
2052 anne@some.where, the "personal" test would incorrectly be true for
2054 To: susanne@some.where
2056 This test is now done by extracting each address from the header in turn,
2057 and checking the entire address. Other tests that are part of "personal"
2058 are now done using regular expressions (for example, to check local parts
2059 of addresses in From: header lines).
2061 (2) The list of non-personal local parts in From: addresses has been
2062 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2063 taken from the Sieve specification recommendations.
2065 (3) If the message contains any header line starting with "List-" it is
2066 treated as non-personal.
2068 (4) The test for "circular" in the Subject: header line has been removed
2069 because it now seems ill-conceived.
2071 15. Minor typos in src/EDITME comments corrected.
2073 16. Installed latest exipick from John Jetmore.
2075 17. If headers_add on a router specified a text string that was too long for
2076 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2077 of string_sprintf() is now avoided.
2079 18. $message_body_size was not set (it was always zero) when running the DATA
2080 ACL and the local_scan() function.
2082 19. For the "mail" command in an Exim filter, no default was being set for
2083 the once_repeat time, causing a random time value to be used if "once" was
2084 specified. (If the value happened to be <= 0, no repeat happened.) The
2085 default is now 0s, meaning "never repeat". The "vacation" command was OK
2086 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2087 (I found it when inspecting the code).
2089 20. There is now an overall timeout for performing a callout verification. It
2090 defaults to 4 times the callout timeout, which applies to individual SMTP
2091 commands during the callout. The overall timeout applies when there is more
2092 than one host that can be tried. The timeout is checked before trying the
2093 next host. This prevents very long delays if there are a large number of
2094 hosts and all are timing out (e.g. when the network connections are timing
2095 out). The value of the overall timeout can be changed by specifying an
2096 additional sub-option for "callout", called "maxwait". For example:
2098 verify = sender/callout=5s,maxwait=20s
2100 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2101 to the end before writing, but this should make it even safer).
2103 22. Exim was forgetting that it had advertised PIPELINING for the second and
2104 subsequent messages on an SMTP connection. It was also not resetting its
2105 memory on STARTTLS and an internal HELO.
2107 23. When Exim logs an SMTP synchronization error within a session, it now
2108 records whether PIPELINING has been advertised or not.
2110 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2111 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2112 rather than long int.
2114 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2116 26. Added the never_mail option to autoreply.
2122 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2123 crash if the getsockname() call failed; this can happen if a connection is
2124 closed very soon after it is established. The problem was simply in the
2125 order in which certain operations were done, causing Exim to try to write
2126 to the SMTP stream before it had set up the file descriptor. The bug has
2127 been fixed by making things happen in the correct order.
2133 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2134 before the connection was closed, thus losing the rejection response.
2136 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2137 some early Solaris releases, but causes trouble in current releases where
2138 socklen_t is defined.
2140 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2143 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2146 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2147 that releasing the top of it at the end releases what was used for sub-
2148 expansions (unless the block got too big). However, discard this block if
2149 the first thing is a variable or header, so that we can use its block when
2150 it is dynamic (useful for very large $message_headers, for example).
2152 6. Lookups now cache *every* query, not just the most recent. A new, separate
2153 store pool is used for this. It can be recovered when all lookup caches are
2154 flushed. Lookups now release memory at the end of their result strings.
2155 This has involved some general refactoring of the lookup sources.
2157 7. Some code has been added to the store_xxx() functions to reduce the amount
2158 of flapping under certain conditions.
2160 8. log_incoming_interface used to affect only the <= reception log lines. Now
2161 it causes the local interface and port to be added to several more SMTP log
2162 lines, for example "SMTP connection from", and rejection lines.
2164 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2166 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2168 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2169 could be overwritten at the end of the current message (or the start of a
2170 new message if it was set in a HELO ACL). The value is now preserved for
2171 the duration of the SMTP connection.
2173 12. If a transport had a headers_rewrite setting, and a matching header line
2174 contained an unqualified address, that address was qualified, even if it
2175 did not match any rewriting rules. The underlying bug was that the values
2176 of the flags that permit the existence of unqualified sender and recipient
2177 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2178 non-local messages, and by -bnq for local messages) were not being
2179 preserved with the message after it was received.
2181 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2182 "next input=" as part of the text comprising the host identity instead of
2183 the correct text. The code was using the same buffer for two different
2184 strings. However, depending on which order the printing function evaluated
2185 its arguments, the bug did not always show up. Under Linux, for example, my
2186 test suite worked just fine.
2188 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2189 doesn't work with some older versions of Perl. It has been changed to "my",
2190 which in any case is probably the better facility to use.
2192 15. A really picky compiler found some instances of statements for creating
2193 error messages that either had too many or two few arguments for the format
2196 16. The size of the buffer for calls to the DNS resolver has been increased
2197 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2198 for addresses that have a lot of PTR records. This alleviates a problem; it
2199 does not fully solve it.
2201 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2202 buffer now truncates the list and logs the incident, which is the same
2203 action as happens when Exim is looking up a host name and its aliases.
2204 Previously in this situation something unpredictable would happen;
2205 sometimes it was "internal error: store_reset failed".
2207 18. If a server dropped the connection unexpectedly when an Exim client was
2208 using GnuTLS and trying to read a response, the client delivery process
2209 crashed while trying to generate an error log message.
2211 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2212 string, for example:
2214 warn message = H1: something\nH2: something
2216 the text was added as a single header line from Exim's point of view
2217 though it ended up OK in the delivered message. However, searching for the
2218 second and subsequent header lines using $h_h2: did not work. This has been
2219 fixed. Similarly, if a system filter added multiple headers in this way,
2220 the routers could not see them.
2222 20. Expanded the error message when iplsearch is called with an invalid key to
2223 suggest using net-iplsearch in a host list.
2225 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2226 ACLs are no longer actually imposed (and a message to that effect is
2229 22. If a "gecos" field in a passwd entry contained escaped characters, in
2230 particular, if it contained a \" sequence, Exim got it wrong when building
2231 a From: or a Sender: header from that name. A second bug also caused
2232 incorrect handling when an unquoted " was present following a character
2233 that needed quoting.
2235 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2236 was not being matched caselessly.
2238 24. Arranged for all hyphens in the exim.8 source to be escaped with
2241 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2242 properly. Recipient callout cache records were still being keyed to include
2243 the sender, even when use_sender was set false. This led to far more
2244 callouts that were necessary. The sender is no longer included in the key
2245 when use_sender is false.
2247 26. Added "control = submission" modifier to ACLs.
2249 27. Added the ${base62d: operator to decode base 62 numbers.
2251 28. dnsdb lookups can now access SRV records.
2253 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2254 the configuration file.
2256 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2257 (-v) mode. This makes the output for a verbose queue run more intelligible.
2259 31. Added a use_postmaster feature to recipient callouts.
2261 32. Added the $body_zerocount variable, containing the number of binary zero
2262 bytes in the message body.
2264 33. The time of last modification of the "new" subdirectory is now used as the
2265 "mailbox time last read" when there is a quota error for a maildir
2268 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2270 35. Added +ignore_unknown as a special item in host lists.
2272 36. Code for decoding IPv6 addresses in host lists is now included, even if
2273 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2274 address was recognized as an IP address, but was then not correctly decoded
2275 into binary, causing unexpected and incorrect effects when compared with
2282 1. Very minor rewording of debugging text in manualroute to say "list of
2283 hosts" instead of "hostlist".
2285 2. If verify=header_syntax was set, and a header line with an unqualified
2286 address (no domain) and a large number of spaces between the end of the
2287 name and the colon was received, the reception process suffered a buffer
2288 overflow, and (when I tested it) crashed. This was caused by some obsolete
2289 code that should have been removed. The fix is to remove it!
2291 3. When running in the test harness, delay a bit after writing a bounce
2292 message to get a bit more predictability in the log output.
2294 4. Added a call to search_tidyup() just before forking a reception process. In
2295 theory, someone could use a lookup in the expansion of smtp_accept_max_
2296 per_host which, without the tidyup, could leave open a database connection.
2298 5. Added the variables $recipient_data and $sender_data which get set from a
2299 lookup success in an ACL "recipients" or "senders" condition, or a router
2300 "senders" option, similar to $domain_data and $local_part_data.
2302 6. Moved the writing of debug_print from before to after the "senders" test
2305 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2306 problems for message scanning, either using a data ACL, or using
2307 local_scan() because the Received: header was not generated till after they
2308 were called (in order to set the time as the time of reception completion).
2309 I have revised the way this works. The header is now generated after the
2310 body is received, but before the ACL or local_scan() are called. After they
2311 are run, the timestamp in the header is updated.
2317 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2318 before starting a queue runner without re-exec. This happened only when
2319 deliver_drop_privilege was set or when the Exim user was set to root. The
2320 effect of the bug was that timeouts during subsequent deliveries caused
2321 crashes instead of being properly handled. The handler is now left at its
2322 default (and expected) setting.
2324 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2325 message, again when deliver_drop_privilege is set or Exim is run as root.
2326 The bug described in (1) was not present in this case, but the tidying up
2327 of the other signals was missing. I have made the two cases consistent.
2329 3. The ignore_target_hosts setting on a manualroute router was being ignored
2330 for hosts that were looked up using the /MX notation.
2332 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2335 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2336 operated on the sender address. After changing the $sender_address to <>
2337 for the sender address verify, Exim was re-instated it as the original
2338 (before rewriting) address, but remembering that it had rewritten it, so it
2339 wasn't rewriting it again. This bug also had the effect of breaking the
2340 sender address verification caching when the sender address was rewritten.
2342 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2343 This has been changed so that if the ip literal address matches
2344 ignore_target_hosts, the router declines.
2346 7. Added expansion conditions match_domain, match_address, and match_local_
2347 part (NOT match_host).
2349 8. The placeholder for the Received: header didn't have a length field set.
2351 9. Added code to Exim itself and to exim_lock to test for a specific race
2352 condition that could lead to file corruption when using MBX delivery. The
2353 issue is with the lockfile that is created in /tmp. If this file is removed
2354 after a process has opened it but before that process has acquired a lock,
2355 there is the potential for a second process to recreate the file and also
2356 acquire a lock. This could lead to two Exim processes writing to the file
2357 at the same time. The added code performs the same test as UW imapd; it
2358 checks after acquiring the lock that its file descriptor still refers to
2359 the same named file.
2361 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2362 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2365 11. Added the smtp_active_hostname option. If used, this will typically be made
2366 to depend on the incoming interface address. Because $interface_address is
2367 not set up until the daemon has forked a reception process, error responses
2368 that can happen earlier (such as "too many connections") no longer contain
2371 12. If an expansion in a condition on a "warn" statement fails because a lookup
2372 defers, the "warn" statement is abandoned, and the next ACL statement is
2373 processed. Previously this caused the whole ACL to be aborted.
2375 13. Added the iplsearch lookup type.
2377 14. Added ident_timeout as a log selector.
2379 15. Added tls_certificate_verified as a log selector.
2381 16. Added a global option tls_require_ciphers (compare the smtp transport
2382 option of the same name). This controls incoming TLS connections.
2384 17. I finally figured out how to make tls_require_ciphers do a similar thing
2385 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2386 before starting the TLS session.
2388 18. Tabs are now shown as \t in -bP output.
2390 19. If the log selector return_path_on_delivery was set, Exim crashed when
2391 bouncing a message because it had too many Received: header lines.
2393 20. If two routers both had headers_remove settings, and the first one included
2394 a superfluous trailing colon, the final name in the first list and the
2395 first name in the second list were incorrectly joined into one item (with a
2396 colon in the middle).
2402 1. Added -C and -D options to the exinext utility, mainly to make it easier
2403 to include in the automated testing, but these could be helpful when
2404 multiple configurations are in use.
2406 2. The exinext utility was not formatting the output nicely when there was
2407 an alternate port involved in the retry record key, nor when there was a
2408 message id as well (for retries that were specific to a specific message
2409 and a specific host). It was also confused by IPv6 addresses, because of
2410 the additional colons they contain. I have fixed the IPv4 problem, and
2411 patched it up to do a reasonable job for IPv6.
2413 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2414 delivery, the log line now contains "pipelined" if PIPELINING was used.
2416 4. An SMTP transport process used to panic and die if the bind() call to set
2417 an explicit outgoing interface failed. This has been changed; it is now
2418 treated in the same way as a connect() failure.
2420 5. A reference to $sender_host_name in the part of a conditional expansion
2421 that was being skipped was still causing a DNS lookup. This no longer
2424 6. The def: expansion condition was not recognizing references to header lines
2425 that used bh_ and bheader_.
2427 7. Added the _cache feature to named lists.
2429 8. The code for checking quota_filecount in the appendfile transport was
2430 allowing one more file than it should have been.
2432 9. For compatibility with Sendmail, the command line option
2440 and sets the incoming protocol and host name (for trusted callers). The
2441 host name and its colon can be omitted when only the protocol is to be set.
2442 Note the Exim already has two private options, -pd and -ps, that refer to
2443 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2444 "s", but I don't think that's a major issue.
2446 10. A number of refactoring changes to the code, none of which should affect
2449 (a) The number of logging options was getting close to filling up the
2450 32-bit word that was used as a bit map. I have split them into two classes:
2451 those that are passed in the argument to log_write(), and those that are
2452 only ever tested independently outside of that function. These are now in
2453 separate 32-bit words, so there is plenty of room for expansion again.
2454 There is no change in the user interface or the logging behaviour.
2456 (b) When building, for example, log lines, the code previously used a
2457 macro that called string_cat() twice, in order to add two strings. This is
2458 not really sufficiently general. Furthermore, there was one instance where
2459 it was actually wrong because one of the argument was used twice, and in
2460 one call a function was used. (As it happened, calling the function twice
2461 did not affect the overall behaviour.) The macro has been replaced by a
2462 function that can join an arbitrary number of extra strings onto a growing
2465 (c) The code for expansion conditions now uses a table and a binary chop
2466 instead of a serial search (which was left over from when there were very
2467 few conditions). Also, it now recognizes conditions like "pam" even when
2468 the relevant support is not compiled in: a suitably worded error message is
2469 given if an attempt is made to use such a condition.
2471 11. Added ${time_interval:xxxxx}.
2473 12. A bug was causing one of the ddress fields not to be passed back correctly
2474 from remote delivery subprocesses. The field in question was not being
2475 subsequently used, so this caused to problems in practice.
2477 13. Added new log selectors queue_time and deliver_time.
2479 14. Might have fixed a bug in maildirsizefile handling that threw up
2480 "unexpected character" debug warnings, and recalculated the data
2481 unnecessarily. In any case, I expanded the warning message to give more
2484 15. Added the message "Restricted characters in address" to the statements in
2485 the default ACL that block characters like @ and % in local parts.
2487 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2488 Three changes have been made:
2490 (a) There was a serious bug; a negative response to MAIL caused the whole
2491 recipient domain to be cached as invalid, thereby blocking all messages
2492 to all local parts at the same domain, from all senders. This bug has
2493 been fixed. The domain is no longer cached after a negative response to
2494 MAIL if the sender used is not empty.
2496 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2499 (c) A new callout option, "use_sender" has been added for people who want
2500 the modified behaviour.
2506 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2509 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2510 indeed breaks things for older releases.
2512 3. Added additional logging to the case where there is a problem reading data
2513 from a filter that is running in a subprocess using a pipe, in order to
2514 try to track down a specific problem.
2516 4. Testing facility fudge: when running in the test harness and attempting
2517 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2518 getting "No route to host". Convert this to a timeout.
2520 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2523 6. Some OS don't have socklen_t but use size_t instead. This affects the
2524 fifth argument of getsockopt() amongst other things. This is now
2525 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2526 can be set for individual OS. I have set it for SunOS5, OSF1, and
2527 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2528 some earlier ones do not.
2530 7. Change 4.30/15 was not doing the test caselessly.
2532 8. The standard form for an IPv6 address literal was being rejected by address
2533 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2534 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2535 this, as well as the form without the "IPv6" on the front (but only when
2536 address literals are enabled, of course).
2538 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2540 10. Exim crashed if a message with an empty sender address specified by -f
2541 encountered a router with an errors_to setting. This could be provoked only
2542 by a command such as
2546 where an empty string was supplied; "<>" did not hit this bug.
2548 11. Installed PCRE release 4.5.
2550 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2551 remained set. It is now erased.
2553 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2554 times from message ids (which are base 36 rather than the normal 62).
2556 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2557 were being counted as actual protocol errors, and logged if the log
2558 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2559 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2560 and DATA following a set of rejected RCPTs do not count as protocol errors.
2561 In other words, Exim assumes they were pipelined, though this may not
2562 actually be the case. Of course, in all cases the client gets an
2563 appropriate error code.
2565 15. If a lookup fails in an ACL condition, a message about the failure may
2566 be available; it is used if testing the ACL cannot continue, because most
2567 such messages specify what the cause of the deferral is. However, some
2568 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2569 that caused an old message to be retained and used if a later statement
2570 caused a defer, replacing the real cause of the deferral.
2572 16. If an IP address had so many PTR records that the DNS lookup buffer
2573 was not large enough to hold them, Exim could crash while trying to process
2574 the truncated data. It now detects and logs this case.
2576 17. Further to 4.21/58, another change has been made: if (and only if) the
2577 first line of a message (the first header line) ends with CRLF, a bare LF
2578 in a subsequent header line has a space inserted after it, so as not to
2579 terminate the header.
2581 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2582 unnecessarily, used atoi() instead of strtol(), and didn't check the
2583 termination when getting file sizes from file names by regex.
2585 19. Completely re-implemented the support for maildirsize files, in the light
2586 of a number of problems with the previous contributed implementation
2587 (4.30/29). In particular:
2589 . If the quota is zero, the maildirsize file is maintained, but no quota is
2592 . If the maildir directory does not exist, it is created before any attempt
2593 to write a maildirsize file.
2595 . The quota value in the file is just a cache; if the quota is changed in
2596 the transport, the new value overrides.
2598 . A regular expression is available for excluding directories from the
2601 20. The autoreply transport checks the characters in options that define the
2602 message's headers; it allows continued headers, but it was checking with
2603 isspace() after an embedded newline instead of explicitly looking for a
2606 21. If all the "regular" hosts to which an address was routed had passed their
2607 expiry times, and had not reached their retry times, the address was
2608 bounced, even if fallback hosts were defined. Now Exim should go on to try
2611 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2612 equivalent code in the SMTP transport. Some hosts send humungous responses
2613 to HELO/EHLO, more than 1024 it seems.
2615 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2616 gives compiler warnings in some environments. I've now done it "properly",
2619 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2620 (because of problems with the built-in one) was declared to return uschar *
2621 instead of char *, causing compiler failure.
2623 25. Fixed a file descriptor leak when processing alias/forward files.
2625 26. Fixed a minor format string issue in dbfn.c.
2627 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2629 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2630 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2633 29. When a sender address is verified, it is cached, to save repeating the test
2634 when there is more than one recipient in a message. However, when the
2635 verification involves a callout, it is possible for different callout
2636 options to be set for different recipients. It is too complicated to keep
2637 track of this in the cache, so now Exim always runs a verification when a
2638 callout is required, relying on the callout cache for the optimization.
2639 The overhead is duplication of the address routing, but this should not be
2642 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2643 to be verified with callout=postmaster, and the main callout worked but the
2644 postmaster check failed, the verification correctly failed. However, if a
2645 subsequent RCPT command asked for sender verification *without* the
2646 postmaster check, incorrect caching caused this verification also to fail,
2649 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2650 it was not caching the DNS options (qualify_single, search_parents) that
2651 were used when the lookup failed. A subsequent lookup with different
2652 options therefore always gave the same answer, though there were cases
2653 where it should not have. (Example: a "domains = !$mx_any" option on a
2654 dnslookup router: the "domains" option is always processed without any
2655 widening, but the router might have qualify_single set.) Now Exim uses the
2656 cached value only when the same options are set.
2658 32. Added John Jetmore's "exipick" utility to the distribution.
2660 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2661 than a timeout (e.g. a certificate is required, and is not provided), an
2662 Exim server now closes the connection immediately. Previously it waited for
2663 the client to close - but if the client is SSL, it seems that they each
2664 wait for each other, leading to a delay before one of them times out.
2666 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2667 maintained 0.8.x compatibility because I don't think many are using it, and
2668 it is clearly obsolete.
2670 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2673 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2674 client certificate was expired. A simple patch fixes this, though I don't
2675 understand the full logic of why the verify callback is called multiple
2678 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2679 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2680 which causes problems with some clients (such as the Certicom SSL Plus
2681 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2682 disables the coutermeasure allowing Eudora to connect."
2684 38. Exim was not checking that a write() to a log file succeeded. This could
2685 lead to Bad Things if a log got too big, in particular if it hit a file
2686 size limit. Exim now panics and dies if it cannot write to a log file, just
2687 as it does if it cannot open a log file.
2689 39. Modified OS/Makefile-Linux so that it now contains
2691 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2693 The two -D definitions ensure that Exim is compiled with large file
2694 support, which makes it possible to handle log files that are bigger than
2697 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2698 instance) a domain was checked against a named list that involved a lookup,
2699 causing $domain_data to be set, then another domain was checked against the
2700 same list, then the first domain was re-checked, the value of $domain_data
2701 after the final check could be wrong. In particular, if the second check
2702 failed, it could be set empty. This bug probably also applied to
2705 41. The strip_trailing_dot option was not being applied to the address given
2706 with the -f command-line option.
2708 42. The code for reading a message's header from the spool was incrementing
2709 $received_count, but never initializing it. This meant that the value was
2710 incorrect (doubled) while delivering a message in the same process in which
2711 it was received. In the most common configuration of Exim, this never
2712 happens - a fresh exec is done - but it can happen when
2713 deliver_drop_privilege is set.
2715 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2716 it now includes up to 150 characters of the unexpected data in the log
2719 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2720 and building data strings. The size of both of these buffers was 10 000
2721 bytes - far larger than anybody would *ever* want, thought I. Needless to
2722 say, somebody hit the limit. I have increased the maximum line length to
2723 20 000 and the maximum data length of concatenated lines to 100 000. I have
2724 also fixed two bugs, because there was no checking on these buffers. Tsk,
2725 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2726 buffer is too small.
2728 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2729 lsearch lookups. Now it does.
2731 46. When parsing a route_list item in a manualroute router, a fixed-length
2732 buffer was used for the list of hosts. I made this 1024 bytes long,
2733 thinking that nobody would ever have a list of hosts that long. Wrong.
2734 Somebody had a whole pile of complicated expansion conditions, and the
2735 string was silently truncated, leading to an expansion error. It turns out
2736 that it is easier to change to an unlimited length (owing to other changes
2737 that have happened since this code was originally written) than to build
2738 structure for giving a limitation error. The length of the item that
2739 expands into the list of hosts is now unlimited.
2741 47. The lsearch lookup could not handle data where the length of text line was
2742 more than 4095 characters. Such lines were truncated, leading to shortened
2743 data being returned. It should now handle lines of any length.
2745 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2746 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2749 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2750 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2751 no longer happens in comment lines. A list of replacements is now placed
2752 at the head of all of the source files, except those whose only change is
2753 to replace PERL_COMMAND in the very first #! line.
2755 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2756 messages on the queue, with a bottom-up merge sort, using code contributed
2757 by Michael Haardt. This should make operations like -bp somewhat faster on
2758 large queues. It won't affect queue runners, except when queue_run_in_order
2761 51. Installed eximstats 1.31 in the distribution.
2763 52. Added support for SRV lookups to the dnslookup router.
2765 53. If an ACL referred to $message_body or $message_body_end, the value was not
2766 reset for any messages that followed in the same SMTP session.
2768 54. The store-handling optimization for building very long strings was not
2769 differentiating between the different store pools. I don't think this
2770 actually made any difference in practice, but I've tidied it.
2772 55. While running the routers to verify a sender address, $sender_address
2773 was still set to the sender address. This is wrong, because when routing to
2774 send a bounce to the sender, it would be empty. Therefore, I have changed
2775 it so that, while verifying a sender address, $sender_address is set to <>.
2776 (There is no change to what happens when verifying a recipient address.)
2778 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2779 target A or AAAA records (if not already returned) without resetting the
2780 qualify_single or search_parents options of the DNS resolver. These are
2781 inappropriate in this case because the targets of MX and SRV records must
2782 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2783 target that, when qualified, matched something in the local domain. These
2784 two options are now turned off when doing these lookups.
2786 57. It seems that at least some releases of Reiserfs (which does not have the
2787 concept of a fixed number of inodes) returns zero and not -1 for the
2788 number of available inodes. This interacted badly with check_spool_inodes,
2789 which assumed that -1 was the "no such thing" setting. What I have done is
2790 to check that the total number of inodes is greater than zero before doing
2791 the test of how many are available.
2793 58. When a "warn" ACL statement has a log_message modifier, the message is
2794 remembered, and not repeated. This is to avoid a lot of repetition when a
2795 message has many recipients that cause the same warning to be written.
2796 Howewer, Exim was preserving the list of already written lines for an
2797 entire SMTP session, which doesn't seem right. The memory is now reset if a
2798 new message is started.
2800 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2801 debugging output unless log_rewrite was also set.
2803 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2804 of (int)(handle) when we know that handle contains (void *)(-1).
2806 61. The Exim daemon panic-logs an error return when it closes the incoming
2807 connection. However "connection reset by peer" seems to be common, and
2808 isn't really an error worthy of noting specially, so that particular error
2811 62. When Exim is trying to find all the local interfaces, it used to panic and
2812 die if the ioctl to get the interface flags failed. However, it seems that
2813 on at least one OS (Solaris 9) it is possible to have an interface that is
2814 included in the list of interfaces, but for which you get a failure error
2815 for this call. This happens when the interface is not "plumbed" into a
2816 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2817 failure of the "get flags" call assumes that the interface is down.
2819 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2820 makes life easier for people who are doing arithmetic on fields extracted
2821 from dates, where you often get leading zeros that should not be
2822 interpreted as octal.
2824 64. Added qualify_domain to the redirect router, to override the global
2827 65. If a pathologically long header line contained very many addresses (the
2828 report of this problem mentioned 10 000) and each of them was rewritten,
2829 Exim could use up a very large amount of memory. (It kept on making new
2830 copies of the header line as it rewrote, and never released the old ones.)
2831 At the expense of a bit more processing, the header rewriting function has
2832 been changed so that it no longer eats memory in this way.
2834 66. The generation of the Received: header has been moved from the time that a
2835 message starts to be received, to the time that it finishes. The timestamp
2836 in the Received: header should now be very close to that of the <= log
2837 line. There are two side-effects of this change:
2839 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2840 logged header lines no longer include the local Received: line, because
2841 it has not yet been created. The same applies to a copy of the message
2842 that is returned to a non-SMTP sender when a message is rejected.
2844 (b) When a filter file is tested using -bf, no additional Received: header
2845 is added to the test message. After some thought, I decided that this
2848 This change does not affect the value of $received_for. It is still set
2849 after address rewriting, but before local_scan() is called.
2851 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2853 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2854 gave an unhelpful panic error message, and a defer error. I have managed to
2855 change this behaviour so that it now rejects any supplied certificate,
2856 which seems right, as the list of acceptable certificates is empty.
2858 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2859 gave an unhelpful defer error. I have not managed to make this reject any
2860 supplied certificates, but the error message it gives is "no certificate
2861 supplied", which is not helpful.
2863 70. exigrep's output now also includes lines that are not associated with any
2864 message, but which match the given pattern. Implemented by a patch from
2865 Martin Sluka, which also tidied up the Perl a bit.
2867 71. Recipient callout verification, like sender verification, was using <> in
2868 the MAIL FROM command. This isn't really the right thing, since the actual
2869 sender may affect whether the remote host accepts the recipient or not. I
2870 have changed it to use the actual sender in the callout; this means that
2871 the cache record is now keyed on a recipient/sender pair, not just the
2872 recipient address. There doesn't seem to be a real danger of callout loops,
2873 since a callout by the remote host to check the sender would use <>.
2874 [SEE ABOVE: changed after hitting problems.]
2876 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2877 temporary errors. However, in the case of such a code being given after
2878 the end of a data transmission (i.e. after ".") Exim was failing to write
2879 a retry record for the message. (Yes, there was some broken host that was
2880 actually sending 8xx at this point.)
2882 73. An unknown lookup type in a host list could cause Exim to panic-die when
2883 the list was checked. (An example that provoked this was putting <; in the
2884 middle of a list instead of at the start.) If this happened during a DATA
2885 ACL check, a -D file could be left lying around. This kind of configuration
2886 error no longer causes Exim to die; instead it causes a defer errror. The
2887 incident is still logged to the main and panic logs.
2889 74. Buglet left over from Exim 3 conversion. The message "too many messages
2890 in one connection" was written to the rejectlog but not the mainlog, except
2891 when address rewriting (yes!) was being logged.
2893 75. Added write_rejectlog option.
2895 76. When a system filter was run not as root (that is, when system_filter_user
2896 was set), the values of the $n variables were not being returned to the
2897 main process; thus, they were not subsequently available in the $sn
2900 77. Added +return_path_on_delivery log selector.
2902 78. A connection timeout was being treated differently from recipients deferred
2903 when testing hosts_max_try with a message that was older than the host's
2904 retry timeout. (The host should not be counted, thus allowing all hosts to
2905 be tried at least once before bouncing.) This may have been the cause of an
2906 occasionally reported bug whereby a message would remain on the queue
2907 longer than the retry timeout, but would be bounced if a delivery was
2908 forced. I say "may" because I never totally pinned down the problem;
2909 setting up timeout/retry tests is difficult. See also the next item.
2911 79. The ultimate address timeout was not being applied to errors that involved
2912 a combination of host plus message (for example, a timeout on a MAIL
2913 command). When an address resolved to a number of possible hosts, and they
2914 were not all tried for each delivery (e.g. because of hosts_max_try), a
2915 message could remain on the queue longer than the retry timeout.
2917 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2920 81. Fixed an obscure SMTP outgoing bug which required at least the following
2921 conditions: (a) there was another message waiting for the same server;
2922 (b) the server returned 5xx to all RCPT commands in the first message so
2923 that the message was not completed; (c) the server dropped the connection
2924 or gave a negative response to the RSET that Exim sends to abort the
2925 transaction. The observed case was a dropped connection after DATA that had
2926 been sent in pipelining mode. That is, the server had advertised PIPELINING
2927 but was not implementing it correctly. The effect of the bug was incorrect
2928 behaviour, such as trying another host, and this could lead to a crash.
2934 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2935 and daemon.c were passed as pointers to ints; they should have been
2936 pointers to socklen_t variables (which are typically unsigned ints).
2938 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2941 3. Fixed a really odd bug that affected only the testing scheme; patching a
2942 certain fixed string in the binary changed the value of another string that
2943 happened to be identical to the end of the original first string.
2945 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2946 name", it returns that address as the IP address. On some operating
2947 systems (e.g. Solaris), it also passes back the IP address string as the
2948 "host name". However, on others (e.g. Linux), it passes back an empty
2949 string. Exim wasn't checking for this, and was changing the host name to an
2950 empty string, assuming it had been canonicized.
2952 5. Although rare, it is permitted to have more than one PTR record for a given
2953 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2954 all the names associated with an address, because they do in Solaris.
2955 However, it seems that they do not in Linux for data that comes from the
2956 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2957 I found this out when I moved to a new Linux workstation and tried to run
2958 the Exim test suite.
2960 To get round this problem I have changed the code so that it now does its
2961 own call to the DNS to look up PTR records when searching for a host name.
2962 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2963 addresses that are only in /etc/hosts are still found.
2965 This behaviour is, however, controlled by an option called host_lookup_
2966 order, which defaults to "bydns:byaddr". If people want to use the other
2967 order, or indeed, just use one or the other means of lookup, they can
2968 specify it in this variable.
2970 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2971 some operating systems, this comes back from gethostbyaddr() as an empty
2972 string, and this is what Exim used to test for. However, it seems that in
2973 other systems, "." is yielded. Exim now tests for this case too.
2975 7. The values of check_spool_space and check_log_space are now held internally
2976 as a number of kilobytes instead of an absolute number of bytes. If a
2977 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2978 kilobyte. This means that much larger values can be stored.
2980 8. Exim monitor: an attempt to get the action menu when not actually pointing
2981 at a message produces an empty menu entitled "No message selected". This
2982 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2983 no entries in it ("Shell widget menu has zero width and/or height"). So I
2984 have added a single, blank menu entry in this case.
2986 9. Added ${quote_local_part.
2988 10. MIME decoding is now applied to the contents of Subject: header lines when
2991 11. Now that a reference to $sender_host_address automatically causes a reverse
2992 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2993 host lookup before query-style lookups in lists that might use this
2994 variable. This has therefore been abolished, and the "net-" prefix is no
2995 longer necessary for query-style lookups.
2997 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2998 have been a typo for LFLAGS, so it has been changed.
3000 13. The install script calls Exim with "-C /dev/null" in order to find the
3001 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3002 to be output. Howeve, since Exim outputs its version number before the
3003 error, it didn't break the script. It just looked ugly. I fixed this by
3004 always allowing "-C /dev/null" if the caller is root.
3006 14. Ignore overlarge ACL variable number when reading spool file - insurance
3007 against a later release with more variables having written the file.
3009 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3010 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3011 this, as well as the form without the "IPv6" on the front.
3013 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3014 OS/Makefile-Darwin file.
3016 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3017 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3019 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3020 variable. However, some OS use ints for inodes. I've added cast to long int
3021 to get rid of the compiler warning.
3023 19. I had forgotten to lock out "/../" in configuration file names when
3024 ALT_CONFIG_PREFIX was set.
3026 20. Routers used for verification do not need to specify transports. However,
3027 if such a router generated a host list, and callout was configured, Exim
3028 crashed, because it could not find a port number from the (non-existent)
3029 transport. It now assumes port 25 in this circumstance.
3031 21. Added the -t option to exigrep.
3033 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3034 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3035 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3036 them is set, LOOKUP_LSEARCH is forced.
3038 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3040 24. Added sender and host information to the "rejected by local_scan()" log
3041 line; previously there was no indication of these.
3043 25. Added .include_if_exists.
3045 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3046 Linux. It turns out that not all file systems support this. Apparently some
3047 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3048 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3049 not supported on the file descriptor, is now ignored when Exim is trying to
3050 sync a directory. This applies only to Linux.
3052 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3054 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3055 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3058 29. Added support for maildirsize files from supplied patch (modified a bit).
3060 30. The use of :fail: followed by an empty string could lead Exim to respond to
3061 sender verification failures with (e.g.):
3063 550 Verification failed for <xxx>
3064 550 Sender verify failed
3066 where the first response line was missing the '-' that indicates it is not
3067 the final line of the response.
3069 31. The loop for finding the name of the user that called Exim had a hardwired
3070 limit of 10; it now uses the value of finduser_retries, which is used for
3071 all other user lookups.
3073 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3076 33. Exim was neglecting to zero errno before one call of strtol() when
3077 expanding a string and expecting an integer value. On some systems this
3078 resulted in spurious "integer overflow" errors. Also, it was casting the
3079 result into an int without checking.
3081 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3082 did not work. The code looks as if it has *never* worked, though it appears
3083 to have been documented since at least releast 1.62. I have made it work.
3085 35. The "timeout_DNS" error in retry rules, also documented since at least
3086 1.62, also never worked. As it isn't clear exactly what this means, and
3087 clearly it isn't a major issue, I have abolished the feature by treating it
3088 as "timeout", and writing a warning to the main and panic logs.
3090 36. The display of retry rules for -brt wasn't always showing the error code
3093 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3094 timeout_connect_A, timeout_connect_MX.
3096 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3097 to the empty sender.
3099 39. The daemon was not analysing the content of -oX till after it had closed
3100 stderr and disconnected from the controlling terminal. This meant that any
3101 syntax errors were only noted on the panic log, and the return code from
3102 the command was 0. By re-arranging the code a little, I've made the
3103 decoding happen first, so such errors now appear on stderr, and the return
3104 code is 1. However, the actual setting up of the sockets still happens in
3105 the disconnected process, so errors there are still only recorded on the
3108 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3109 connections (as happens on some IP stacks) was logged at start up time as
3110 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3111 it from "IPv6 and IPv4", which means that two separate sockets are being
3114 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3115 says whether AF_INET or AF_INET6 was passed as an argument.
3117 42. Exiwhat output was messed up when time zones were included in log
3120 43. Exiwhat now gives more information about the daemon's listening ports,
3121 and whether -tls-on-connect was used.
3123 44. The "port" option of the smtp transport is now expanded.
3125 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3126 silently ignored. Now an error message is written to the main and panic
3129 46. There's a new ACL modifier called "logwrite" which writes to a log file
3130 as soon as it is encountered.
3132 47. Added $local_user_uid and $local_user_gid at routing time.
3134 48. Exim crashed when trying to verify a sender address that was being
3137 49. Exim was recognizing only a space character after ".include". It now also
3138 recognizes a tab character.
3140 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3141 extracting the relevant information from the specification. The man page no
3142 longer contains scrambled data for the -d option, and I've added a section
3143 at the front about calling Exim under different names.
3145 51. Added "extra_headers" argument to the "mail" command in filter files.
3147 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3150 53. Installed eximstats 1.29.
3152 54. Added transport_filter_timeout as a generic transport option.
3154 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3155 Cc: header lines. This was required by RFC 822, but it not required by RFC
3158 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3159 incoming messages that did not have them. Now it does so only if the
3160 message originates locally, that is, if there is no associated remote host
3161 address. When Resent- header lines are present, this applies to the Resent-
3162 lines rather than the non-Resent- lines.
3164 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3165 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3167 58. Messages for configuration errors now include the name of the main
3168 configuration file - useful now that there may be more than one file in a
3169 list (.included file names were always shown).
3171 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3172 for those rare installations that do not start the daemon as root or run it
3173 setuid root. I've cut out the call to initgroups() if the daemon is not
3176 60. The Exim user and group can now be bound into the binary as text strings
3177 that are looked up at the start of Exim's processing.
3179 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3181 62. Added $mailstore_basename variable.
3183 63. Installed patch to sieve.c from Michael Haardt.
3185 64. When Exim failed to open the panic log after failing to open the main log,
3186 the original message it was trying to log was written to stderr and debug
3187 output, but if they were not available (the usual case in production), it
3188 was lost. Now it is written to syslog before the two lines that record the
3189 failures to open the logs.
3191 65. Users' Exim filters run in subprocesses under the user's uid. It is
3192 possible for a "deliver" command or an alias in a "personal" command to
3193 provoke an address rewrite. If logging of address rewriting is configured,
3194 this fails because the process is not running as root or exim. There may be
3195 a better way of dealing with this, but for the moment (because 4.30 needs
3196 to be released), I have disabled address rewrite logging when running a
3197 filter in a non-root, non-exim process.
3203 1. The buildconfig auxiliary program wasn't quoting the value set for
3204 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3205 not defined. This bug was masked in 4.22 by the effect that was fixed in
3208 2. Some messages that were rejected after a message id was allocated were
3209 shown as "incomplete" by exigrep. It no longer does this for messages that
3210 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3212 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3213 have allow_domain_literals set, the ID did not get logged in the <= line.
3214 Domain literals are now always recognized in Message-ID: header lines.
3216 4. The first argument for a ${extract expansion item is the key name or field
3217 number. Leading and trailing spaces in this item were not being ignored,
3218 causing some misleading effects.
3220 5. When deliver_drop_privilege was set, single queue runner processes started
3221 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3222 same command in the process it spins off) were not dropping privilege.
3224 6. When the daemon running as "exim" started a queue runner, it always
3225 re-executed Exim in the spun-off process. This is a waste of effort when
3226 deliver_drop_privilege is set. The new process now just calls the
3227 queue-runner function directly.
3233 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3236 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3237 ignored. Though the use of -oP was forcing the writing of a pid file, it
3238 was always written to the default place.
3240 3. If the message "no IP address found for host xxxx" is generated during
3241 incoming verification, it is now followed by identification of the incoming
3242 connection (so you can more easily find what provoked it).
3244 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3246 5. Added some features to "harden" Exim a bit more against certain attacks:
3248 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3249 be put in Local/Makefile. This is like the never_users runtime option,
3250 but it cannot be overridden. The default setting is "root".
3252 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3253 prefix string with which any file named in a -C command line option
3256 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3257 is retained for -C and -D only if the caller of Exim is root. Without
3258 it, the exim user may also use -C and -D and retain privilege.
3260 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3261 command line option is disabled.
3263 6. Macro names set by the -D option must start with an upper case letter, just
3264 like macro names defined in the configuration file.
3266 7. Added "dereference=" facility to LDAP.
3268 8. Two instances of the typo "uknown" in the source files are fixed.
3270 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3271 the Configure-Makefile script screwed up while processing it.
3273 10. Incorporated PCRE 4.4.
3275 11. The SMTP synchronization check was not operating right at the start of an
3276 SMTP session. For example, it could not catch a HELO sent before the client
3277 waited for the greeting. There is now a check for outstanding input at the
3278 point when the greeting is written. Because of the duplex, asynchronous
3279 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3280 way, but not yet received, when the check is performed.
3282 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3283 on TCP/IP sockets, because this apparently causes some broken clients to
3286 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3287 unchanged) from the Cygwin maintainer.
3289 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3290 was supported instead of testing for mbx. Effectively a typo.
3292 15. The spa authenticator server code was not checking that the input it
3293 received was valid base64.
3295 16. The debug output line for the "set" modifier in ACLs was not showing the
3296 name of the variable that was being set.
3298 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3300 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3301 lookup on its own. Something else was needed to trigger the lookup. For
3302 example, a match in host_lookup or the need for a host name in a host list.
3303 Now, if $sender_host_name is referenced and the host name has not yet been
3304 looked up, a lookup is performed. If the lookup fails, the variable remains
3305 empty, and $host_lookup_failed is set to "1".
3307 19. Added "eqi" as a case-independent comparison operator.
3309 20. The saslauthd authentication condition could segfault if neither service
3310 nor realm was specified.
3312 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3313 error message that was logged was misleading, and incoming SMTP
3314 connections were dropped. The message is now more accurate, and temporary
3315 errors are given to SMTP connections.
3317 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3318 (except RSET) with a 421 error, until QUIT is received. However, it was
3319 failing to send a response to QUIT.
3321 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3322 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3323 ACL is now run after the helo_try_verify_hosts code.
3325 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3326 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3327 case-independent, so other case variants are also recognized). Apparently
3328 some systems use these upper case variants.
3330 25. If more than two messages were waiting for the same host, and a transport
3331 filter was specified for the transport, Exim sent two messages over the
3332 same TCP/IP connection, and then failed with "socket operation on non-
3333 socket" when it tried to send the third.
3335 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3337 27. The extern definition of crypt16() in expand.c was not being excluded when
3338 the OS had its own crypt16() function.
3340 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3341 as a preferred synonym for return_size_limit, both as an option and as an
3344 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3346 30. Changed the default configuration ACL to relax the local part checking rule
3347 for addresses that are not in any local domains. For these addresses,
3348 slashes and pipe symbols are allowed within local parts, but the sequence
3349 /../ is explicitly forbidden.
3351 31. SPA server authentication was not clearing the challenge buffer before
3354 32. log_message in a "warn" ACL statement was writing to the reject log as
3355 well as to the main log, which contradicts the documentation and doesn't
3356 seem right (because no rejection is happening). So I have stopped it.
3358 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3359 However, I am unable to do any testing of this.
3361 34. Fixed an infelicity in the appendfile transport. When checking directories
3362 for a mailbox, to see if any needed to be created, it was accidentally
3363 using path names with one or more superfluous leading slashes; tracing
3364 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3366 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3367 contents are added to the log line that is written for every discarded
3368 recipient. (Previously a log_message setting was ignored.)
3370 36. The ${quote: operator now quotes the string if it is empty.
3372 37. The install script runs exim in order to find its version number. If for
3373 some reason other than non-existence or emptiness, which it checks, it
3374 could not run './exim', it was installing it with an empty version number,
3375 i.e. as "exim-". This error state is now caught, and the installation is
3378 38. An argument was missing from the function that creates an error message
3379 when Exim fails to connect to the socket for saslauthd authentication.
3380 This could cause Exim to crash, or give a corrupted message.
3382 39. Added isip, isip4, and isip6 to ${if conditions.
3384 40. The ACL variables $acl_xx are now saved with the message, and can be
3385 accessed later in routers, transports, and filters.
3387 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3388 strings in the file are not string-expanded.
3390 42. If a MAIL command specified a SIZE value that was too large to fit into an
3391 int variable, the check against message_size_limit failed. Such values are
3392 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3393 day this will have to be increased, but I don't think I want to be around
3394 when emails are that large.
3401 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3402 iconv() is not standard in FreeBSD.
3404 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3405 IPv6 enabled. The observed symptom was a segmentation fault on return from
3406 the function os_common_find_running_interfaces() in src/os.c.
3408 3. In the check_special_case() function in daemon.c I had used "errno" as an
3409 argument name, which causes warnings on some systems. This was basically a
3410 typo, since it was named "eno" in the comments!
3412 4. The code that waits for the clock to tick (at a resolution of some fraction
3413 of a second) so as to ensure message-id uniqueness was always waiting for
3414 at least one whole tick, when it could have waited for less. [This is
3415 almost certainly not relevant at current processor speeds, where it is
3416 unlikely to ever wait at all. But we try to future-proof.]
3418 5. The function that sleeps for a time interval that includes fractions of a
3419 second contained a race. It did not block SIGALRM between setting the
3420 timer, and suspending (a couple of lines later). If the interval was short
3421 and the sigsuspend() was delayed until after it had expired, the suspension
3422 never ended. On busy systems this could lead to processes getting stuck for
3425 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3426 process, before it forks any delivery processes. The open lookup caching
3427 mechanism meant that the open file or database connection was passed into
3428 the delivery process. The problem was that delivery processes always tidy
3429 up cached lookup data. This could cause a problem for the next delivery
3430 process started by the queue runner, because the external queue runner
3431 process does not know about the closure. So the next delivery process
3432 still has data in the lookup cache. In the case of a file lookup, there was
3433 no problem because closing a file descriptor in a subprocess doesn't affect
3434 the parent. However, if the lookup was caching a connection to a database,
3435 the connection was closed, and the second delivery process was likely to
3436 see errors such as "PGSQL: query failed: server closed the connection
3437 unexpectedly". The problem has been fixed by closing all cached lookups
3438 in a queue runner before running a delivery process.
3440 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3441 seem to have the "const" qualifier which it has on other OS. I've
3444 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3445 *of the same type* (client or server) with the same public name that an
3446 error should be diagnosed.
3448 9. When Exim looked up a host name for an IP address, but failed to find the
3449 original IP address when looking up the host name (a safety check), it
3450 output the message "<ip address> does not match any IP for NULL", which was
3451 confusing, to say the least. The bug was that the host name should have
3452 appeared instead of "NULL".
3454 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3455 user that is built into the binary, and the -C or -D options is used, root
3456 privilege is dropped before the configuration file is read. In addition,
3457 logging is switched to stderr instead of the normal log files. If the
3458 configuration then re-defines the Exim user, the unprivileged environment
3459 is probably not what is expected, so Exim logs a panic warning message (but
3462 However, if deliver_drop_privilege is set, the unprivileged state may well
3463 be exactly what is intended, so the warning has been cut out in that case,
3464 and Exim is allowed to try to write to its normal log files.
3470 1. smtp_return_error_details was not giving details for temporary sender
3471 or receiver verification errors.
3473 2. Diagnose a configuration error if two authenticators have the same public
3476 3. Exim used not to create the message log file for a message until the first
3477 delivery attempt. This could be confusing when incoming messages were held
3478 for policy or load reasons. The message log file is now created at the time
3479 the message is received, and an initial "Received" line is written to it.
3481 4. The automatically generated man page for command line options had a minor
3482 bug that caused no ill effects; however, a more serious problem was that
3483 the procedure for building the man page automatically didn't always
3484 operate. Consequently, release 4.20 contains an out-of-date version. This
3485 shouldn't happen again.
3487 5. When building Exim with embedded Perl support, the script that builds the
3488 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3489 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3491 6. The freeze_tell option was not being used for messages that were frozen on
3492 arrival, either by an ACL or by local_scan().
3494 7. Added the smtp_incomplete_transaction log selector.
3496 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3497 was accepting AUTH without a new EHLO.
3499 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3500 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3503 10. Logging of TCP/IP connections (when configured) now happens in the main
3504 daemon process instead of the child process, so that the TCP/IP connection
3505 count is more accurate (but it can never be perfect).
3507 11. The use of "drop" in a nested ACL was not being handled correctly in the
3508 outer ACL. Now, if condition failure induced by the nested "drop" causes
3509 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3510 or "require"), the connection is dropped.
3512 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3513 that yield "discard" can now be used with an "accept" or a "discard" verb,
3514 but an error is generated for any others (because I can't see a useful way
3515 to define what should happen).
3517 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3518 are now processed in the same way as lines in the Exim configuration file.
3519 In particular, continuation lines are supported.
3521 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3523 15. Added -ti meaning -t -i.
3525 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3526 domains, and warn by logging if others are found.
3528 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3529 returned from GIFCONF when Exim is trying to find the list of interfaces on
3530 a host. The code in os.c has been modified to copy each ifreq to an aligned
3531 structure in all cases.
3533 Also, in some cases, the returned ifreq's were being copied to a 'struct
3534 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3535 means the last couple of bytes of an IPv6 address could be chopped if the
3536 ifreq contained only a normal sockaddr (14 bytes storage).
3538 18. Named domain lists were not supported in the hosts_treat_as_local option.
3539 An entry such as +xxxx was not recognized, and was treated as a literal
3542 19. Ensure that header lines added by a DATA ACL are included in the reject log
3543 if the ACL subsequently rejects the message.
3545 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3546 MD5 (which is deprecated).
3548 21. When testing a filter file using -bf, Exim was writing a message when it
3549 took the sender from a "From " line in the message, but it was not doing so
3550 when it took $return_path from a Return-Path: header line. It now does.
3552 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3553 with a valid header line field name (a series of printing characters
3554 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3556 23. Changed "disc" in the source to "disk" to conform to the documentation and
3557 the book and for uniformity.
3559 24. Ignore Sendmail's -Ooption=value command line item.
3561 25. When execve() failed while trying to run a command in a pipe transport,
3562 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3563 could be confused with a return value of 69 from the command itself. This
3564 has been changed to 127, the value the shell returns if it is asked to run
3565 a non-existent command. The wording for the related log line suggests a
3566 non-existent command as the problem.
3568 26. If received_header_text expands to an empty string, do not add a Received:
3569 header line to the message. (Well, it adds a token one on the spool, but
3570 marks it "old" so that it doesn't get used or transmitted.)
3572 27. Installed eximstats 1.28 (addition of -nt option).
3574 28. There was no check for failure on the call to getsockname() in the daemon
3575 code. This can fail if there is a shortage of resources on the system, with
3576 ENOMEM, for example. A temporary error is now given on failure.
3578 29. Contrary to the C standard, it seems that in some environments, the
3579 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3580 program. Exim now does this explicitly; it affects the formatting of
3581 timestamps using strftime().
3583 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3584 complaints. I've now initialized all the variables, to avoid this.
3586 32. Header lines added by a system filter were not being "seen" during
3587 transport-time rewrites.
3589 33. The info_callback() function passed to OpenSSL is set up with type void
3590 (*)(SSL *, int, int), as described somewhere. However, when calling the
3591 function (actually a macro) that sets it up, the type void(*)() is
3592 expected. I've put in a cast to prevent warnings from picky compilers.
3594 34. If a DNS black list lookup found a CNAME record, but there were no A
3595 records associated with the domain it pointed at, Exim crashed.
3597 35. If a DNS black list lookup returned more than one A record, Exim ignored
3598 all but the first. It now scans all returned addresses if a particular IP
3599 value is being sought. In this situation, the contents of the
3600 $dnslist_value variable are a list of all the addresses, separated by a
3603 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3604 to accept a host name and all its aliases if the forward lookup for any of
3605 them yielded the IP address of the incoming connection. Now it accepts only
3606 those names whose forward lookup yields the correct IP address. Any other
3607 names are discarded. This closes a loophole whereby a rogue DNS
3608 administrator could create reverse DNS records to break through a
3609 wildcarded host restriction in an ACL.
3611 37. If a user filter or a system filter that ran in a subprocess used any of
3612 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3613 the wrong values were passed to the pipe command ($thisaddress had the
3614 value of $0, $0 had the value of $1, etc). This bug was introduced by
3615 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3617 38. Improved the line breaking for long SMTP error messages from ACLs.
3618 Previously, if there was no break point between 40 and 75 characters, Exim
3619 left the rest of the message alone. Two changes have been made: (a) I've
3620 reduced the minimum length to 35 characters; (b) if it can't find a break
3621 point between 35 and 75 characters, it looks ahead and uses the first one
3622 that it finds. This may give the occasional overlong line, but at least the
3623 remaining text gets split now.
3625 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3626 file descriptors might be low, and that setting 1000 would always raise it.
3627 It turns out that in some environments, the limit is already over 1000 and
3628 that lowering it causes trouble. So now Exim takes care not to decrease it.
3630 40. When delivering a message, the value of $return_path is set to $sender_
3631 address at the start of routing (routers may change the value). By an
3632 oversight, this default was not being set up when an address was tested by
3633 -bt or -bv, which affected the outcome if any router or filter referred to
3636 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3637 log only when "message" or "log_message" are set. However, if one of the
3638 conditions was an address verification, or a call to a nested ACL, the
3639 messages generated by the underlying test were being passed through. This
3640 no longer happens. The underlying message is available in $acl_verify_
3641 message for both "message" and "log_message" expansions, so it can be
3642 passed through if needed.
3644 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3645 new expansion $bh_ to give the encoded byte string without charset
3646 translation. Translation happens only if iconv() is available; HAVE_ICONV
3647 indicates this at build time. HEADERS_CHARSET gives the charset to
3648 translate to; headers_charset can change it in the configuration, and
3649 "headers charset" can change it in an individual filter file.
3651 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3652 that creates RFC 2047 encoded "words" labels them as that charset instead
3653 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3654 expansion operator; (ii) when Exim creates a From: line for a local
3655 message; (iii) when a header line is rewritten to include a "phrase" part.
3657 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3658 buggy, causing it to skip the first lines of messages whose message ID
3659 ended in 'D'. This would not have bitten before Exim release 4.14, because
3660 message IDs were unlikely to end in 'D' before then. The effect was to have
3661 incorrect size information for certain domains.
3663 45. #include "config.h" was missing at the start of the crypt16.c module. This
3664 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3667 46. If there was a timeout during a "random" callout check, Exim treated it as
3668 a failure of the random address, and carried on sending RSET and the real
3669 address. If the delay was just some slowness somewhere, the response to the
3670 original RCPT would be taken as a response to RSET and so on, causing
3671 mayhem of various kinds.
3673 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3674 when I implemented it. It didn't allow for the fact that some option values
3675 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3676 the right test for positive values.
3678 48. Domain names in DNS records are case-independent. Exim always looks them up
3679 in lower case. Some resolvers return domain names in exactly the case they
3680 appear in the zone file, that is, they may contain uppercase letters. Not
3681 all resolvers do this - some return always lower case. Exim was treating a
3682 change of case by a resolver as a change of domain, similar to a widening
3683 of a domain abbreviation. This triggered its re-routing code and so it was
3684 trying to route what was effectively the same domain again. This normally
3685 caused routing to fail (because the router wouldn't handle the domain
3686 twice). Now Exim checks for this case specially, and just changes the
3687 casing of the domain that it ultimately uses when it transmits the message
3690 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3693 50. If a filter generated a file delivery with a non-absolute name (possible if
3694 no home directory exists for the router), the forbid_file option was not
3697 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3698 the existing equality matching.
3700 52. Exim was using ints instead of ino_t variables in some places where it was
3701 dealing with inode numbers.
3703 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3704 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3705 called TMPDIR, and if it finds it is different, it changes its value.
3707 54. The smtp_printf() function is now made available to local_scan() so
3708 additional output lines can be written before returning. There is also an
3709 smtp_fflush() function to enable the detection of a dropped connection.
3710 The variables smtp_input and smtp_batched_input are exported to
3713 55. Changed the default runtime configuration: the message "Unknown user"
3714 has been removed from the ACL, and instead placed on the localuser router,
3715 using the cannot_route_message feature. This means that any verification
3716 failures that generate their own messages won't get overridden. Similarly,
3717 the "Unrouteable address" message that was in the ACL for unverifiable
3718 relay addresses has also been removed.
3720 56. Added hosts_avoid_esmtp to the smtp transport.
3722 57. The exicyclog script was not checking for the esoteric option
3723 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3724 will work only if exicyclog is run under the appropriate euid.
3726 58. Following a discussion on the list, the rules by which Exim recognises line
3727 endings on incoming messages have been changed. The -dropcr and drop_cr
3728 options are now no-ops, retained only for backwards compatibility. The
3729 following line terminators are recognized: LF CRLF CR. However, special
3730 processing applies to CR:
3732 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3733 nor a local message in the state where . is a terminator.
3735 (ii) If a bare CR is encountered in a header line, an extra space is added
3736 after the line terminator so as not to end the header. The reasoning
3737 behind this is that bare CRs in header lines are most likely either
3738 to be mistakes, or people trying to play silly games.
3740 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3741 was being incorrectly given as 18 bytes larger than it should have been.
3742 This is a VOB (very old bug).
3744 60. This may never have affected anything current, but just in case it has:
3745 When the local host is found other than at the start of a list of hosts,
3746 the local host, those with the same MX, and any that follow, are discarded.
3747 When the list in question was part of a longer list of hosts, the following
3748 hosts (not currently being processed) were also being discarded. This no
3749 longer happens. I'm not sure if this situation could ever has previously
3752 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3755 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3756 header. This is something that is recommended in a new Internet Draft, and
3757 is something that is documented as being done by Sendmail. There are two
3758 possible values. For messages generated by the autoreply transport, Exim
3761 Auto-Submitted: auto-replied
3763 whereas for all other generated messages (e.g. bounces) it adds
3765 Auto-Submitted: auto-generated
3767 63. The "personal" condition in filters now includes a test for the
3768 Auto-Submitted: header. If it contains the string "auto-" the message it
3769 not considered personal.
3771 64. Added rcpt_include_affixes as a generic transport option.
3773 65. Added queue_only_override (default true).
3775 66. Added the syslog_duplication option.
3777 67. If what should have been the first header line of a message consisted of
3778 a space followed by a colon, Exim was mis-interpreting it as a header line.
3779 It isn't of course - it is syntactically invalid and should therefore be
3780 treated as the start of the message body. The misbehaviour could have
3781 caused a number of strange effects, including loss of data in subsequent
3782 header lines, and spool format errors.
3784 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3785 client host had authenticated. This control can now be exercised by an ACL
3786 for more flexibility.
3788 69. By default, callouts do not happen when testing with -bh. There is now a
3789 variant, -bhc, which does actually run the callout code, including
3790 consulting and updating the callout cache.
3792 70. Added support for saslauthd authentication, courtesy of Alexander
3795 71. If statvfs() failed on the spool or log directories while checking their
3796 size for availability, Exim confusingly gave the error "space shortage".
3797 Furthermore, in debugging mode it crashed with a floating point exception.
3798 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3799 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3800 serious problem, Exim now writes to the main and panic logs when this
3801 happens, with details of the failure. It then refuses to accept the
3802 incoming message, giving the message "spool directory problem" or "log
3803 directory problem" with a 421 code for SMTP messages.
3805 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3806 0, 1, and 2 exist, because some OS complain for execs without them (see
3807 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3808 descriptors. However, the code omitted to check that the open succeeded,
3809 causing mysterious errors if for some reason the permissions on /dev/null
3810 got screwed. Now Exim writes a message to the main and panic logs, and
3811 bombs out if it can't open /dev/null.
3813 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3814 interact so that it is all more flexible. It is supposed to remain
3815 backwards compatible. Also added extra_local_interfaces.
3817 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3818 to bomb out with an assertion failure - to the client this appears as a
3819 connection drop. This problem occurs in the part of the code that was taken
3820 from the Samba project. Fortunately, the assertion is in a very simple
3821 function, so I have fixed this by reproducing the function inline in the
3822 one place where it is called, and arranging for authentication to fail
3823 instead of killing the process with assert().
3825 75. The SPA client code was not working when the server requested OEM rather
3826 than Unicode encoding.
3828 76. Added code to make require_files with a specific uid setting more usable in
3829 the case where statting the file as root fails - usually a non-root-mounted
3830 NFS file system. When this happens and the failure is EACCES, Exim now
3831 forks a subprocess and does the per-uid checking as the relevant uid.
3833 77. Added process_log_path.
3835 78. If log_file_path was not explicitly set, a setting of check_log_space or
3836 check_log_inodes was ignored.
3838 79. If a space check for the spool or log partitions fails, the incident is now
3839 logged. Of course, in the latter case the data may get lost...
3841 80. Added the %p formatting code to string_format() so that it can be used to
3842 print addresses in debug_print(). Adjusted all the address printing in the
3843 debugging in store.c to use %p rather than %d.
3845 81. There was a concern that a line of code in smtp_in.c could overflow a
3846 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3847 initially expressed, the concern was not well-founded, because trailing
3848 spaces are removed early. However, if the trailing spaces were followed by
3849 a NULL, they did not get removed, so the overflow was possible. Two fixes
3852 (a) I re-wrote the offending code in a cleaner fashion.
3853 (b) If an incoming SMTP command contains a NULL character, it is rejected
3856 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3857 runs initgroups(), so that if the Exim user is in any additional groups,
3858 they will be used during message reception.
3864 The change log for 4.20 and earlier releases has been archived.