1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.344 2006/04/20 14:11:29 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
10 other effects) broke the use of negated acl sub-conditions.
12 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
15 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
16 "Deny" causes Exim to reject the incoming connection with a 554 error.
17 Unfortunately, if there is a major crisis, such as a disk failure,
18 tcp-wrappers gives "deny", whereas what one would like would be some
19 kind of temporary error. A kludge has been added to help with this.
20 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
21 554 error is used if errno is still zero or contains ENOENT (which occurs
22 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
25 PH/03 Add -lutil to the default FreeBSD LIBS setting.
27 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
28 errors. Otherwise a message that provokes a temporary error (when other
29 messages do not) can cause a whole host to time out.
31 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
32 addresses were routed directly to files or pipes from a redirect router.
33 File deliveries just didn't batch; pipe deliveries might have suffered
40 PH/01 The code for finding all the local interface addresses on a FreeBSD
41 system running IPv6 was broken. This may well have applied to all BSD
42 systems, as well as to others that have similar system calls. The broken
43 code found IPv4 interfaces correctly, but gave incorrect values for the
44 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
45 that it would not match correctly against @[] and not recognize the IPv6
48 PH/02 The ipliteral router was not recognizing addresses of the form user@
49 [ipv6:....] because it didn't know about the "ipv6:" prefix.
51 PH/03 Added disable_ipv6.
53 PH/04 Changed $reply_address to use the raw form of the headers instead of the
54 decoded form, because it is most often used to construct To: headers
55 lines in autoreplies, and the decoded form may well be syntactically
56 invalid. However, $reply_address has leading white space removed, and all
57 newlines turned into spaces so that the autoreply transport does not
60 PH/05 If group was specified without a user on a router, and no group or user
61 was specified on a transport, the group from the router was ignored.
63 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
64 for visible compile-time settings that can be used to change these
65 numbers, for those that want even more. Backwards compatibility with old
66 spool files has been maintained. However, going back to a previous Exim
67 release will lost any variables that are in spool files.
69 PH/07 Two small changes when running in the test harness: increase delay when
70 passing a TCP/IP connection to a new process, in case the original
71 process has to generate a bounce, and remove special handling of
72 127.0.0.2 (sic), which is no longer necessary.
74 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
75 be the same on different OS.
77 PH/09 Moved a debug statement in filter processing to avoid a race problem when
80 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
81 whether --show-vars was specified or not
83 JJ/02 exipick: Added support for new ACL variable spool format introduced
86 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
87 syntactically invalid From: or Reply-to: line, and a filter used this to
88 generate an autoreply, and therefore failed to obtain an address for the
89 autoreply, Exim could try to deliver to a non-existent relative file
90 name, causing unrelated and misleading errors. What now happens is that
91 it logs this as a hard delivery error, but does not attempt to create a
94 PH/11 The exinext utility has a -C option for testing purposes, but although
95 the given file was scanned by exinext itself; it wasn't being passed on
98 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
99 an end-of-file indication when reading a command response.
101 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
102 compiled. In many other places in Exim, IPv6 addresses are always
103 recognized, so I have changed this. It also means that IPv4 domain
104 literals of the form [IPV4:n.n.n.n] are now always recognized.
106 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
107 used if the router is not running as root, for example, when verifying at
108 ACL time, or when using -bh. The debugging output from this situation was
109 non-existent - all you got was a failure to exec. I have made two
112 (a) Failures to set uid/gid, the current directory, or a process leader
113 in a subprocess such as that created by queryprogram now generate
114 suitable debugging ouput when -d is set.
116 (b) The queryprogram router detects when it is not running as root,
117 outputs suitable debugging information if -d is set, and then runs
118 the subprocess without attempting to change uid/gid.
120 PH/15 Minor change to Makefile for building test_host (undocumented testing
123 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
124 additional section of a DNS packet that returns MX or SRV records.
125 Instead, it always explicitly searches for A/AAAA records. This avoids
126 major problems that occur when a DNS server includes only records of one
127 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
128 fixed another bug: if SRV records were looked up and the corresponding
129 address records were *not* found in the additional section, the port
130 values from the SRV records were lost.
132 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
133 using the correct key (the original address) when searching the retry
134 rules in order to find which one to use for generating the retry hint.
136 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
137 adding the default one. Similarly, if it contains a Reply-To: header, the
138 errors_reply_to option, if set, is not used.
140 PH/19 When calculating a retry time, Exim used to measure the "time since
141 failure" by looking at the "first failed" field in the retry record. Now
142 it does not use this if it is later than than the arrival time of the
143 message. Instead it uses the arrival time. This makes for better
144 behaviour in cases where some deliveries succeed, thus re-setting the
145 "first failed" field. An example is a quota failure for a huge message
146 when small messages continue to be delivered. Without this change, the
147 "time since failure" will always be short, possible causing more frequent
148 delivery attempts for the huge message than are intended.
149 [Note: This change was subsequently modified - see PH/04 for 4.62.]
151 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
152 $1, $2, $3) because the numerical variables can be reset during some
153 expansion items (e.g. "match"), thereby losing the authentication data.
155 PH/21 Make -bV show the size of off_t variables so that the test suite can
156 decide whether to run tests for quotas > 2G.
158 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
159 mailbox_size, and mailbox_filecount in the appendfile transport. If a
160 filecount value is greater than 2G or if a quota value is greater than 2G
161 on a system where the size of off_t is not greater than 4, a panic error
164 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
165 never match. The debug and -bh output now contains an explicit error
166 message indicating a malformed IPv4 address or mask.
168 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
169 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
172 PH/25 Do not write to syslog when running in the test harness. The only
173 occasion when this arises is a failure to open the main or panic logs
174 (for which there is an explicit test).
176 PH/26 Added the /no_tell option to "control=freeze".
178 PH/27 If a host name lookup failed very early in a connection, for example, if
179 the IP address matched host_lookup and the reverse lookup yielded a name
180 that did not have a forward lookup, an error message of the form "no IP
181 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
182 could be logged. Now it outputs the IP address instead of "NULL".
184 PH/28 An enabling patch from MH: add new function child_open_exim2() which
185 allows the sender and the authenticated sender to be set when
186 submitting a message from within Exim. Since child_open_exim() is
187 documented for local_scan(), the new function should be too.
189 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
190 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
191 results in an empty string is now treated as unset.
193 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
195 PH/31 Added sender_verify_fail logging option.
197 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
198 needed by RFC 822 but not by RFC 2822 was commented out. I have now
199 tidied the source and removed it altogether.
201 PH/33 When a queue run was abandoned because the load average was too high, a
202 log line was always written; now it is written only if the queue_run log
203 selector is set. In addition, the log line for abandonment now contains
204 information about the queue run such as the pid. This is always present
205 in "start" and "stop" lines but was omitted from the "abandon" line.
207 PH/34 Omit spaces between a header name and the colon in the error message that
208 is given when verify = headers_syntax fails (if there are lots of them,
209 the message gets confusing).
211 PH/35 Change the default for dns_check_names_pattern to allow slashes within
212 names, as there are now some PTR records that contain slashes. This check
213 is only to protect against broken name servers that fall over on strange
214 characters, so the fact that it applies to all lookups doesn't matter.
216 PH/36 Now that the new test suite is complete, we can remove some of the
217 special code in Exim that was needed for the old test suite. For example,
218 sorting DNS records because real resolvers return them in an arbitrary
219 order. The new test suite's fake resolver always returns records in the
222 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
223 bounces) except when queue_only is set, to avoid logging races between
224 the different processes.
226 PH/38 Panic-die if .include specifies a non-absolute path.
228 PH/39 A tweak to the "H" retry rule from its user.
230 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
231 a label. They prevented compilation on older perls.
233 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
234 a warning to be raised on newish perls.
236 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
237 on queue. Changes to match documented behaviour of showing count of
238 messages matching specified criteria.
240 PH/40 Changed the default ident timeout from 30s to 5s.
242 PH/41 Added support for the use of login_cap features, on those BSD systems
243 that have them, for controlling the resources used by pipe deliveries.
245 PH/42 The content-scanning code uses fopen() to create files in which to put
246 message data. Previously it was not paying any attention to the mode of
247 the files. Exim runs with umask(0) because the rest of the code creates
248 files with open(), and sets the required mode explicitly. Thus, these
249 files were ending up world-writeable. This was not a big issue, because,
250 being within the spool directory, they were not world-accessible. I have
251 created a function called modefopen, which takes an additional mode
252 argument. It sets umask(777), creates the file, chmods it to the required
253 mode, then resets the umask. All the relevant calls to fopen() in the
254 content scanning code have been changed to use this function.
256 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
257 to 24 hours. This avoids potential overflow problems when processing G
258 and H retry rules. I suspect nobody ever tinkers with this value.
260 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
262 PH/45 When the plaintext authenticator is running as a client, the server's
263 challenges are checked to ensure they are valid base64 strings. By
264 default, the authentication attempt is cancelled if an invalid string is
265 received. Setting client_ignore_invalid_base64 true ignores these errors.
266 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
267 they are received. Thus, the responses can be made to depend on the
268 challenges. If an invalid string is ignored, an empty string is placed in
271 PH/46 Messages that are created by the autoreply transport now contains a
272 References: header, in accordance with RFCs 2822 and 3834.
274 PH/47 Added authenticated_sender_force to the smtp transport.
276 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
278 PH/49 Installed latest patch from the Sieve maintainer.
280 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
281 was also set, the appendfile transport was unnecessarily scanning a
282 directory of message files (e.g. for maildir delivery) to find the count
283 of files (along with the size), even though it did not need this
284 information. It now does the scan only if it needs to find either the
285 size of the count of files.
287 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
289 PH/52 Two bugs concerned with error handling when the smtp transport is
292 (i) Exim was not creating retry information for temporary errors given
293 for individual recipients after the DATA command when the smtp transport
294 was used in LMTP mode. This meant that they could be retried too
295 frequently, and not timed out correctly.
297 (ii) Exim was setting the flag that allows error details to be returned
298 for LMTP errors on RCPT commands, but not for LMTP errors for individual
299 recipients that were returned after the DATA command.
301 PH/53 This is related to PH/52, but is more general: for any failing address,
302 when detailed error information was permitted to be returned to the
303 sender, but the error was temporary, then after the final timeout, only
304 "retry timeout exceeded" was returned. Now it returns the full error as
305 well as "retry timeout exceeded".
307 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
308 do this, and (what is worse) MTAs that accept it.
310 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
311 will now be deprecated.
313 PH/56 New os.c-cygwin from the Cygwin maintainer.
315 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
316 formats (previously only available in exim formats via -bpr, -bpru,
317 and -bpra. Now also available in native and exiqgrep formats)
319 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
320 with very large, slow to parse queues
322 JJ/08 exipick: added ! as generic prefix to negate any criteria format
324 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
326 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
327 responses to authentication challenges, though it was showing the
328 challenges; (ii) I've removed the CR characters from the debug output for
331 PH/58 Allow for the insertion of a newline as well as a space when a string
332 is turned into more than one encoded-word during RFC 2047 encoding. The
333 Sieve code now uses this.
335 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
336 data_4xx, lost_connection, tls_required.
338 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
339 message was being sent as an SMTP response.
341 PH/61 Add -l and -k options to exicyclog.
343 PH/62 When verifying, if an address was redirected to one new address, so that
344 verification continued, and the new address failed or deferred after
345 having set something in $address_data, the value of $address_data was not
346 passed back to the ACL. This was different to the case when no
347 redirection occurred. The value is now passed back in both cases.
349 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
350 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
351 use login_cap.h, so on its own it isn't the distinguishing feature. The
352 new name refers directly to the setclassresources() function.
354 PH/65 Added configuration files for NetBSD3.
356 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
358 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
359 is preferred over IPv4.
361 PH/68 The bounce_return_message and bounce_return_body options were not being
362 honoured for bounces generated during the reception of non-SMTP messages.
363 In particular, this applied to messages rejected by the ACL. This bug has
364 been fixed. However, if bounce_return_message is true and bounce_return_
365 body is false, the headers that are returned for a non-SMTP message
366 include only those that have been read before the error was detected.
367 (In the case of an ACL rejection, they have all been read.)
369 PH/69 The HTML version of the specification is now built in a directory called
370 spec_html instead of spec.html, because the latter looks like a path with
371 a MIME-type, and this confuses some software.
373 PH/70 Catch two compiler warnings in sieve.c.
375 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
376 function verify_get_ident() calls ip_connect() to connect a socket, but
377 if the "connect()" function timed out, ip_connect() used to close the
378 socket. However, verify_get_ident() also closes the socket later, and in
379 between Exim writes to the log, which may get opened at this point. When
380 the socket was closed in ip_connect(), the log could get the same file
381 descriptor number as the socket. This naturally causes chaos. The fix is
382 not to close the socket in ip_connect(); the socket should be closed by
383 the function that creates it. There was only one place in the code where
384 this was missing, in the iplookup router, which I don't think anybody now
385 uses, but I've fixed it anyway.
387 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
388 well as to direct DNS lookups. Otherwise the handling of names in host
389 lists is inconsistent and therefore confusing.
395 PH/01 Two changes to the default runtime configuration:
397 (1) Move the checks for relay_from_hosts and authenticated clients from
398 after to before the (commented out) DNS black list checks.
400 (2) Add control=submission to the relay_from_hosts and authenticated
401 clients checks, on the grounds that messages accepted by these
402 statements are most likely to be submissions.
404 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
406 (1) Generate an error if the third argument for the ${prvs expansion is
409 (2) Treat a missing third argument of ${prvscheck as if it were an empty
412 (3) Reset the variables that are obtained from the first argument of
413 ${prvscheck and used in the second argument before leaving the code,
414 because their memory is reclaimed, so using them afterwards may do
417 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
418 one (it's much easier than Tom thought :-).
420 (5) Because of (4), we can now allow for the use of $prvscheck_result
421 inside the third argument.
423 PH/03 For some reason, the default setting of PATH when running a command from
424 a pipe transport was just "/usr/bin". I have changed it to
427 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
428 anything to be listed in the output from -bV.
430 PH/05 When a filter generated an autoreply, the entire To: header line was
431 quoted in the delivery log line, like this:
433 => >A.N.Other <ano@some.domain> <original@ddress> ...
435 This has been changed so that it extracts the operative address. There
436 may be more than one such address. If so, they are comma-separated, like
439 => >ano@some.domain,ona@other.domain <original@ddress> ...
441 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
442 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
443 not being looked up in the rDNS to get a host name, Exim was showing the
444 IP address twice in Received: lines, even though the IP addresses were
445 identical. For example:
447 Received: from [1.2.3.4] (helo=[1.2.3.4])
449 However, if the real host name was known, it was omitting the HELO data
450 if it matched the actual IP address. This has been tidied up so that it
451 doesn't show the same IP address twice.
453 PH/07 When both +timestamp and +memory debugging was on, the value given by
454 $tod_xxx expansions could be wrong, because the tod_stamp() function was
455 called by the debug printing, thereby overwriting the timestamp buffer.
456 Debugging no longer uses the tod_stamp() function when +timestamp is set.
458 PH/08 When the original message was included in an autoreply transport, it
459 always said "this is a copy of the message, including all the headers",
460 even if body_only or headers_only was set. It now gives an appropriate
463 PH/09 Applied a patch from the Sieve maintainer which:
465 o fixes some comments
466 o adds the (disabled) notify extension core
467 o adds some debug output for the result of if/elsif tests
468 o points to the current vacation draft in the documentation
469 and documents the missing references header update
473 o fixes a bug in processing the envelope test (when testing
474 multiple envelope elements, the last element determinted the
477 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
478 Electronic Mail") by including:
480 Auto-submitted: auto-generated
482 in the messages that it generates (bounce messages and others, such as
483 warnings). In the case of bounce messages for non-SMTP mesages, there was
484 also a typo: it was using "Auto_submitted" (underscore instead of
485 hyphen). Since every message generated by Exim is necessarily in response
486 to another message, thes have all been changed to:
488 Auto-Submitted: auto-replied
490 in accordance with these statements in the RFC:
492 The auto-replied keyword:
494 - SHOULD be used on messages sent in direct response to another
495 message by an automatic process,
497 - MUST NOT be used on manually-generated messages,
499 - MAY be used on Delivery Status Notifications (DSNs) and Message
500 Disposition Notifications (MDNs),
502 - MUST NOT be used on messages generated by automatic or periodic
503 processes, except for messages which are automatic responses to
506 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
507 to the default Received: header definition.
509 PH/12 Added log selector acl_warn_skipped (default on).
511 PH/13 After a successful wildlsearch lookup, discard the values of numeric
512 variables because (a) they are in the wrong storage pool and (b) even if
513 they were copied, it wouldn't work properly because of the caching.
515 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
516 checking when decoding. Apparently there are clients that generate
517 overlong encoded strings. Why am I not surprised?
519 PH/15 If the first argument of "${if match_address" was not empty, but did not
520 contain an "@" character, Exim crashed. Now it writes a panic log message
521 and treats the condition as false.
523 PH/16 In autoreply, treat an empty string for "once" the same as unset.
525 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
526 extension "envelope-auth". The code is finished and in agreement with
527 other implementations, but there is no documentation so far and in fact,
528 nobody wrote the draft yet. This extension is currently #undef'ed, thus
529 not changing the active code.
531 Print executed "if" and "elsif" statements when debugging is used. This
532 helps a great deal to understand what a filter does.
534 Document more things not specified clearly in RFC3028. I had all this
535 sorted out, when out of a sudden new issues came to my mind. Oops."
537 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
540 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
541 canonical form (as documented). However, after a host name lookup from
542 the IP address, check_host() was doing a simple string comparison with
543 addresses acquired from the DNS when checking that the found name did
544 have the original IP as one of its addresses. Since any found IPv6
545 addresses are likely to be in abbreviated form, the comparison could
546 fail. Luckily, there already exists a function for doing the comparison
547 by converting both addresses to binary, so now that is used instead of
550 PH/20 There was another similar case to PH/19, when a complete host name was
551 given in a host list; looking up its IP address could give an abbreviated
552 form, whereas the current host's name might or might not be abbreviated.
553 The same fix has been applied.
559 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
560 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
563 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
564 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
566 PH/03 Typo: missing ".o" in src/pcre/Makefile.
568 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
569 header line, restrict the check to what is listed in RFCs 2369 and 2929.
570 Also, for "Auto-Submitted", treat anything other than "no" as
571 non-personal, in accordance with RFC 3834. (Previously it treated
572 anything starting "auto-" as non-personal.)
574 TF/01 The control=submission/name=... option had a problem with syntax
575 errors if the name included a slash character. The /name= option
576 now slurps the rest of the string, so it can include any characters
577 but it must come last in the list of options (after /sender_retain
580 PH/05 Some modifications to the interface to the fake nameserver for the new
588 TK/01 Added the "success_on_redirect" address verification option. See
589 NewStuff for rationale and an example.
591 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
593 PH/02 Patch to exigrep to allow it to work on syslog lines.
595 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
596 fread() to skip over the body file's header line, because in Cygwin the
597 header line is locked and is inaccessible.
599 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
600 co-exist for some time) to make it clear that it is the Exim ID that is
601 referenced, not the Message-ID: header line.
603 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
604 string_format() function, because snprintf() does not exist on all
607 PH/06 The use of forbid_filter_existstest now also locks out the use of the
608 ${stat: expansion item.
610 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
611 protocol synchronization error", to keep the pedants happy.
613 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
614 well as for IRIX systems, when gcc is being used. See the host.c source
617 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
619 PH/10 Named domain lists were not working if used in a queue_smtp_domains
622 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
623 transport and to the smtp transport in LMTP mode.
625 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
627 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
628 run a filter in a subprocess. This could lead to confusion in subsequent
629 lookups in the parent process. There should also be a search_tidyup() at
630 the end of the subprocess.
632 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
633 only if the host matched helo_try_verify_hosts, which caused the
634 verification to occur when the EHLO/HELO command was issued. The ACL just
635 tested the remembered result. Now, if a previous verification attempt has
636 not happened, "verify = helo" does it there and then.
638 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
640 TK/03 Fix log output including CR from clamd.
642 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
643 exist provoked a memory error which could cause a segfault.
645 PH/15 Installed PCRE 6.2
647 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
649 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
650 of the problem. Specifically, suggested +O2 rather than +O1 for the
653 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
655 PH/20 If a delivery was routed to a non-standard port by means of an SRV
656 record, the port was not correctly logged when the outgoing_port log
657 selector was set (it logged the transort's default port).
659 PH/21 Added support for host-specific ports to manualroute, queryprogram,
660 fallback_hosts, and "hosts" in the smtp transport.
662 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
663 host errors such as "Connection refused".
665 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
666 authentication with radiusclient 0.4.9:
668 - Error returned from rc_read_config was caught wrongly
669 - Username/password not passed on to radius server due to wrong length.
671 The presumption is that some radiusclient API changes for 4.51/PH/17
672 were not taken care of correctly. The code is still untested by me (my
673 Linux distribution still has 0.3.2 of radiusclient), but it was
674 contributed by a Radius user.
676 PH/24 When doing a callout, the value of $domain wasn't set correctly when
677 expanding the "port" option of the smtp transport.
679 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
680 while reading a MIME header. Thanks to Tom Hughes for a patch.
682 PH/24 Include config.h inside local_scan.h so that configuration settings are
685 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
686 that in an ACL for RCPT (for example), you can examine exactly what was
689 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
690 commands, but it was not correctly comparing the address with the actual
691 client host address. Thus, it would show the EHLO address in Received:
692 header lines when this was not necessary.
694 PH/27 Added the % operator to ${eval:}.
696 PH/28 Exim tries to create and chdir to its spool directory when it starts;
697 it should be ignoring failures (because with -C, for example, it has lost
698 privilege). It wasn't ignoring creation failures other than "already
701 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
704 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
705 because an input line was too long, either on its own, or by virtue of
706 too many continuations, the temporary file was not being removed, and the
707 return code was incorrect.
709 PH/31 Missing "BOOL" in function definition in filtertest.c.
711 PH/32 Applied Sieve patches from the maintainer.
713 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
715 PH/33 Added "verify = not_blind".
717 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
718 Local/Makefile (with some defaults set). These are used in built scripts
719 such as exicyclog, but they have never been used in the exim_install
720 script (though there are many overriding facilities there). I have
721 arranged that the exim_install script now takes note of these two
724 PH/35 Installed configuration files for Dragonfly.
726 PH/36 When a locally submitted message by a trusted user did not contain a
727 From: header, and the sender address was obtained from -f or from an SMTP
728 MAIL command, and the trusted user did not use -F to supply a sender
729 name, $originator_name was incorrectly used when constructing a From:
730 header. Furthermore, $originator_name was used for submission mode
731 messages from external hosts without From: headers in a similar way,
732 which is clearly wrong.
734 PH/37 Added control=suppress_local_fixups.
736 PH/38 When log_selector = +received_sender was set, and the addition of the
737 sender made the log line's construction buffer exactly full, or one byte
738 less than full, an overflow happened when the terminating "\n" was
741 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
742 when the result of a list match is failure because a DNS lookup failed.
744 PH/40 RM_COMMAND is now used in the building process.
746 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
747 the "build-* directories that it finds.
749 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
750 address in a domain literal was a prefix of an interface address.
752 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
753 when verifying a sender address, unless rewrite_headers is false.
755 PH/44 Wrote a long comment about why errors_to addresses are verified as
756 recipients, not senders.
758 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
759 the ratelimit ACL was added.
761 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
763 PH/46 Added extra information about PostgreSQL errors to the error string.
765 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
766 avoiding the need to install special zones in a real server. This is
767 backwards compatible; if it can't find the fake resolver, it drops back.
768 Thus, both old and new test suites can be run.
770 TF/02 Added util/ratelimit.pl
772 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
773 clock is set back in time.
775 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
776 Candler <B.Candler@pobox.com>.
778 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
779 OK for addresses that are the result of redirections.
781 PH/48 A number of further additions for the benefit of the new test suite,
782 including a fake gethostbyname() that interfaces to the fake DNS resolver
785 TF/06 The fix for widen_domains has also been applied to qualify_single and
786 search_parents which are the other dnslookup options that can cause
789 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
792 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
794 TF/07 Exim produced the error message "an SRV record indicated no SMTP
795 service" if it encountered an MX record with an empty target hostname.
796 The message is now "an MX or SRV record indicated no SMTP service".
798 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
799 if the DNS of the sending site is misconfigured. This is quite a
800 common situation. This change restores the behaviour of treating a
801 helo verification defer as a failure.
803 PH/51 If self=fail was set on a router, the bounce message did not include the
804 actual error message.
810 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
812 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
813 command itself ended in error, the underlying message about the transport
814 filter timeout was being overwritten with the pipe command error. Now the
815 underlying error message should be appended to the second error message.
817 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
819 PH/02 Reduce the amount of output that "make" produces by default. Full output
820 can still be requested.
822 PH/03 The warning log line about a condition test deferring for a "warn" verb
823 was being output only once per connection, rather than after each
824 occurrence (because it was using the same function as for successful
825 "warn" verbs). This seems wrong, so I have changed it.
827 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
828 it should not have, which might have caused a crash in the right
829 circumstances, but probably never did.
831 PH/04 Installed a modified version of Tony Finch's patch to make submission
832 mode fix the return path as well as the Sender: header line, and to
833 add a /name= option so that you can make the user's friendly name appear
836 TF/03 Added the control = fakedefer ACL modifier.
838 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
839 Mark Lowes for thorough testing.
841 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
843 TK/03 Merged latest SRS patch from Miles Wilton.
845 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
846 with the definition in sysexits.h (which is #included earlier).
847 Fortunately, Exim does not actually use EX_OK. The code used to try to
848 preserve the sysexits.h value, by assumimg that macro definitions were
849 scanned for macro replacements. I have been disabused of this notion,
850 so now the code just undefines EX_OK before #including unistd.h.
852 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
853 in the smtp transport. When a block could not be written in a single
854 write() function, the timeout was being re-applied to each part-write.
855 This seems wrong - if the receiver was accepting one byte at a time it
856 would take for ever. The timeout is now adjusted when this happens. It
857 doesn't have to be particularly precise.
859 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
860 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
862 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
863 without a domain if the check to <postmaster@domain> fails.
865 SC/01 Eximstats: added -xls and the ability to specify output files
866 (patch written by Frank Heydlauf).
868 SC/02 Eximstats: use FileHandles for outputing results.
870 SC/03 Eximstats: allow any combination of xls, txt, and html output.
872 SC/04 Eximstats: fixed display of large numbers with -nvr option
874 SC/05 Eximstats: fixed merging of reports with empty tables.
876 SC/06 Eximstats: added the -include_original_destination flag
878 SC/07 Eximstats: removed tabs and trailing whitespace.
880 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
882 TK/06 MBOX spool code: Add real "From " MBOX separator line
883 so the .eml file is really in mbox format (even though
884 most programs do not really care). Patch from Alex Miller.
886 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
887 The latter is generated from $received_to and is only set if the
888 message has one envelope recipient. SA can use these headers,
889 obviously out-of-the-box. Patch from Alex Miller.
891 PH/08 The ${def test on a variable was returning false if the variable's
892 value was "0", contrary to what the specification has always said!
893 The result should be true unless the variable is empty.
895 PH/09 The syntax error of a character other than { following "${if
896 def:variable_name" (after optional whitespace) was not being diagnosed.
897 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
898 accidental colon was present, for example, could give incorrect results.
900 PH/10 Tidied the code in a number of places where the st_size field of a stat()
901 result is used (not including appendfile, where other changes are about
904 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
905 This involved changing a lot of size variables from int to off_t. It
906 should work with maildirs and everything.
908 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
909 spamd dying while we are connected to it.
911 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
914 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
915 function for BDB 4.3.
917 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
919 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
920 the sources for PCRE have been re-arranged and more files are now
923 PH/15 The code I had for printing potentially long long variables in PH/11
924 above was not the best (it lost precision). The length of off_t variables
925 is now inspected at build time, and an appropriate printing format (%ld
926 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
927 to be "long long int" or "long int". This is needed for the internal
928 formatting function string_vformat().
930 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
931 the configuration file to be ":syslog", then the script "guesses" where
932 the logs files are, rather than using the compiled in default. In our
933 case the guess is not the same as the compiled default, so the script
934 suddenly stopped working when I started to use syslog. The patch checks
935 to see if log_file_path is "". If so, it attempts to read it from exim
936 with no configuration file to get the compiled in version, before it
937 falls back to the previous guessing code."
939 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
940 implementing BATV in an Exim configuration. See NewStuff for the gory
943 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
944 Makefile that are specific to HP-UX.
946 PH/18 If the "use_postmaster" option was set for a recipient callout together
947 with the "random" option, the postmaster address was used as the MAIL
948 FROM address for the random test, but not for the subsequent recipient
949 test. It is now used for both.
951 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
952 patch removes a few documentation additions to RFC 3028, because the
953 latest draft now contains them. It adds the new en;ascii-case comparator
954 and a new error check for 8bit text in MIME parts. Comparator and
955 require names are now matched exactly. I enabled the subaddress
956 extension, but it is not well tested yet (read: it works for me)."
958 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
959 rework some of the code of TK/09 above to avoid the hardwired use of
960 "%lld" and "long long". Replaced the call to snprintf() with a call to
963 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
964 records point to non-existent hosts", "retry timeout exceeded", and
965 "retry time not reached for any host after a long failure period".
967 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
968 experimental DomainKeys support:
970 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
971 (2) On an error such as an illegally used "control", the wrong name for
972 the control was given.
974 These problems did NOT occur unless DomainKeys support was compiled.
976 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
978 PH/24 Added ${if match_ip condition.
980 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
981 databases so that it will be absolutely obvious if a crash occurs in the
982 DB library. This is a regular occurrence (often caused by mis-matched
985 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
986 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
987 picked up on a user's system that detects such things. There doesn't seem
988 to be a gcc warning option for this - only an attribute that has to be
989 put on the function's prototype. It seems that in Fedora Core 4 they have
990 set this on a number of new functions. No doubt there will be more in due
993 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
994 specify a transport. However, if an address that was verified by such a
995 router was the subject of a callout, Exim crashed because it tried to
996 read the rcpt_include_affixes from the non-existent transport. Now it
997 just assumes that the setting of that option is false. This bug was
998 introduced by 4.51/PH/31.
1000 PH/28 Changed -d+all to exclude +memory, because that information is very
1001 rarely of interest, but it makes the output a lot bigger. People tend to
1002 do -d+all out of habit.
1004 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1005 code in os-type was giving problems when libc.so lives in lib64, like on
1008 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1009 aren't the modern standard, and it seems that some systems' include files
1010 don't always have them. Exim was already checking for some of the newer
1011 ones like T_AAAA, and defining it itself. I've added checks for all the
1012 record types that Exim uses.
1014 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1015 not automatically generating a new one, as it is supposed to. This
1016 prevented TLS from working. If the file did exist, but contained invalid
1017 data, a new version was generated, as expected. It was only the case of a
1018 non-existent file that was broken.
1020 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1021 with a change in libdomainkeys > 0.64.
1023 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1024 from DNS. If the selector record carries the flag, it now has
1025 precedence over the domain-wide flag.
1027 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1029 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1030 the use of a port name that isn't defined in /etc/services) occurred, the
1031 message was deferred as in a normal delivery, and thus remained on the
1032 spool, instead of being failed because of the mua_wrapper setting. This
1033 is now fixed, and I tidied up some of the mua_wrapper messages at the
1036 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1037 the messages in a hash of arrays rather than using individual hashes.
1038 This is a bit cleaner and results in dramatic memory savings, albeit
1039 at a slight CPU cost.
1041 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1042 as requested by Marc Sherman.
1044 SC/10 Eximstats: added histograms for user specified patterns as requested
1047 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1049 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1050 fopen() in the content-scanning modules that did not already have it.
1056 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1057 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1059 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1061 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1063 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1064 to test Sieve filters that use "vacation".
1066 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1067 that changes the way the GnuTLS parameters are stored in the cache file.
1068 The new format can be generated externally. For backward compatibility,
1069 if the data in the cache doesn't make sense, Exim assumes it has read an
1070 old-format file, and it generates new data and writes a new file. This
1071 means that you can't go back to an older release without removing the
1074 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1075 work if there are any delivery delays because "one_time" forces the
1076 parent to be marked "delivered", so its unseen clone is never tried
1077 again. For this reason, Exim now forbids the simultaneous setting of
1080 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1081 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1082 entry if you want to know the details. The fix, however, neglected to
1083 consider the case when local delivery batching is involved. The test for
1084 "previously delivered" was not happening when checking to see if an
1085 address could be batched with a previous (undelivered) one; under
1086 certain circumstances this could lead to multiple deliveries to the same
1089 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1090 in its include files, and this causes problems building Exim.
1092 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1093 header_syntax) but Exim was just ignoring anything given after a slash.
1094 In particular, this caused confusion with an attempt to use "verify =
1095 reverse_host_lookup/defer_ok". An error is now given when options are
1096 supplied for verify items that do not have them. (Maybe reverse_host_
1097 lookup should have a defer_ok option, but that's a different point.)
1099 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1100 defined by RFC 821) to 2048, because there were problems with some AUTH
1101 commands, and RFC 1869 says the size should be increased for extended
1102 SMTP commands that take arguments.
1104 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1107 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1108 "unknown" error; now it says that the functionality isn't in the binary.
1110 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1111 an address' error message when a string expansion fails (syntax or
1112 whatever). Otherwise the password may appear in the log. Following change
1113 PH/42 below, there is no longer a chance of it appearing in a bounce
1116 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1118 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1119 of its addresses was ever tried. (Bugzilla bug #2.)
1121 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1122 the result incorrectly in the debug output. (It correctly added a newline
1123 to what was transported.)
1125 TF/01 Added $received_time.
1127 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1128 commented out examples of how to interface to a virus scanner and to
1129 SpamAssassin. Also added commented examples of av_scanner and
1130 spamd_address settings.
1132 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1133 and controls are allowed in which ACLs. There were a couple of minor
1134 errors. Some of the entries in the conditions table (which is a table of
1135 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1136 negation of where the condition IS allowed.
1138 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1140 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1141 header file does not have a version number, so I've had to invent a new
1142 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1143 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1144 radiusclient), but it was contributed by a Radius user.
1146 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1147 files or directories, for OpenSSL.
1149 PH/19 When an Exim process that is running as root has to create an Exim log
1150 file, it does so in a subprocess that runs as exim:exim so as to get the
1151 ownership right at creation (otherwise, other Exim processes might see
1152 the file with the wrong ownership). There was no test for failure of this
1153 fork() call, which would lead to the process getting stuck as it waited
1154 for a non-existent subprocess. Forks do occasionally fail when resources
1155 run out. I reviewed all the other calls to fork(); they all seem to check
1158 PH/20 When checking for unexpected SMTP input at connect time (before writing
1159 the banner), Exim was not dealing correctly with a non-positive return
1160 from the read() function. If the client had disconnected by this time,
1161 the result was a log entry for a synchronization error with an empty
1162 string after "input=" when read() returned zero. If read() returned -1
1163 (an event I could not check), uninitialized data bytes were printed.
1164 There were reports of junk text (parts of files, etc) appearing after
1167 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1169 PH/22 Added support for macro redefinition, and (re)definition in between
1170 driver and ACL definitions.
1172 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1173 forgetting to use the resulting value; it was using the unexpanded value.
1175 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1176 hadn't been configured. The fix is from Juergen Kreileder, who
1177 understands it better than I do:
1179 "Here's what I see happening with three configured cyrus_sasl
1180 authenticators configured (plain, login, cram-md5):
1182 On startup auth_cyrus_sasl_init() gets called for each of these.
1183 This means three calls to sasl_listmech() without a specified mech_list.
1184 => SASL tests which mechs of all available mechs actually work
1185 => three warnings about OTP not working
1186 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1188 With the patch, sasl_listmech() also gets called three times. But now
1189 SASL's mech_list option is set to the server_mech specified in the the
1190 authenticator. Or in other words, the answer from sasl_listmech()
1191 gets limited to just the mech you're testing for (which is different
1193 => the return list contains just 'plain' or 'login', 'cram-md5' or
1194 nothing depending on the value of ob->server_mech.
1196 I've just tested the patch: Authentication still works fine,
1197 unavailable mechs specified in the exim configuration are still
1198 caught, and the auth.log warnings about OTP are gone."
1200 PH/25 When debugging is enabled, the contents of the command line are added
1201 to the debugging output, even when log_selector=+arguments is not
1204 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1205 answer is "GNU", and only if the return is "GNU/something" is the answer
1208 PH/27 $acl_verify_message is now set immediately after the failure of a
1209 verification in an ACL, and so is available in subsequent modifiers. In
1210 particular, the message can be preserved by coding like this:
1212 warn !verify = sender
1213 set acl_m0 = $acl_verify_message
1215 Previously, $acl_verify_message was set only while expanding "message"
1216 and "log_message" when a very denied access.
1218 PH/28 Modified OS/os.c-Linux with
1220 -#ifndef OS_LOAD_AVERAGE
1221 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1223 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1224 nomenclature these days.)
1226 PH/29 Installed patch from the Sieve maintainer that adds the options
1227 sieve_useraddress and sieve_subaddress to the redirect router.
1229 PH/30 In these circumstances:
1230 . Two addresses routed to the same list of hosts;
1231 . First host does not offer TLS;
1232 . First host accepts first address;
1233 . First host gives temporary error to second address;
1234 . Second host offers TLS and a TLS session is established;
1235 . Second host accepts second address.
1236 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1237 and peerdn, if requested) that were in fact used only for the second
1240 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1241 attention to any local part prefix or suffix that was matched by the
1242 router that accepted the address. It now behaves in the same way as it
1243 does for delivery: the affixes are removed from the local part unless
1244 rcpt_include_affixes is set on the transport.
1246 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1247 timeout during the DATA phase of an incoming message.
1249 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1250 applied a patch sanctioned by the Sieve maintainer.
1252 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1253 the uid or gid is negative. A case of a negative gid caused this to be
1254 noticed. The fix allows for either to be negative.
1256 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1257 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1260 PH/36 The change PH/12 above was broken. Fixed it.
1262 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1263 the grounds that routing the same address twice would always produce the
1264 same answer. This might have been true once, but it is certainly no
1265 longer true now. Routing a child address may depend on the previous
1266 routing that produced that child. Some complicated redirection strategies
1267 went wrong when messages had multiple recipients, and made Exim's
1268 behaviour dependent on the order in which the addresses were given.
1270 I have moved the duplicate checking until after the routing is complete.
1271 Exim scans the addresses that are assigned to local and remote
1272 transports, and removes any duplicates. This means that more work will be
1273 done, as duplicates will always all be routed, but duplicates are
1274 presumably rare, so I don't expect this is of any significance.
1276 For deliveries to pipes, files, and autoreplies, the duplicate checking
1277 still happens during the routing process, since they are not going to be
1280 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1281 It corrects a timeout issue with spamd. This is Ian's comment: "The
1282 background is that sometimes spamd either never reads data from a
1283 connection it has accepted, or it never writes response data. The exiscan
1284 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1285 blindly assumes that writes won't block so it may never time out."
1287 PH/39 Allow G after quota size as well as K and M.
1289 PH/40 The value set for $authenticated_id in an authenticator may not contain
1290 binary zeroes or newlines because the value is written to log lines and
1291 to spool files. There was no check on this. Now the value is run through
1292 the string_printing() function so that such characters are converted to
1293 printable escape sequences.
1295 PH/41 $message_linecount is a new variable that contains the total number of
1296 lines in the message. Compare $body_linecount, which is the count for the
1299 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1300 bounce and delay warning messages, except in certain special cases, which
1303 (a) An SMTP error message from a remote host;
1304 (b) A message specified in a :fail: redirection;
1305 (c) A message specified in a "fail" command in a system filter;
1306 (d) A message specified in a FAIL return from the queryprogram router;
1307 (e) A message specified by the cannot_route_message router option.
1309 In these cases only, Exim does include the error details in bounce and
1310 warning messages. There are also a few cases where bland messages such
1311 as "unrouteable address" or "local delivery error" are given.
1313 PH/43 $value is now also set for the "else" part of a ${run expansion.
1315 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1316 being worked on, but at least Exim now implements the latest version to
1319 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1320 process to complete was treated as a delivery failure, a timeout while
1321 writing the message to the pipe was logged, but erroneously treated as a
1322 successful delivery. Such timeouts include transport filter timeouts. For
1323 consistency with the overall process timeout, these timeouts are now
1324 treated as errors, giving rise to delivery failures by default. However,
1325 there is now a new Boolean option for the pipe transport called
1326 timeout_defer, which, if set TRUE, converts the failures into defers for
1327 both kinds of timeout. A transport filter timeout is now identified in
1330 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1331 systems where "make" and "gmake" are different, calling "gmake" at top
1332 level broke things. I've arranged for the value of $(MAKE) to be passed
1333 from the Makefile to this script so that it can call the same version of
1337 A note about Exim versions 4.44 and 4.50
1338 ----------------------------------------
1340 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1341 changes of various kinds. As a consequence, a big documentation update was
1342 needed. This delayed the release for rather longer than seemed good, especially
1343 in the light of a couple of (minor) security issues. Therefore, the changes
1344 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1345 release. So 4.44 and 4.50 are in effect two different branches that both start
1348 I have left the 4.50 change log unchanged; it contains all the changes since
1349 4.43. The change log for 4.44 is below; many of its items are identical to
1350 those for 4.50. This seems to be the most sensible way to preserve the
1351 historical information.
1357 1. Minor wording change to the doc/README.SIEVE file.
1359 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1360 computation of the current number of files was incorrect.
1362 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1363 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1366 4. Give more explanation in the error message when the command for a transport
1367 filter fails to execute.
1369 5. There are several places where Exim runs a non-Exim command in a
1370 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1371 was being done only for the command run by the queryprogram router. It is
1372 now done for all such subprocesses. The other cases are: ${run, transport
1373 filters, and the commands run by the lmtp and pipe transports.
1375 6. Added CONFIGURE_GROUP build-time option.
1377 7. Some older OS have a limit of 256 on the maximum number of file
1378 descriptors. Exim was using setrlimit() to set 1000 as a large value
1379 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1380 systems. I've change it so that if it can't get 1000, it tries for 256.
1382 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1383 was an oversight, and furthermore, ever since the addition of extra
1384 controls (e.g. 4.43/32), the checks on when to allow different forms of
1385 "control" were broken. There should now be diagnostics for all cases when a
1386 control that does not make sense is encountered.
1388 9. Added the /retain_sender option to "control=submission".
1390 10. $recipients is now available in the predata ACL (oversight).
1392 11. Tidy the search cache before the fork to do a delivery from a message
1393 received from the command line. Otherwise the child will trigger a lookup
1394 failure and thereby defer the delivery if it tries to use (for example) a
1395 cached ldap connection that the parent has called unbind on.
1397 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1398 of $address_data from the recipient verification was clobbered by the
1399 sender verification.
1401 13. The value of address_data from a sender verification is now available in
1402 $sender_address_data in subsequent conditions in the ACL statement.
1404 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1406 15. Added a new option "connect=<time>" to callout options, to set a different
1409 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1410 was its contents. (It was OK if the option was not defined at all.)
1412 17. A "Completed" log line is now written for messages that are removed from
1413 the spool by the -Mrm option.
1415 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1416 information about exactly what failed.
1418 19. Added -dd to debug only the daemon process.
1420 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1421 handles timeouts, both on the server side and network timeouts. Renamed the
1422 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1424 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1425 It is now set to "smtps".
1427 22. $host_address is now set to the target address during the checking of
1428 ignore_target_hosts.
1430 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1431 being passed; this would have caused $sender_host_name to have been used if
1432 matching the list had actually called for a host name (not very likely,
1433 since this list is usually IP addresses). A host name is now passed as
1436 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1437 code that specifies a non-restarting handler (typically sigaction() in
1438 modern systems) in an attempt to fix a rare and obscure crash bug.
1440 25. Narrowed the window for a race in the daemon that could cause it to ignore
1441 SIGCHLD signals. This is not a major problem, because they are used only to
1442 wake it up if nothing else does.
1444 26. A malformed maildirsize file could cause Exim to calculate negative values
1445 for the mailbox size or file count. Odd effects could occur as a result.
1446 The maildirsize information is now recalculated if the size or filecount
1449 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1450 support for a long time. Removed HAVE_SYS_VFS_H.
1452 28. Installed the latest version of exipick from John Jetmore.
1454 29. In an address list, if the pattern was not a regular expression, an empty
1455 subject address (from a bounce message) matched only if the pattern was an
1456 empty string. Non-empty patterns were not even tested. This was the wrong
1457 because it is perfectly reasonable to use an empty address as part of a
1458 database query. An empty address is now tested by patterns that are
1459 lookups. However, all the other forms of pattern expect the subject to
1460 contain a local part and a domain, and therefore, for them, an empty
1461 address still always fails if the pattern is not itself empty.
1463 30. Exim went into a mad DNS loop when attempting to do a callout where the
1464 host was specified on an smtp transport, and looking it up yielded more
1465 than one IP address.
1467 31. Re-factored the code for checking spool and log partition space into a
1468 function that finds that data and another that does the check. The former
1469 is then used to implement four new variables: $spool_space, $log_space,
1470 $spool_inodes, and $log_inodes.
1472 32. The RFC2047 encoding function was originally intended for short strings
1473 such as real names; it was not keeping to the 75-character limit for
1474 encoded words that the RFC imposes. It now respects the limit, and
1475 generates multiple encoded words if necessary. To be on the safe side, I
1476 have increased the buffer size for the ${rfc2047: expansion operator from
1479 33. It is now permitted to omit both strings after an "if" condition; if the
1480 condition is true, the result is "true". As before, when the second string
1481 is omitted, a false condition yields an empty string. This makes it less
1482 cumbersome to write custom ACL and router conditions.
1484 34. Failure to deliver a bounce message always caused it to be frozen, even if
1485 there was an errors_to setting on the router. The errors_to setting is now
1488 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1489 canonical form (fully expanded) before being placed in
1490 $sender_host_address.
1492 36. The table in the code that translates DNS record types into text (T_A to
1493 "A" for instance) was missing entries for NS and CNAME. It is just possible
1494 that this could have caused confusion if both these types were looked up
1495 for the same domain, because the text type is used as part of Exim's
1496 per-process caching. But the chance of anyone hitting this buglet seems
1499 37. The dnsdb lookup has been extended in a number of ways.
1501 (1) There is a new type, "zns", which walks up the domain tree until it
1502 finds some nameserver records. It should be used with care.
1504 (2) There is a new type, "mxh", which is like "mx" except that it returns
1505 just the host names, not the priorities.
1507 (3) It is now possible to give a list of domains (or IP addresses) to be
1508 looked up. The behaviour when one of the lookups defers can be
1509 controlled by a keyword.
1511 (4) It is now possible to specify the separator character for use when
1512 multiple records are returned.
1514 38. The dnslists ACL condition has been extended: it is now possible to supply
1515 a list of IP addresses and/or domains to be looked up in a particular DNS
1518 39. Added log_selector=+queue_time_overall.
1520 40. When running the queue in the test harness, wait just a tad after forking a
1521 delivery process, to get repeatability of debugging output.
1523 41. Include certificate and key file names in error message when GnuTLS fails
1524 to set them up, because the GnuTLS error message doesn't include the name
1525 of the failing file when there is a problem reading it.
1527 42. Allow both -bf and -bF in the same test run.
1529 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1531 44. The "Exiscan patch" is now merged into the mainline Exim source.
1533 45. Sometimes the final signoff response after QUIT could fail to get
1534 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1535 < 0 before doing a fflush(). This bug looks as though it goes back to the
1536 introduction of TLS in release 3.20, but "sometimes" must have been rare
1537 because the tests only now provoked it.
1539 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1540 (this can affect the format of dates).
1542 47. exim_tidydb, when checking for the continued existence of a message for
1543 which it has found a message-specific retry record, was not finding
1544 messages that were in split spool directories. Consequently, it was
1545 deleting retry records that should have stayed in existence.
1547 48. Steve fixed some bugs in eximstats.
1549 49. The SPA authentication driver was not abandoning authentication and moving
1550 on to the next authenticator when an expansion was forced to fail,
1551 contradicting the general specification for all authenticators. Instead it
1552 was generating a temporary error. It now behaves as specified.
1554 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1555 (the order specifies the preference for clients). The order is now AES256,
1556 AES128, 3DES, ARCFOUR128.
1558 51. Small patch to Sieve code - explicitly set From: when generating an
1561 52. Exim crashed if a remote delivery caused a very long error message to be
1562 recorded - for instance if somebody sent an entire SpamAssassin report back
1563 as a large number of 550 error lines. This bug was coincidentally fixed by
1564 increasing the size of one of Exim's internal buffers (big_buffer) that
1565 happened as part of the Exiscan merge. However, to be on the safe side, I
1566 have made the code more robust (and fixed the comments that describe what
1569 53. Now that there can be additional text after "Completed" in log lines (if
1570 the queue_time_overall log selector is set), a one-byte patch to exigrep
1571 was needed to allow it to recognize "Completed" as not the last thing in
1574 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1575 patch that reportedly fixes this has been added. I am not expert enough to
1576 create a test for it. This is what the patch creator wrote:
1578 "I found a little strange behaviour of ldap code when working with
1579 Windows 2003 AD Domain, where users was placed in more than one
1580 Organization Units. When I tried to give exim partial DN, the exit code
1581 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1582 But simultaneously result of request was absolutely normal ldap result,
1583 so I produce this patch..."
1585 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1586 REFERENCE, so I have modified the code to exclude the patch when that macro
1589 55. Some experimental protocols are using DNS PTR records for new purposes. The
1590 keys for these records are domain names, not reversed IP addresses. The
1591 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1592 leaves it alone. Component reversal etc. now happens only for IP addresses.
1595 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1597 57. Double the size of the debug message buffer (to 2048) so that more of very
1598 long debug lines gets shown.
1600 58. The exicyclog utility now does better if the number of log files to keep
1601 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1603 59. Two changes related to the smtp_active_hostname option:
1605 (1) $smtp_active_hostname is now available as a variable.
1606 (2) The default for smtp_banner uses $smtp_active_hostname instead
1607 of $primary_hostname.
1609 60. The host_aton() function is supposed to be passed a string that is known
1610 to be a valid IP address. However, in the case of IPv6 addresses, it was
1611 not checking this. This is a hostage to fortune. Exim now panics and dies
1612 if the condition is not met. A case was found where this could be provoked
1613 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1614 components; fortuitously, this particular loophole had already been fixed
1615 by change 4.50/55 above.
1617 If there are any other similar loopholes, the new check in host_aton()
1618 itself should stop them being exploited. The report I received stated that
1619 data on the command line could provoke the exploit when Exim was running as
1620 exim, but did not say which command line option was involved. All I could
1621 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1622 running as the user.
1625 61. There was a buffer overflow vulnerability in the SPA authentication code
1626 (which came originally from the Samba project). I have added a test to the
1627 spa_base64_to_bits() function which I hope fixes it.
1630 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1631 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1633 63. The daemon start-up calls getloadavg() while still root for those OS that
1634 need the first call to be done as root, but it missed one case: when
1635 deliver_queue_load_max is set with deliver_drop_privilege. This is
1636 necessary for the benefit of the queue runner, because there is no re-exec
1637 when deliver_drop_privilege is set.
1639 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1640 This has been fixed.
1642 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1643 was in use, was not putting the data itself into the right store pool;
1644 consequently, it could be overwritten for a subsequent message in the same
1645 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1648 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1650 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1651 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1652 the function were treating the return as a boolean value, which happened to
1653 work because 0=false and not-0=true, but is not correct code.
1655 68. The host_aton() function was not handling scoped IPv6 addresses (those
1656 with, for example, "%eth0" on the end) correctly.
1658 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1659 negated items (that is, ~something) in unsigned ints. Some compilers
1660 apparently mutter when there is no cast.
1662 70. If an address verification called from an ACL failed, and did not produce a
1663 user-specific message (i.e. there was only a "system" message), nothing was
1664 put in $acl_verify_message. In this situation, it now puts the system
1667 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1668 session; change 4.31/43 added the unwanted input to the log line - except
1669 that it did not do this in the start of session case. It now does.
1671 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1672 This is wrong and can cause the other end to generate a synchronization
1673 error if it is another Exim or anything else that does the synchronization
1674 check. A QUIT command is no longer sent after a timeout.
1676 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1677 during host lookups.
1679 74. The defer_ok option of callout verification was not working if it was used
1680 when verifying addresses in header lines, that is, for this case:
1682 verify = header_sender/callout=defer_ok
1684 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1685 those file descriptors could be used for SMTP connections. If anything
1686 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1687 could be sent to the SMTP client, causing chaos. The daemon now opens
1688 stdin, stdout, and stderr to /dev/null when it puts itself into the
1691 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1692 log by default. The user can override this with suitable Perl magic.
1694 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1695 the log message when discard triggers, was not working for the DATA ACL or
1696 for the non-SMTP ACL.
1698 78. Error message wording change in sieve.c.
1700 79. If smtp_accept_max_per_host was set, the number of connections could be
1701 restricted to fewer than expected, because the daemon was trying to set up
1702 a new connection before checking whether the processes handling previous
1703 connections had finished. The check for completed processes is now done
1704 earlier. On busy systems, this bug wouldn't be noticed because something
1705 else would have woken the daemon, and it would have reaped the completed
1708 80. If a message was submitted locally by a user whose login name contained one
1709 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1710 It caused a spool format error. I have fixed the spool reading code. A
1711 related problem was that the "from" clause in the Received: line became
1712 illegal because of the space(s). It is now covered by ${quote_local_part.
1714 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1717 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1719 83. Patch from Sieve maintainer to fix unterminated string problem in
1720 "vacation" handling.
1722 84. Some minor changes to the Linux configuration files to help with other
1723 OS variants using glibc.
1725 85. One more patch for Sieve to update vacation handling to latest spec.
1728 ----------------------------------------------------
1729 See the note above about the 4.44 and 4.50 releases.
1730 ----------------------------------------------------
1736 1. Change 4.43/35 introduced a bug that caused file counts to be
1737 incorrectly computed when quota_filecount was set in an appendfile
1740 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1741 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1744 3. Give more explanation in the error message when the command for a transport
1745 filter fails to execute.
1747 4. There are several places where Exim runs a non-Exim command in a
1748 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1749 was being done only for the command run by the queryprogram router. It is
1750 now done for all such subprocesses. The other cases are: ${run, transport
1751 filters, and the commands run by the lmtp and pipe transports.
1753 5. Some older OS have a limit of 256 on the maximum number of file
1754 descriptors. Exim was using setrlimit() to set 1000 as a large value
1755 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1756 systems. I've change it so that if it can't get 1000, it tries for 256.
1758 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1759 was an oversight, and furthermore, ever since the addition of extra
1760 controls (e.g. 4.43/32), the checks on when to allow different forms of
1761 "control" were broken. There should now be diagnostics for all cases when a
1762 control that does not make sense is encountered.
1764 7. $recipients is now available in the predata ACL (oversight).
1766 8. Tidy the search cache before the fork to do a delivery from a message
1767 received from the command line. Otherwise the child will trigger a lookup
1768 failure and thereby defer the delivery if it tries to use (for example) a
1769 cached ldap connection that the parent has called unbind on.
1771 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1772 of $address_data from the recipient verification was clobbered by the
1773 sender verification.
1775 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1776 was its contents. (It was OK if the option was not defined at all.)
1778 11. A "Completed" log line is now written for messages that are removed from
1779 the spool by the -Mrm option.
1781 12. $host_address is now set to the target address during the checking of
1782 ignore_target_hosts.
1784 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1785 being passed; this would have caused $sender_host_name to have been used if
1786 matching the list had actually called for a host name (not very likely,
1787 since this list is usually IP addresses). A host name is now passed as
1790 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1791 code that specifies a non-restarting handler (typically sigaction() in
1792 modern systems) in an attempt to fix a rare and obscure crash bug.
1794 15. Narrowed the window for a race in the daemon that could cause it to ignore
1795 SIGCHLD signals. This is not a major problem, because they are used only to
1796 wake it up if nothing else does.
1798 16. A malformed maildirsize file could cause Exim to calculate negative values
1799 for the mailbox size or file count. Odd effects could occur as a result.
1800 The maildirsize information is now recalculated if the size or filecount
1803 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1804 support for a long time. Removed HAVE_SYS_VFS_H.
1806 18. Updated exipick to current release from John Jetmore.
1808 19. Allow an empty sender to be matched against a lookup in an address list.
1809 Previously the only cases considered were a regular expression, or an
1812 20. Exim went into a mad DNS lookup loop when doing a callout where the
1813 host was specified on the transport, if the DNS lookup yielded more than
1816 21. The RFC2047 encoding function was originally intended for short strings
1817 such as real names; it was not keeping to the 75-character limit for
1818 encoded words that the RFC imposes. It now respects the limit, and
1819 generates multiple encoded words if necessary. To be on the safe side, I
1820 have increased the buffer size for the ${rfc2047: expansion operator from
1823 22. Failure to deliver a bounce message always caused it to be frozen, even if
1824 there was an errors_to setting on the router. The errors_to setting is now
1827 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1828 canonical form (fully expanded) before being placed in
1829 $sender_host_address.
1831 24. Updated eximstats to version 1.33
1833 25. Include certificate and key file names in error message when GnuTLS fails
1834 to set them up, because the GnuTLS error message doesn't include the name
1835 of the failing file when there is a problem reading it.
1837 26. Expand error message when OpenSSL has problems setting up cert/key files.
1840 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1841 (this can affect the format of dates).
1843 28. exim_tidydb, when checking for the continued existence of a message for
1844 which it has found a message-specific retry record, was not finding
1845 messages that were in split spool directories. Consequently, it was
1846 deleting retry records that should have stayed in existence.
1848 29. eximstats updated to version 1.35
1849 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1850 1.35 - bugfix such that pie charts by volume are generated correctly
1852 30. The SPA authentication driver was not abandoning authentication and moving
1853 on to the next authenticator when an expansion was forced to fail,
1854 contradicting the general specification for all authenticators. Instead it
1855 was generating a temporary error. It now behaves as specified.
1857 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1858 (the order specifies the preference for clients). The order is now AES256,
1859 AES128, 3DES, ARCFOUR128.
1861 31. Small patch to Sieve code - explicitly set From: when generating an
1864 32. Exim crashed if a remote delivery caused a very long error message to be
1865 recorded - for instance if somebody sent an entire SpamAssassin report back
1866 as a large number of 550 error lines. This bug was coincidentally fixed by
1867 increasing the size of one of Exim's internal buffers (big_buffer) that
1868 happened as part of the Exiscan merge. However, to be on the safe side, I
1869 have made the code more robust (and fixed the comments that describe what
1872 33. Some experimental protocols are using DNS PTR records for new purposes. The
1873 keys for these records are domain names, not reversed IP addresses. The
1874 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1875 leaves it alone. Component reversal etc. now happens only for IP addresses.
1878 34. The host_aton() function is supposed to be passed a string that is known
1879 to be a valid IP address. However, in the case of IPv6 addresses, it was
1880 not checking this. This is a hostage to fortune. Exim now panics and dies
1881 if the condition is not met. A case was found where this could be provoked
1882 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1883 components; fortuitously, this particular loophole had already been fixed
1884 by change 4.50/55 or 4.44/33 above.
1886 If there are any other similar loopholes, the new check in host_aton()
1887 itself should stop them being exploited. The report I received stated that
1888 data on the command line could provoke the exploit when Exim was running as
1889 exim, but did not say which command line option was involved. All I could
1890 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1891 running as the user.
1894 35. There was a buffer overflow vulnerability in the SPA authentication code
1895 (which came originally from the Samba project). I have added a test to the
1896 spa_base64_to_bits() function which I hope fixes it.
1899 36. The daemon start-up calls getloadavg() while still root for those OS that
1900 need the first call to be done as root, but it missed one case: when
1901 deliver_queue_load_max is set with deliver_drop_privilege. This is
1902 necessary for the benefit of the queue runner, because there is no re-exec
1903 when deliver_drop_privilege is set.
1905 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1906 was in use, was not putting the data itself into the right store pool;
1907 consequently, it could be overwritten for a subsequent message in the same
1908 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1911 38. Sometimes the final signoff response after QUIT could fail to get
1912 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1913 < 0 before doing a fflush(). This bug looks as though it goes back to the
1914 introduction of TLS in release 3.20, but "sometimes" must have been rare
1915 because the tests only now provoked it.
1921 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1922 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1923 Now it returns an integer. A number of calls were still expecting a T/F
1924 return. Fortuitously, in all cases, the tests worked in OK situations,
1925 which is the norm. However, things would have gone wrong on any write
1926 failures on the smtp file descriptor. This function is used when sending
1927 messages over SMTP and also when doing verify callouts.
1929 2. When Exim is called to do synchronous delivery of a locally submitted
1930 message (the -odf or -odi options), it no longer closes stderr before doing
1933 3. Implemented the mua_wrapper option.
1935 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1937 5. Implemented the functions header_remove(), header_testname(),
1938 header_add_at_position(), and receive_remove_recipient(), and exported them
1941 6. If an ACL "warn" statement specified the addition of headers, Exim already
1942 inserted X-ACL-Warn: at the start if there was no header name. However, it
1943 was not making this test for the second and subsequent header lines if
1944 there were newlines in the string. This meant that an invalid header could
1945 be inserted if Exim was badly configured.
1947 7. Allow an ACL "warn" statement to add header lines at the start or after all
1948 the Received: headers, as well as at the end.
1950 8. Added the rcpt_4xx retry error code.
1952 9. Added postmaster_mailfrom=xxx to callout verification option.
1954 10. Added mailfrom=xxxx to the callout verification option, for verify=
1957 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1958 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1960 12. Inserted some casts to stop certain compilers warning when using pointer
1961 differences as field lengths or precisions in printf-type calls (mostly
1962 affecting debugging statements).
1964 13. Added optional readline() support for -be (dynamically loaded).
1966 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1967 same clock tick as a message's arrival, so that its received time was the
1968 same as the "first fail" time on the retry record, and that message
1969 remained on the queue past the ultimate address timeout, every queue runner
1970 would try a delivery (because it was past the ultimate address timeout) but
1971 after another failure, the ultimate address timeout, which should have then
1972 bounced the address, did not kick in. This was a "< instead of <=" error;
1973 in most cases the first failure would have been in the next clock tick
1974 after the received time, and all would be well.
1976 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1977 being recognized when the domain list was tested by the match_domain
1978 condition in an expansion string.
1980 16. Added the ${str2b64: operator.
1982 17. Exim was always calling setrlimit() to set a large limit for the number of
1983 processes, without checking whether the existing limit was already
1984 adequate. (It did check for the limit on file descriptors.) Furthermore,
1985 errors from getrlimit() and setrlimit() were being ignored. Now they are
1986 logged to the main and panic logs, but Exim does carry on, to try to do its
1987 job under whatever limits there are.
1989 18. Imported PCRE 5.0.
1991 19. Trivial typo in log message " temporarily refused connection" (the leading
1994 20. If the log selector return_path_on_delivery was set and an address was
1995 redirected to /dev/null, the delivery process crashed because it assumed
1996 that a return path would always be set for a "successful" delivery. In this
1997 case, the whole delivery is bypassed as an optimization, and therefore no
2000 21. Internal re-arrangement: the function for sending a challenge and reading
2001 a response while authentication was assuming a zero-terminated challenge
2002 string. It's now changed to take a pointer and a length, to allow for
2003 binary data in such strings.
2005 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2007 23. Exim was not respecting finduser_retries when seeking the login of the
2008 uid under which it was called; it was always trying 10 times. (The default
2009 setting of finduser_retries is zero.) Also, it was sleeping after the final
2010 failure, which is pointless.
2012 24. Implemented tls_on_connect_ports.
2014 25. Implemented acl_smtp_predata.
2016 26. If the domain in control=submission is set empty, Exim assumes that the
2017 authenticated id is a complete email address when it generates From: or
2018 Sender: header lines.
2020 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2021 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2022 chown and chgrp in /bin and hostname in /usr/bin.
2024 28. Exim was keeping the "process log" file open after each use, just as it
2025 does for the main log. This opens the possibility of it remaining open for
2026 long periods when the USR1 signal hits a daemon. Occasional processlog
2027 errors were reported, that could have been caused by this. Anyway, it seems
2028 much more sensible not to leave this file open at all, so that is what now
2031 29. The long-running daemon process does not normally write to the log once it
2032 has entered its main loop, and it closes the log before doing so. This is
2033 so that log files can straightforwardly be renamed and moved. However,
2034 there are a couple of unusual error situations where the daemon does write
2035 log entries, and I had neglected to close the log afterwards.
2037 30. The text of an SMTP error response that was received during a remote
2038 delivery was being truncated at 512 bytes. This is too short for some of
2039 the long messages that one sometimes sees. I've increased the limit to
2042 31. It is now possible to make retry rules that apply only when a message has a
2043 specific sender, in particular, an empty sender.
2045 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2046 it possible to be selective about when SMTP synchronization is enforced.
2048 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2050 32. Implemented hosts_connection_nolog.
2052 33. Added an ACL for QUIT.
2054 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2057 35. Added mailbox_size and mailbox_filecount to appendfile.
2059 36. Added control = no_multiline_responses to ACLs.
2061 37. There was a bug in the logic of the code that waits for the clock to tick
2062 in the case where the clock went backwards by a substantial amount such
2063 that the microsecond fraction of "now" was more than the microsecond
2064 fraction of "then" (but the whole seconds number was less).
2066 38. Added support for the libradius Radius client library this is found on
2067 FreeBSD (previously only the radiusclient library was supported).
2073 1. When certain lookups returned multiple values in the form name=value, the
2074 quoting of the values was not always being done properly. Specifically:
2075 (a) If the value started with a double quote, but contained no whitespace,
2077 (b) If the value contained whitespace other than a space character (i.e.
2078 tabs or newlines or carriage returns) it was not quoted.
2079 This fix has been applied to the mysql and pgsql lookups by writing a
2080 separate quoting function and calling it from the lookup code. The fix
2081 should probably also be applied to nisplus, ibase and oracle lookups, but
2082 since I cannot test any of those, I have not disturbed their existing code.
2084 2. A hit in the callout cache for a specific address caused a log line with no
2085 reason for rejecting RCPT. Now it says "Previous (cached) callout
2086 verification failure".
2088 3. There was an off-by-one bug in the queryprogram router. An over-long
2089 return line was truncated at 256 instead of 255 characters, thereby
2090 overflowing its buffer with the terminating zero. As well as fixing this, I
2091 have increased the buffer size to 1024 (and made a note to document this).
2093 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2094 when Exim is waiting for an SMTP response from a remote server, Exim
2095 restarts its select() call on the socket, thereby resetting its timeout.
2096 This is not a problem when such interrupts are rare. Somebody set up a cron
2097 job to run exiwhat every 2 minutes, which is less than the normal select()
2098 timeout (5 or 10 minutes). This meant that the select() timeout never
2099 kicked in because it was always reset. I have fixed this by comparing the
2100 time when an interrupt arrives with the time at the start of the first call
2101 to select(). If more time than the timeout has elapsed, the interrupt is
2102 treated as a timeout.
2104 5. Some internal re-factoring in preparation for the addition of Sieve
2105 extensions (by MH). In particular, the "personal" test is moved to a
2106 separate function, and given an option for scanning Cc: and Bcc: (which is
2107 not set for Exim filters).
2109 6. When Exim created an email address using the login of the caller as the
2110 local part (e.g. when creating a From: or Sender: header line), it was not
2111 quoting the local part when it contained special characters such as @.
2113 7. Installed new OpenBSD configuration files.
2115 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2116 try to make them clearer.
2118 9. Callout options, other than the timeout value, were being ignored when
2119 verifying sender addresses in header lines. For example, when using
2121 verify = header_sender/callout=no_cache
2123 the cache was (incorrectly) being used.
2125 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2126 only the Cygwin environment.
2128 11. When return_path_on_delivery was set as a log selector, if different remote
2129 addresses in the same message used different return paths and parallel
2130 remote delivery occurred, the wrong values would sometimes be logged.
2131 (Whenever a remote delivery process finished, the return path value from
2132 the most recently started remote delivery process was logged.)
2134 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2135 lines when AUTH and/or TLS are in use. This is the "received protocol"
2136 field. Exim used to use "asmtp" for authenticated SMTP, without any
2137 indication (in the protocol name) for TLS use. Now it follows the RFC and
2138 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2139 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2140 names appear in log lines as well as in Received: header lines.
2142 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2143 extensions, and comparison tests, and to fix some bugs.
2145 14. Changes to the "personal" filter test:
2147 (1) The test was buggy in that it was just doing the equivalent of
2148 "contains" tests on header lines. For example, if a user's address was
2149 anne@some.where, the "personal" test would incorrectly be true for
2151 To: susanne@some.where
2153 This test is now done by extracting each address from the header in turn,
2154 and checking the entire address. Other tests that are part of "personal"
2155 are now done using regular expressions (for example, to check local parts
2156 of addresses in From: header lines).
2158 (2) The list of non-personal local parts in From: addresses has been
2159 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2160 taken from the Sieve specification recommendations.
2162 (3) If the message contains any header line starting with "List-" it is
2163 treated as non-personal.
2165 (4) The test for "circular" in the Subject: header line has been removed
2166 because it now seems ill-conceived.
2168 15. Minor typos in src/EDITME comments corrected.
2170 16. Installed latest exipick from John Jetmore.
2172 17. If headers_add on a router specified a text string that was too long for
2173 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2174 of string_sprintf() is now avoided.
2176 18. $message_body_size was not set (it was always zero) when running the DATA
2177 ACL and the local_scan() function.
2179 19. For the "mail" command in an Exim filter, no default was being set for
2180 the once_repeat time, causing a random time value to be used if "once" was
2181 specified. (If the value happened to be <= 0, no repeat happened.) The
2182 default is now 0s, meaning "never repeat". The "vacation" command was OK
2183 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2184 (I found it when inspecting the code).
2186 20. There is now an overall timeout for performing a callout verification. It
2187 defaults to 4 times the callout timeout, which applies to individual SMTP
2188 commands during the callout. The overall timeout applies when there is more
2189 than one host that can be tried. The timeout is checked before trying the
2190 next host. This prevents very long delays if there are a large number of
2191 hosts and all are timing out (e.g. when the network connections are timing
2192 out). The value of the overall timeout can be changed by specifying an
2193 additional sub-option for "callout", called "maxwait". For example:
2195 verify = sender/callout=5s,maxwait=20s
2197 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2198 to the end before writing, but this should make it even safer).
2200 22. Exim was forgetting that it had advertised PIPELINING for the second and
2201 subsequent messages on an SMTP connection. It was also not resetting its
2202 memory on STARTTLS and an internal HELO.
2204 23. When Exim logs an SMTP synchronization error within a session, it now
2205 records whether PIPELINING has been advertised or not.
2207 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2208 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2209 rather than long int.
2211 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2213 26. Added the never_mail option to autoreply.
2219 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2220 crash if the getsockname() call failed; this can happen if a connection is
2221 closed very soon after it is established. The problem was simply in the
2222 order in which certain operations were done, causing Exim to try to write
2223 to the SMTP stream before it had set up the file descriptor. The bug has
2224 been fixed by making things happen in the correct order.
2230 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2231 before the connection was closed, thus losing the rejection response.
2233 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2234 some early Solaris releases, but causes trouble in current releases where
2235 socklen_t is defined.
2237 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2240 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2243 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2244 that releasing the top of it at the end releases what was used for sub-
2245 expansions (unless the block got too big). However, discard this block if
2246 the first thing is a variable or header, so that we can use its block when
2247 it is dynamic (useful for very large $message_headers, for example).
2249 6. Lookups now cache *every* query, not just the most recent. A new, separate
2250 store pool is used for this. It can be recovered when all lookup caches are
2251 flushed. Lookups now release memory at the end of their result strings.
2252 This has involved some general refactoring of the lookup sources.
2254 7. Some code has been added to the store_xxx() functions to reduce the amount
2255 of flapping under certain conditions.
2257 8. log_incoming_interface used to affect only the <= reception log lines. Now
2258 it causes the local interface and port to be added to several more SMTP log
2259 lines, for example "SMTP connection from", and rejection lines.
2261 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2263 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2265 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2266 could be overwritten at the end of the current message (or the start of a
2267 new message if it was set in a HELO ACL). The value is now preserved for
2268 the duration of the SMTP connection.
2270 12. If a transport had a headers_rewrite setting, and a matching header line
2271 contained an unqualified address, that address was qualified, even if it
2272 did not match any rewriting rules. The underlying bug was that the values
2273 of the flags that permit the existence of unqualified sender and recipient
2274 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2275 non-local messages, and by -bnq for local messages) were not being
2276 preserved with the message after it was received.
2278 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2279 "next input=" as part of the text comprising the host identity instead of
2280 the correct text. The code was using the same buffer for two different
2281 strings. However, depending on which order the printing function evaluated
2282 its arguments, the bug did not always show up. Under Linux, for example, my
2283 test suite worked just fine.
2285 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2286 doesn't work with some older versions of Perl. It has been changed to "my",
2287 which in any case is probably the better facility to use.
2289 15. A really picky compiler found some instances of statements for creating
2290 error messages that either had too many or two few arguments for the format
2293 16. The size of the buffer for calls to the DNS resolver has been increased
2294 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2295 for addresses that have a lot of PTR records. This alleviates a problem; it
2296 does not fully solve it.
2298 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2299 buffer now truncates the list and logs the incident, which is the same
2300 action as happens when Exim is looking up a host name and its aliases.
2301 Previously in this situation something unpredictable would happen;
2302 sometimes it was "internal error: store_reset failed".
2304 18. If a server dropped the connection unexpectedly when an Exim client was
2305 using GnuTLS and trying to read a response, the client delivery process
2306 crashed while trying to generate an error log message.
2308 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2309 string, for example:
2311 warn message = H1: something\nH2: something
2313 the text was added as a single header line from Exim's point of view
2314 though it ended up OK in the delivered message. However, searching for the
2315 second and subsequent header lines using $h_h2: did not work. This has been
2316 fixed. Similarly, if a system filter added multiple headers in this way,
2317 the routers could not see them.
2319 20. Expanded the error message when iplsearch is called with an invalid key to
2320 suggest using net-iplsearch in a host list.
2322 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2323 ACLs are no longer actually imposed (and a message to that effect is
2326 22. If a "gecos" field in a passwd entry contained escaped characters, in
2327 particular, if it contained a \" sequence, Exim got it wrong when building
2328 a From: or a Sender: header from that name. A second bug also caused
2329 incorrect handling when an unquoted " was present following a character
2330 that needed quoting.
2332 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2333 was not being matched caselessly.
2335 24. Arranged for all hyphens in the exim.8 source to be escaped with
2338 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2339 properly. Recipient callout cache records were still being keyed to include
2340 the sender, even when use_sender was set false. This led to far more
2341 callouts that were necessary. The sender is no longer included in the key
2342 when use_sender is false.
2344 26. Added "control = submission" modifier to ACLs.
2346 27. Added the ${base62d: operator to decode base 62 numbers.
2348 28. dnsdb lookups can now access SRV records.
2350 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2351 the configuration file.
2353 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2354 (-v) mode. This makes the output for a verbose queue run more intelligible.
2356 31. Added a use_postmaster feature to recipient callouts.
2358 32. Added the $body_zerocount variable, containing the number of binary zero
2359 bytes in the message body.
2361 33. The time of last modification of the "new" subdirectory is now used as the
2362 "mailbox time last read" when there is a quota error for a maildir
2365 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2367 35. Added +ignore_unknown as a special item in host lists.
2369 36. Code for decoding IPv6 addresses in host lists is now included, even if
2370 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2371 address was recognized as an IP address, but was then not correctly decoded
2372 into binary, causing unexpected and incorrect effects when compared with
2379 1. Very minor rewording of debugging text in manualroute to say "list of
2380 hosts" instead of "hostlist".
2382 2. If verify=header_syntax was set, and a header line with an unqualified
2383 address (no domain) and a large number of spaces between the end of the
2384 name and the colon was received, the reception process suffered a buffer
2385 overflow, and (when I tested it) crashed. This was caused by some obsolete
2386 code that should have been removed. The fix is to remove it!
2388 3. When running in the test harness, delay a bit after writing a bounce
2389 message to get a bit more predictability in the log output.
2391 4. Added a call to search_tidyup() just before forking a reception process. In
2392 theory, someone could use a lookup in the expansion of smtp_accept_max_
2393 per_host which, without the tidyup, could leave open a database connection.
2395 5. Added the variables $recipient_data and $sender_data which get set from a
2396 lookup success in an ACL "recipients" or "senders" condition, or a router
2397 "senders" option, similar to $domain_data and $local_part_data.
2399 6. Moved the writing of debug_print from before to after the "senders" test
2402 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2403 problems for message scanning, either using a data ACL, or using
2404 local_scan() because the Received: header was not generated till after they
2405 were called (in order to set the time as the time of reception completion).
2406 I have revised the way this works. The header is now generated after the
2407 body is received, but before the ACL or local_scan() are called. After they
2408 are run, the timestamp in the header is updated.
2414 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2415 before starting a queue runner without re-exec. This happened only when
2416 deliver_drop_privilege was set or when the Exim user was set to root. The
2417 effect of the bug was that timeouts during subsequent deliveries caused
2418 crashes instead of being properly handled. The handler is now left at its
2419 default (and expected) setting.
2421 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2422 message, again when deliver_drop_privilege is set or Exim is run as root.
2423 The bug described in (1) was not present in this case, but the tidying up
2424 of the other signals was missing. I have made the two cases consistent.
2426 3. The ignore_target_hosts setting on a manualroute router was being ignored
2427 for hosts that were looked up using the /MX notation.
2429 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2432 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2433 operated on the sender address. After changing the $sender_address to <>
2434 for the sender address verify, Exim was re-instated it as the original
2435 (before rewriting) address, but remembering that it had rewritten it, so it
2436 wasn't rewriting it again. This bug also had the effect of breaking the
2437 sender address verification caching when the sender address was rewritten.
2439 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2440 This has been changed so that if the ip literal address matches
2441 ignore_target_hosts, the router declines.
2443 7. Added expansion conditions match_domain, match_address, and match_local_
2444 part (NOT match_host).
2446 8. The placeholder for the Received: header didn't have a length field set.
2448 9. Added code to Exim itself and to exim_lock to test for a specific race
2449 condition that could lead to file corruption when using MBX delivery. The
2450 issue is with the lockfile that is created in /tmp. If this file is removed
2451 after a process has opened it but before that process has acquired a lock,
2452 there is the potential for a second process to recreate the file and also
2453 acquire a lock. This could lead to two Exim processes writing to the file
2454 at the same time. The added code performs the same test as UW imapd; it
2455 checks after acquiring the lock that its file descriptor still refers to
2456 the same named file.
2458 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2459 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2462 11. Added the smtp_active_hostname option. If used, this will typically be made
2463 to depend on the incoming interface address. Because $interface_address is
2464 not set up until the daemon has forked a reception process, error responses
2465 that can happen earlier (such as "too many connections") no longer contain
2468 12. If an expansion in a condition on a "warn" statement fails because a lookup
2469 defers, the "warn" statement is abandoned, and the next ACL statement is
2470 processed. Previously this caused the whole ACL to be aborted.
2472 13. Added the iplsearch lookup type.
2474 14. Added ident_timeout as a log selector.
2476 15. Added tls_certificate_verified as a log selector.
2478 16. Added a global option tls_require_ciphers (compare the smtp transport
2479 option of the same name). This controls incoming TLS connections.
2481 17. I finally figured out how to make tls_require_ciphers do a similar thing
2482 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2483 before starting the TLS session.
2485 18. Tabs are now shown as \t in -bP output.
2487 19. If the log selector return_path_on_delivery was set, Exim crashed when
2488 bouncing a message because it had too many Received: header lines.
2490 20. If two routers both had headers_remove settings, and the first one included
2491 a superfluous trailing colon, the final name in the first list and the
2492 first name in the second list were incorrectly joined into one item (with a
2493 colon in the middle).
2499 1. Added -C and -D options to the exinext utility, mainly to make it easier
2500 to include in the automated testing, but these could be helpful when
2501 multiple configurations are in use.
2503 2. The exinext utility was not formatting the output nicely when there was
2504 an alternate port involved in the retry record key, nor when there was a
2505 message id as well (for retries that were specific to a specific message
2506 and a specific host). It was also confused by IPv6 addresses, because of
2507 the additional colons they contain. I have fixed the IPv4 problem, and
2508 patched it up to do a reasonable job for IPv6.
2510 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2511 delivery, the log line now contains "pipelined" if PIPELINING was used.
2513 4. An SMTP transport process used to panic and die if the bind() call to set
2514 an explicit outgoing interface failed. This has been changed; it is now
2515 treated in the same way as a connect() failure.
2517 5. A reference to $sender_host_name in the part of a conditional expansion
2518 that was being skipped was still causing a DNS lookup. This no longer
2521 6. The def: expansion condition was not recognizing references to header lines
2522 that used bh_ and bheader_.
2524 7. Added the _cache feature to named lists.
2526 8. The code for checking quota_filecount in the appendfile transport was
2527 allowing one more file than it should have been.
2529 9. For compatibility with Sendmail, the command line option
2537 and sets the incoming protocol and host name (for trusted callers). The
2538 host name and its colon can be omitted when only the protocol is to be set.
2539 Note the Exim already has two private options, -pd and -ps, that refer to
2540 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2541 "s", but I don't think that's a major issue.
2543 10. A number of refactoring changes to the code, none of which should affect
2546 (a) The number of logging options was getting close to filling up the
2547 32-bit word that was used as a bit map. I have split them into two classes:
2548 those that are passed in the argument to log_write(), and those that are
2549 only ever tested independently outside of that function. These are now in
2550 separate 32-bit words, so there is plenty of room for expansion again.
2551 There is no change in the user interface or the logging behaviour.
2553 (b) When building, for example, log lines, the code previously used a
2554 macro that called string_cat() twice, in order to add two strings. This is
2555 not really sufficiently general. Furthermore, there was one instance where
2556 it was actually wrong because one of the argument was used twice, and in
2557 one call a function was used. (As it happened, calling the function twice
2558 did not affect the overall behaviour.) The macro has been replaced by a
2559 function that can join an arbitrary number of extra strings onto a growing
2562 (c) The code for expansion conditions now uses a table and a binary chop
2563 instead of a serial search (which was left over from when there were very
2564 few conditions). Also, it now recognizes conditions like "pam" even when
2565 the relevant support is not compiled in: a suitably worded error message is
2566 given if an attempt is made to use such a condition.
2568 11. Added ${time_interval:xxxxx}.
2570 12. A bug was causing one of the ddress fields not to be passed back correctly
2571 from remote delivery subprocesses. The field in question was not being
2572 subsequently used, so this caused to problems in practice.
2574 13. Added new log selectors queue_time and deliver_time.
2576 14. Might have fixed a bug in maildirsizefile handling that threw up
2577 "unexpected character" debug warnings, and recalculated the data
2578 unnecessarily. In any case, I expanded the warning message to give more
2581 15. Added the message "Restricted characters in address" to the statements in
2582 the default ACL that block characters like @ and % in local parts.
2584 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2585 Three changes have been made:
2587 (a) There was a serious bug; a negative response to MAIL caused the whole
2588 recipient domain to be cached as invalid, thereby blocking all messages
2589 to all local parts at the same domain, from all senders. This bug has
2590 been fixed. The domain is no longer cached after a negative response to
2591 MAIL if the sender used is not empty.
2593 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2596 (c) A new callout option, "use_sender" has been added for people who want
2597 the modified behaviour.
2603 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2606 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2607 indeed breaks things for older releases.
2609 3. Added additional logging to the case where there is a problem reading data
2610 from a filter that is running in a subprocess using a pipe, in order to
2611 try to track down a specific problem.
2613 4. Testing facility fudge: when running in the test harness and attempting
2614 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2615 getting "No route to host". Convert this to a timeout.
2617 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2620 6. Some OS don't have socklen_t but use size_t instead. This affects the
2621 fifth argument of getsockopt() amongst other things. This is now
2622 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2623 can be set for individual OS. I have set it for SunOS5, OSF1, and
2624 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2625 some earlier ones do not.
2627 7. Change 4.30/15 was not doing the test caselessly.
2629 8. The standard form for an IPv6 address literal was being rejected by address
2630 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2631 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2632 this, as well as the form without the "IPv6" on the front (but only when
2633 address literals are enabled, of course).
2635 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2637 10. Exim crashed if a message with an empty sender address specified by -f
2638 encountered a router with an errors_to setting. This could be provoked only
2639 by a command such as
2643 where an empty string was supplied; "<>" did not hit this bug.
2645 11. Installed PCRE release 4.5.
2647 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2648 remained set. It is now erased.
2650 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2651 times from message ids (which are base 36 rather than the normal 62).
2653 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2654 were being counted as actual protocol errors, and logged if the log
2655 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2656 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2657 and DATA following a set of rejected RCPTs do not count as protocol errors.
2658 In other words, Exim assumes they were pipelined, though this may not
2659 actually be the case. Of course, in all cases the client gets an
2660 appropriate error code.
2662 15. If a lookup fails in an ACL condition, a message about the failure may
2663 be available; it is used if testing the ACL cannot continue, because most
2664 such messages specify what the cause of the deferral is. However, some
2665 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2666 that caused an old message to be retained and used if a later statement
2667 caused a defer, replacing the real cause of the deferral.
2669 16. If an IP address had so many PTR records that the DNS lookup buffer
2670 was not large enough to hold them, Exim could crash while trying to process
2671 the truncated data. It now detects and logs this case.
2673 17. Further to 4.21/58, another change has been made: if (and only if) the
2674 first line of a message (the first header line) ends with CRLF, a bare LF
2675 in a subsequent header line has a space inserted after it, so as not to
2676 terminate the header.
2678 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2679 unnecessarily, used atoi() instead of strtol(), and didn't check the
2680 termination when getting file sizes from file names by regex.
2682 19. Completely re-implemented the support for maildirsize files, in the light
2683 of a number of problems with the previous contributed implementation
2684 (4.30/29). In particular:
2686 . If the quota is zero, the maildirsize file is maintained, but no quota is
2689 . If the maildir directory does not exist, it is created before any attempt
2690 to write a maildirsize file.
2692 . The quota value in the file is just a cache; if the quota is changed in
2693 the transport, the new value overrides.
2695 . A regular expression is available for excluding directories from the
2698 20. The autoreply transport checks the characters in options that define the
2699 message's headers; it allows continued headers, but it was checking with
2700 isspace() after an embedded newline instead of explicitly looking for a
2703 21. If all the "regular" hosts to which an address was routed had passed their
2704 expiry times, and had not reached their retry times, the address was
2705 bounced, even if fallback hosts were defined. Now Exim should go on to try
2708 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2709 equivalent code in the SMTP transport. Some hosts send humungous responses
2710 to HELO/EHLO, more than 1024 it seems.
2712 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2713 gives compiler warnings in some environments. I've now done it "properly",
2716 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2717 (because of problems with the built-in one) was declared to return uschar *
2718 instead of char *, causing compiler failure.
2720 25. Fixed a file descriptor leak when processing alias/forward files.
2722 26. Fixed a minor format string issue in dbfn.c.
2724 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2726 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2727 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2730 29. When a sender address is verified, it is cached, to save repeating the test
2731 when there is more than one recipient in a message. However, when the
2732 verification involves a callout, it is possible for different callout
2733 options to be set for different recipients. It is too complicated to keep
2734 track of this in the cache, so now Exim always runs a verification when a
2735 callout is required, relying on the callout cache for the optimization.
2736 The overhead is duplication of the address routing, but this should not be
2739 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2740 to be verified with callout=postmaster, and the main callout worked but the
2741 postmaster check failed, the verification correctly failed. However, if a
2742 subsequent RCPT command asked for sender verification *without* the
2743 postmaster check, incorrect caching caused this verification also to fail,
2746 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2747 it was not caching the DNS options (qualify_single, search_parents) that
2748 were used when the lookup failed. A subsequent lookup with different
2749 options therefore always gave the same answer, though there were cases
2750 where it should not have. (Example: a "domains = !$mx_any" option on a
2751 dnslookup router: the "domains" option is always processed without any
2752 widening, but the router might have qualify_single set.) Now Exim uses the
2753 cached value only when the same options are set.
2755 32. Added John Jetmore's "exipick" utility to the distribution.
2757 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2758 than a timeout (e.g. a certificate is required, and is not provided), an
2759 Exim server now closes the connection immediately. Previously it waited for
2760 the client to close - but if the client is SSL, it seems that they each
2761 wait for each other, leading to a delay before one of them times out.
2763 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2764 maintained 0.8.x compatibility because I don't think many are using it, and
2765 it is clearly obsolete.
2767 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2770 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2771 client certificate was expired. A simple patch fixes this, though I don't
2772 understand the full logic of why the verify callback is called multiple
2775 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2776 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2777 which causes problems with some clients (such as the Certicom SSL Plus
2778 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2779 disables the coutermeasure allowing Eudora to connect."
2781 38. Exim was not checking that a write() to a log file succeeded. This could
2782 lead to Bad Things if a log got too big, in particular if it hit a file
2783 size limit. Exim now panics and dies if it cannot write to a log file, just
2784 as it does if it cannot open a log file.
2786 39. Modified OS/Makefile-Linux so that it now contains
2788 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2790 The two -D definitions ensure that Exim is compiled with large file
2791 support, which makes it possible to handle log files that are bigger than
2794 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2795 instance) a domain was checked against a named list that involved a lookup,
2796 causing $domain_data to be set, then another domain was checked against the
2797 same list, then the first domain was re-checked, the value of $domain_data
2798 after the final check could be wrong. In particular, if the second check
2799 failed, it could be set empty. This bug probably also applied to
2802 41. The strip_trailing_dot option was not being applied to the address given
2803 with the -f command-line option.
2805 42. The code for reading a message's header from the spool was incrementing
2806 $received_count, but never initializing it. This meant that the value was
2807 incorrect (doubled) while delivering a message in the same process in which
2808 it was received. In the most common configuration of Exim, this never
2809 happens - a fresh exec is done - but it can happen when
2810 deliver_drop_privilege is set.
2812 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2813 it now includes up to 150 characters of the unexpected data in the log
2816 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2817 and building data strings. The size of both of these buffers was 10 000
2818 bytes - far larger than anybody would *ever* want, thought I. Needless to
2819 say, somebody hit the limit. I have increased the maximum line length to
2820 20 000 and the maximum data length of concatenated lines to 100 000. I have
2821 also fixed two bugs, because there was no checking on these buffers. Tsk,
2822 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2823 buffer is too small.
2825 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2826 lsearch lookups. Now it does.
2828 46. When parsing a route_list item in a manualroute router, a fixed-length
2829 buffer was used for the list of hosts. I made this 1024 bytes long,
2830 thinking that nobody would ever have a list of hosts that long. Wrong.
2831 Somebody had a whole pile of complicated expansion conditions, and the
2832 string was silently truncated, leading to an expansion error. It turns out
2833 that it is easier to change to an unlimited length (owing to other changes
2834 that have happened since this code was originally written) than to build
2835 structure for giving a limitation error. The length of the item that
2836 expands into the list of hosts is now unlimited.
2838 47. The lsearch lookup could not handle data where the length of text line was
2839 more than 4095 characters. Such lines were truncated, leading to shortened
2840 data being returned. It should now handle lines of any length.
2842 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2843 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2846 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2847 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2848 no longer happens in comment lines. A list of replacements is now placed
2849 at the head of all of the source files, except those whose only change is
2850 to replace PERL_COMMAND in the very first #! line.
2852 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2853 messages on the queue, with a bottom-up merge sort, using code contributed
2854 by Michael Haardt. This should make operations like -bp somewhat faster on
2855 large queues. It won't affect queue runners, except when queue_run_in_order
2858 51. Installed eximstats 1.31 in the distribution.
2860 52. Added support for SRV lookups to the dnslookup router.
2862 53. If an ACL referred to $message_body or $message_body_end, the value was not
2863 reset for any messages that followed in the same SMTP session.
2865 54. The store-handling optimization for building very long strings was not
2866 differentiating between the different store pools. I don't think this
2867 actually made any difference in practice, but I've tidied it.
2869 55. While running the routers to verify a sender address, $sender_address
2870 was still set to the sender address. This is wrong, because when routing to
2871 send a bounce to the sender, it would be empty. Therefore, I have changed
2872 it so that, while verifying a sender address, $sender_address is set to <>.
2873 (There is no change to what happens when verifying a recipient address.)
2875 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2876 target A or AAAA records (if not already returned) without resetting the
2877 qualify_single or search_parents options of the DNS resolver. These are
2878 inappropriate in this case because the targets of MX and SRV records must
2879 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2880 target that, when qualified, matched something in the local domain. These
2881 two options are now turned off when doing these lookups.
2883 57. It seems that at least some releases of Reiserfs (which does not have the
2884 concept of a fixed number of inodes) returns zero and not -1 for the
2885 number of available inodes. This interacted badly with check_spool_inodes,
2886 which assumed that -1 was the "no such thing" setting. What I have done is
2887 to check that the total number of inodes is greater than zero before doing
2888 the test of how many are available.
2890 58. When a "warn" ACL statement has a log_message modifier, the message is
2891 remembered, and not repeated. This is to avoid a lot of repetition when a
2892 message has many recipients that cause the same warning to be written.
2893 Howewer, Exim was preserving the list of already written lines for an
2894 entire SMTP session, which doesn't seem right. The memory is now reset if a
2895 new message is started.
2897 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2898 debugging output unless log_rewrite was also set.
2900 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2901 of (int)(handle) when we know that handle contains (void *)(-1).
2903 61. The Exim daemon panic-logs an error return when it closes the incoming
2904 connection. However "connection reset by peer" seems to be common, and
2905 isn't really an error worthy of noting specially, so that particular error
2908 62. When Exim is trying to find all the local interfaces, it used to panic and
2909 die if the ioctl to get the interface flags failed. However, it seems that
2910 on at least one OS (Solaris 9) it is possible to have an interface that is
2911 included in the list of interfaces, but for which you get a failure error
2912 for this call. This happens when the interface is not "plumbed" into a
2913 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2914 failure of the "get flags" call assumes that the interface is down.
2916 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2917 makes life easier for people who are doing arithmetic on fields extracted
2918 from dates, where you often get leading zeros that should not be
2919 interpreted as octal.
2921 64. Added qualify_domain to the redirect router, to override the global
2924 65. If a pathologically long header line contained very many addresses (the
2925 report of this problem mentioned 10 000) and each of them was rewritten,
2926 Exim could use up a very large amount of memory. (It kept on making new
2927 copies of the header line as it rewrote, and never released the old ones.)
2928 At the expense of a bit more processing, the header rewriting function has
2929 been changed so that it no longer eats memory in this way.
2931 66. The generation of the Received: header has been moved from the time that a
2932 message starts to be received, to the time that it finishes. The timestamp
2933 in the Received: header should now be very close to that of the <= log
2934 line. There are two side-effects of this change:
2936 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2937 logged header lines no longer include the local Received: line, because
2938 it has not yet been created. The same applies to a copy of the message
2939 that is returned to a non-SMTP sender when a message is rejected.
2941 (b) When a filter file is tested using -bf, no additional Received: header
2942 is added to the test message. After some thought, I decided that this
2945 This change does not affect the value of $received_for. It is still set
2946 after address rewriting, but before local_scan() is called.
2948 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2950 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2951 gave an unhelpful panic error message, and a defer error. I have managed to
2952 change this behaviour so that it now rejects any supplied certificate,
2953 which seems right, as the list of acceptable certificates is empty.
2955 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2956 gave an unhelpful defer error. I have not managed to make this reject any
2957 supplied certificates, but the error message it gives is "no certificate
2958 supplied", which is not helpful.
2960 70. exigrep's output now also includes lines that are not associated with any
2961 message, but which match the given pattern. Implemented by a patch from
2962 Martin Sluka, which also tidied up the Perl a bit.
2964 71. Recipient callout verification, like sender verification, was using <> in
2965 the MAIL FROM command. This isn't really the right thing, since the actual
2966 sender may affect whether the remote host accepts the recipient or not. I
2967 have changed it to use the actual sender in the callout; this means that
2968 the cache record is now keyed on a recipient/sender pair, not just the
2969 recipient address. There doesn't seem to be a real danger of callout loops,
2970 since a callout by the remote host to check the sender would use <>.
2971 [SEE ABOVE: changed after hitting problems.]
2973 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2974 temporary errors. However, in the case of such a code being given after
2975 the end of a data transmission (i.e. after ".") Exim was failing to write
2976 a retry record for the message. (Yes, there was some broken host that was
2977 actually sending 8xx at this point.)
2979 73. An unknown lookup type in a host list could cause Exim to panic-die when
2980 the list was checked. (An example that provoked this was putting <; in the
2981 middle of a list instead of at the start.) If this happened during a DATA
2982 ACL check, a -D file could be left lying around. This kind of configuration
2983 error no longer causes Exim to die; instead it causes a defer errror. The
2984 incident is still logged to the main and panic logs.
2986 74. Buglet left over from Exim 3 conversion. The message "too many messages
2987 in one connection" was written to the rejectlog but not the mainlog, except
2988 when address rewriting (yes!) was being logged.
2990 75. Added write_rejectlog option.
2992 76. When a system filter was run not as root (that is, when system_filter_user
2993 was set), the values of the $n variables were not being returned to the
2994 main process; thus, they were not subsequently available in the $sn
2997 77. Added +return_path_on_delivery log selector.
2999 78. A connection timeout was being treated differently from recipients deferred
3000 when testing hosts_max_try with a message that was older than the host's
3001 retry timeout. (The host should not be counted, thus allowing all hosts to
3002 be tried at least once before bouncing.) This may have been the cause of an
3003 occasionally reported bug whereby a message would remain on the queue
3004 longer than the retry timeout, but would be bounced if a delivery was
3005 forced. I say "may" because I never totally pinned down the problem;
3006 setting up timeout/retry tests is difficult. See also the next item.
3008 79. The ultimate address timeout was not being applied to errors that involved
3009 a combination of host plus message (for example, a timeout on a MAIL
3010 command). When an address resolved to a number of possible hosts, and they
3011 were not all tried for each delivery (e.g. because of hosts_max_try), a
3012 message could remain on the queue longer than the retry timeout.
3014 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3017 81. Fixed an obscure SMTP outgoing bug which required at least the following
3018 conditions: (a) there was another message waiting for the same server;
3019 (b) the server returned 5xx to all RCPT commands in the first message so
3020 that the message was not completed; (c) the server dropped the connection
3021 or gave a negative response to the RSET that Exim sends to abort the
3022 transaction. The observed case was a dropped connection after DATA that had
3023 been sent in pipelining mode. That is, the server had advertised PIPELINING
3024 but was not implementing it correctly. The effect of the bug was incorrect
3025 behaviour, such as trying another host, and this could lead to a crash.
3031 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3032 and daemon.c were passed as pointers to ints; they should have been
3033 pointers to socklen_t variables (which are typically unsigned ints).
3035 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3038 3. Fixed a really odd bug that affected only the testing scheme; patching a
3039 certain fixed string in the binary changed the value of another string that
3040 happened to be identical to the end of the original first string.
3042 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3043 name", it returns that address as the IP address. On some operating
3044 systems (e.g. Solaris), it also passes back the IP address string as the
3045 "host name". However, on others (e.g. Linux), it passes back an empty
3046 string. Exim wasn't checking for this, and was changing the host name to an
3047 empty string, assuming it had been canonicized.
3049 5. Although rare, it is permitted to have more than one PTR record for a given
3050 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3051 all the names associated with an address, because they do in Solaris.
3052 However, it seems that they do not in Linux for data that comes from the
3053 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3054 I found this out when I moved to a new Linux workstation and tried to run
3055 the Exim test suite.
3057 To get round this problem I have changed the code so that it now does its
3058 own call to the DNS to look up PTR records when searching for a host name.
3059 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3060 addresses that are only in /etc/hosts are still found.
3062 This behaviour is, however, controlled by an option called host_lookup_
3063 order, which defaults to "bydns:byaddr". If people want to use the other
3064 order, or indeed, just use one or the other means of lookup, they can
3065 specify it in this variable.
3067 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3068 some operating systems, this comes back from gethostbyaddr() as an empty
3069 string, and this is what Exim used to test for. However, it seems that in
3070 other systems, "." is yielded. Exim now tests for this case too.
3072 7. The values of check_spool_space and check_log_space are now held internally
3073 as a number of kilobytes instead of an absolute number of bytes. If a
3074 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3075 kilobyte. This means that much larger values can be stored.
3077 8. Exim monitor: an attempt to get the action menu when not actually pointing
3078 at a message produces an empty menu entitled "No message selected". This
3079 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3080 no entries in it ("Shell widget menu has zero width and/or height"). So I
3081 have added a single, blank menu entry in this case.
3083 9. Added ${quote_local_part.
3085 10. MIME decoding is now applied to the contents of Subject: header lines when
3088 11. Now that a reference to $sender_host_address automatically causes a reverse
3089 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3090 host lookup before query-style lookups in lists that might use this
3091 variable. This has therefore been abolished, and the "net-" prefix is no
3092 longer necessary for query-style lookups.
3094 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3095 have been a typo for LFLAGS, so it has been changed.
3097 13. The install script calls Exim with "-C /dev/null" in order to find the
3098 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3099 to be output. Howeve, since Exim outputs its version number before the
3100 error, it didn't break the script. It just looked ugly. I fixed this by
3101 always allowing "-C /dev/null" if the caller is root.
3103 14. Ignore overlarge ACL variable number when reading spool file - insurance
3104 against a later release with more variables having written the file.
3106 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3107 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3108 this, as well as the form without the "IPv6" on the front.
3110 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3111 OS/Makefile-Darwin file.
3113 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3114 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3116 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3117 variable. However, some OS use ints for inodes. I've added cast to long int
3118 to get rid of the compiler warning.
3120 19. I had forgotten to lock out "/../" in configuration file names when
3121 ALT_CONFIG_PREFIX was set.
3123 20. Routers used for verification do not need to specify transports. However,
3124 if such a router generated a host list, and callout was configured, Exim
3125 crashed, because it could not find a port number from the (non-existent)
3126 transport. It now assumes port 25 in this circumstance.
3128 21. Added the -t option to exigrep.
3130 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3131 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3132 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3133 them is set, LOOKUP_LSEARCH is forced.
3135 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3137 24. Added sender and host information to the "rejected by local_scan()" log
3138 line; previously there was no indication of these.
3140 25. Added .include_if_exists.
3142 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3143 Linux. It turns out that not all file systems support this. Apparently some
3144 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3145 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3146 not supported on the file descriptor, is now ignored when Exim is trying to
3147 sync a directory. This applies only to Linux.
3149 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3151 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3152 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3155 29. Added support for maildirsize files from supplied patch (modified a bit).
3157 30. The use of :fail: followed by an empty string could lead Exim to respond to
3158 sender verification failures with (e.g.):
3160 550 Verification failed for <xxx>
3161 550 Sender verify failed
3163 where the first response line was missing the '-' that indicates it is not
3164 the final line of the response.
3166 31. The loop for finding the name of the user that called Exim had a hardwired
3167 limit of 10; it now uses the value of finduser_retries, which is used for
3168 all other user lookups.
3170 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3173 33. Exim was neglecting to zero errno before one call of strtol() when
3174 expanding a string and expecting an integer value. On some systems this
3175 resulted in spurious "integer overflow" errors. Also, it was casting the
3176 result into an int without checking.
3178 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3179 did not work. The code looks as if it has *never* worked, though it appears
3180 to have been documented since at least releast 1.62. I have made it work.
3182 35. The "timeout_DNS" error in retry rules, also documented since at least
3183 1.62, also never worked. As it isn't clear exactly what this means, and
3184 clearly it isn't a major issue, I have abolished the feature by treating it
3185 as "timeout", and writing a warning to the main and panic logs.
3187 36. The display of retry rules for -brt wasn't always showing the error code
3190 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3191 timeout_connect_A, timeout_connect_MX.
3193 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3194 to the empty sender.
3196 39. The daemon was not analysing the content of -oX till after it had closed
3197 stderr and disconnected from the controlling terminal. This meant that any
3198 syntax errors were only noted on the panic log, and the return code from
3199 the command was 0. By re-arranging the code a little, I've made the
3200 decoding happen first, so such errors now appear on stderr, and the return
3201 code is 1. However, the actual setting up of the sockets still happens in
3202 the disconnected process, so errors there are still only recorded on the
3205 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3206 connections (as happens on some IP stacks) was logged at start up time as
3207 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3208 it from "IPv6 and IPv4", which means that two separate sockets are being
3211 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3212 says whether AF_INET or AF_INET6 was passed as an argument.
3214 42. Exiwhat output was messed up when time zones were included in log
3217 43. Exiwhat now gives more information about the daemon's listening ports,
3218 and whether -tls-on-connect was used.
3220 44. The "port" option of the smtp transport is now expanded.
3222 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3223 silently ignored. Now an error message is written to the main and panic
3226 46. There's a new ACL modifier called "logwrite" which writes to a log file
3227 as soon as it is encountered.
3229 47. Added $local_user_uid and $local_user_gid at routing time.
3231 48. Exim crashed when trying to verify a sender address that was being
3234 49. Exim was recognizing only a space character after ".include". It now also
3235 recognizes a tab character.
3237 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3238 extracting the relevant information from the specification. The man page no
3239 longer contains scrambled data for the -d option, and I've added a section
3240 at the front about calling Exim under different names.
3242 51. Added "extra_headers" argument to the "mail" command in filter files.
3244 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3247 53. Installed eximstats 1.29.
3249 54. Added transport_filter_timeout as a generic transport option.
3251 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3252 Cc: header lines. This was required by RFC 822, but it not required by RFC
3255 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3256 incoming messages that did not have them. Now it does so only if the
3257 message originates locally, that is, if there is no associated remote host
3258 address. When Resent- header lines are present, this applies to the Resent-
3259 lines rather than the non-Resent- lines.
3261 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3262 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3264 58. Messages for configuration errors now include the name of the main
3265 configuration file - useful now that there may be more than one file in a
3266 list (.included file names were always shown).
3268 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3269 for those rare installations that do not start the daemon as root or run it
3270 setuid root. I've cut out the call to initgroups() if the daemon is not
3273 60. The Exim user and group can now be bound into the binary as text strings
3274 that are looked up at the start of Exim's processing.
3276 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3278 62. Added $mailstore_basename variable.
3280 63. Installed patch to sieve.c from Michael Haardt.
3282 64. When Exim failed to open the panic log after failing to open the main log,
3283 the original message it was trying to log was written to stderr and debug
3284 output, but if they were not available (the usual case in production), it
3285 was lost. Now it is written to syslog before the two lines that record the
3286 failures to open the logs.
3288 65. Users' Exim filters run in subprocesses under the user's uid. It is
3289 possible for a "deliver" command or an alias in a "personal" command to
3290 provoke an address rewrite. If logging of address rewriting is configured,
3291 this fails because the process is not running as root or exim. There may be
3292 a better way of dealing with this, but for the moment (because 4.30 needs
3293 to be released), I have disabled address rewrite logging when running a
3294 filter in a non-root, non-exim process.
3300 1. The buildconfig auxiliary program wasn't quoting the value set for
3301 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3302 not defined. This bug was masked in 4.22 by the effect that was fixed in
3305 2. Some messages that were rejected after a message id was allocated were
3306 shown as "incomplete" by exigrep. It no longer does this for messages that
3307 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3309 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3310 have allow_domain_literals set, the ID did not get logged in the <= line.
3311 Domain literals are now always recognized in Message-ID: header lines.
3313 4. The first argument for a ${extract expansion item is the key name or field
3314 number. Leading and trailing spaces in this item were not being ignored,
3315 causing some misleading effects.
3317 5. When deliver_drop_privilege was set, single queue runner processes started
3318 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3319 same command in the process it spins off) were not dropping privilege.
3321 6. When the daemon running as "exim" started a queue runner, it always
3322 re-executed Exim in the spun-off process. This is a waste of effort when
3323 deliver_drop_privilege is set. The new process now just calls the
3324 queue-runner function directly.
3330 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3333 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3334 ignored. Though the use of -oP was forcing the writing of a pid file, it
3335 was always written to the default place.
3337 3. If the message "no IP address found for host xxxx" is generated during
3338 incoming verification, it is now followed by identification of the incoming
3339 connection (so you can more easily find what provoked it).
3341 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3343 5. Added some features to "harden" Exim a bit more against certain attacks:
3345 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3346 be put in Local/Makefile. This is like the never_users runtime option,
3347 but it cannot be overridden. The default setting is "root".
3349 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3350 prefix string with which any file named in a -C command line option
3353 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3354 is retained for -C and -D only if the caller of Exim is root. Without
3355 it, the exim user may also use -C and -D and retain privilege.
3357 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3358 command line option is disabled.
3360 6. Macro names set by the -D option must start with an upper case letter, just
3361 like macro names defined in the configuration file.
3363 7. Added "dereference=" facility to LDAP.
3365 8. Two instances of the typo "uknown" in the source files are fixed.
3367 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3368 the Configure-Makefile script screwed up while processing it.
3370 10. Incorporated PCRE 4.4.
3372 11. The SMTP synchronization check was not operating right at the start of an
3373 SMTP session. For example, it could not catch a HELO sent before the client
3374 waited for the greeting. There is now a check for outstanding input at the
3375 point when the greeting is written. Because of the duplex, asynchronous
3376 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3377 way, but not yet received, when the check is performed.
3379 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3380 on TCP/IP sockets, because this apparently causes some broken clients to
3383 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3384 unchanged) from the Cygwin maintainer.
3386 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3387 was supported instead of testing for mbx. Effectively a typo.
3389 15. The spa authenticator server code was not checking that the input it
3390 received was valid base64.
3392 16. The debug output line for the "set" modifier in ACLs was not showing the
3393 name of the variable that was being set.
3395 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3397 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3398 lookup on its own. Something else was needed to trigger the lookup. For
3399 example, a match in host_lookup or the need for a host name in a host list.
3400 Now, if $sender_host_name is referenced and the host name has not yet been
3401 looked up, a lookup is performed. If the lookup fails, the variable remains
3402 empty, and $host_lookup_failed is set to "1".
3404 19. Added "eqi" as a case-independent comparison operator.
3406 20. The saslauthd authentication condition could segfault if neither service
3407 nor realm was specified.
3409 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3410 error message that was logged was misleading, and incoming SMTP
3411 connections were dropped. The message is now more accurate, and temporary
3412 errors are given to SMTP connections.
3414 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3415 (except RSET) with a 421 error, until QUIT is received. However, it was
3416 failing to send a response to QUIT.
3418 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3419 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3420 ACL is now run after the helo_try_verify_hosts code.
3422 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3423 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3424 case-independent, so other case variants are also recognized). Apparently
3425 some systems use these upper case variants.
3427 25. If more than two messages were waiting for the same host, and a transport
3428 filter was specified for the transport, Exim sent two messages over the
3429 same TCP/IP connection, and then failed with "socket operation on non-
3430 socket" when it tried to send the third.
3432 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3434 27. The extern definition of crypt16() in expand.c was not being excluded when
3435 the OS had its own crypt16() function.
3437 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3438 as a preferred synonym for return_size_limit, both as an option and as an
3441 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3443 30. Changed the default configuration ACL to relax the local part checking rule
3444 for addresses that are not in any local domains. For these addresses,
3445 slashes and pipe symbols are allowed within local parts, but the sequence
3446 /../ is explicitly forbidden.
3448 31. SPA server authentication was not clearing the challenge buffer before
3451 32. log_message in a "warn" ACL statement was writing to the reject log as
3452 well as to the main log, which contradicts the documentation and doesn't
3453 seem right (because no rejection is happening). So I have stopped it.
3455 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3456 However, I am unable to do any testing of this.
3458 34. Fixed an infelicity in the appendfile transport. When checking directories
3459 for a mailbox, to see if any needed to be created, it was accidentally
3460 using path names with one or more superfluous leading slashes; tracing
3461 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3463 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3464 contents are added to the log line that is written for every discarded
3465 recipient. (Previously a log_message setting was ignored.)
3467 36. The ${quote: operator now quotes the string if it is empty.
3469 37. The install script runs exim in order to find its version number. If for
3470 some reason other than non-existence or emptiness, which it checks, it
3471 could not run './exim', it was installing it with an empty version number,
3472 i.e. as "exim-". This error state is now caught, and the installation is
3475 38. An argument was missing from the function that creates an error message
3476 when Exim fails to connect to the socket for saslauthd authentication.
3477 This could cause Exim to crash, or give a corrupted message.
3479 39. Added isip, isip4, and isip6 to ${if conditions.
3481 40. The ACL variables $acl_xx are now saved with the message, and can be
3482 accessed later in routers, transports, and filters.
3484 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3485 strings in the file are not string-expanded.
3487 42. If a MAIL command specified a SIZE value that was too large to fit into an
3488 int variable, the check against message_size_limit failed. Such values are
3489 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3490 day this will have to be increased, but I don't think I want to be around
3491 when emails are that large.
3498 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3499 iconv() is not standard in FreeBSD.
3501 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3502 IPv6 enabled. The observed symptom was a segmentation fault on return from
3503 the function os_common_find_running_interfaces() in src/os.c.
3505 3. In the check_special_case() function in daemon.c I had used "errno" as an
3506 argument name, which causes warnings on some systems. This was basically a
3507 typo, since it was named "eno" in the comments!
3509 4. The code that waits for the clock to tick (at a resolution of some fraction
3510 of a second) so as to ensure message-id uniqueness was always waiting for
3511 at least one whole tick, when it could have waited for less. [This is
3512 almost certainly not relevant at current processor speeds, where it is
3513 unlikely to ever wait at all. But we try to future-proof.]
3515 5. The function that sleeps for a time interval that includes fractions of a
3516 second contained a race. It did not block SIGALRM between setting the
3517 timer, and suspending (a couple of lines later). If the interval was short
3518 and the sigsuspend() was delayed until after it had expired, the suspension
3519 never ended. On busy systems this could lead to processes getting stuck for
3522 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3523 process, before it forks any delivery processes. The open lookup caching
3524 mechanism meant that the open file or database connection was passed into
3525 the delivery process. The problem was that delivery processes always tidy
3526 up cached lookup data. This could cause a problem for the next delivery
3527 process started by the queue runner, because the external queue runner
3528 process does not know about the closure. So the next delivery process
3529 still has data in the lookup cache. In the case of a file lookup, there was
3530 no problem because closing a file descriptor in a subprocess doesn't affect
3531 the parent. However, if the lookup was caching a connection to a database,
3532 the connection was closed, and the second delivery process was likely to
3533 see errors such as "PGSQL: query failed: server closed the connection
3534 unexpectedly". The problem has been fixed by closing all cached lookups
3535 in a queue runner before running a delivery process.
3537 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3538 seem to have the "const" qualifier which it has on other OS. I've
3541 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3542 *of the same type* (client or server) with the same public name that an
3543 error should be diagnosed.
3545 9. When Exim looked up a host name for an IP address, but failed to find the
3546 original IP address when looking up the host name (a safety check), it
3547 output the message "<ip address> does not match any IP for NULL", which was
3548 confusing, to say the least. The bug was that the host name should have
3549 appeared instead of "NULL".
3551 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3552 user that is built into the binary, and the -C or -D options is used, root
3553 privilege is dropped before the configuration file is read. In addition,
3554 logging is switched to stderr instead of the normal log files. If the
3555 configuration then re-defines the Exim user, the unprivileged environment
3556 is probably not what is expected, so Exim logs a panic warning message (but
3559 However, if deliver_drop_privilege is set, the unprivileged state may well
3560 be exactly what is intended, so the warning has been cut out in that case,
3561 and Exim is allowed to try to write to its normal log files.
3567 1. smtp_return_error_details was not giving details for temporary sender
3568 or receiver verification errors.
3570 2. Diagnose a configuration error if two authenticators have the same public
3573 3. Exim used not to create the message log file for a message until the first
3574 delivery attempt. This could be confusing when incoming messages were held
3575 for policy or load reasons. The message log file is now created at the time
3576 the message is received, and an initial "Received" line is written to it.
3578 4. The automatically generated man page for command line options had a minor
3579 bug that caused no ill effects; however, a more serious problem was that
3580 the procedure for building the man page automatically didn't always
3581 operate. Consequently, release 4.20 contains an out-of-date version. This
3582 shouldn't happen again.
3584 5. When building Exim with embedded Perl support, the script that builds the
3585 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3586 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3588 6. The freeze_tell option was not being used for messages that were frozen on
3589 arrival, either by an ACL or by local_scan().
3591 7. Added the smtp_incomplete_transaction log selector.
3593 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3594 was accepting AUTH without a new EHLO.
3596 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3597 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3600 10. Logging of TCP/IP connections (when configured) now happens in the main
3601 daemon process instead of the child process, so that the TCP/IP connection
3602 count is more accurate (but it can never be perfect).
3604 11. The use of "drop" in a nested ACL was not being handled correctly in the
3605 outer ACL. Now, if condition failure induced by the nested "drop" causes
3606 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3607 or "require"), the connection is dropped.
3609 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3610 that yield "discard" can now be used with an "accept" or a "discard" verb,
3611 but an error is generated for any others (because I can't see a useful way
3612 to define what should happen).
3614 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3615 are now processed in the same way as lines in the Exim configuration file.
3616 In particular, continuation lines are supported.
3618 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3620 15. Added -ti meaning -t -i.
3622 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3623 domains, and warn by logging if others are found.
3625 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3626 returned from GIFCONF when Exim is trying to find the list of interfaces on
3627 a host. The code in os.c has been modified to copy each ifreq to an aligned
3628 structure in all cases.
3630 Also, in some cases, the returned ifreq's were being copied to a 'struct
3631 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3632 means the last couple of bytes of an IPv6 address could be chopped if the
3633 ifreq contained only a normal sockaddr (14 bytes storage).
3635 18. Named domain lists were not supported in the hosts_treat_as_local option.
3636 An entry such as +xxxx was not recognized, and was treated as a literal
3639 19. Ensure that header lines added by a DATA ACL are included in the reject log
3640 if the ACL subsequently rejects the message.
3642 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3643 MD5 (which is deprecated).
3645 21. When testing a filter file using -bf, Exim was writing a message when it
3646 took the sender from a "From " line in the message, but it was not doing so
3647 when it took $return_path from a Return-Path: header line. It now does.
3649 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3650 with a valid header line field name (a series of printing characters
3651 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3653 23. Changed "disc" in the source to "disk" to conform to the documentation and
3654 the book and for uniformity.
3656 24. Ignore Sendmail's -Ooption=value command line item.
3658 25. When execve() failed while trying to run a command in a pipe transport,
3659 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3660 could be confused with a return value of 69 from the command itself. This
3661 has been changed to 127, the value the shell returns if it is asked to run
3662 a non-existent command. The wording for the related log line suggests a
3663 non-existent command as the problem.
3665 26. If received_header_text expands to an empty string, do not add a Received:
3666 header line to the message. (Well, it adds a token one on the spool, but
3667 marks it "old" so that it doesn't get used or transmitted.)
3669 27. Installed eximstats 1.28 (addition of -nt option).
3671 28. There was no check for failure on the call to getsockname() in the daemon
3672 code. This can fail if there is a shortage of resources on the system, with
3673 ENOMEM, for example. A temporary error is now given on failure.
3675 29. Contrary to the C standard, it seems that in some environments, the
3676 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3677 program. Exim now does this explicitly; it affects the formatting of
3678 timestamps using strftime().
3680 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3681 complaints. I've now initialized all the variables, to avoid this.
3683 32. Header lines added by a system filter were not being "seen" during
3684 transport-time rewrites.
3686 33. The info_callback() function passed to OpenSSL is set up with type void
3687 (*)(SSL *, int, int), as described somewhere. However, when calling the
3688 function (actually a macro) that sets it up, the type void(*)() is
3689 expected. I've put in a cast to prevent warnings from picky compilers.
3691 34. If a DNS black list lookup found a CNAME record, but there were no A
3692 records associated with the domain it pointed at, Exim crashed.
3694 35. If a DNS black list lookup returned more than one A record, Exim ignored
3695 all but the first. It now scans all returned addresses if a particular IP
3696 value is being sought. In this situation, the contents of the
3697 $dnslist_value variable are a list of all the addresses, separated by a
3700 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3701 to accept a host name and all its aliases if the forward lookup for any of
3702 them yielded the IP address of the incoming connection. Now it accepts only
3703 those names whose forward lookup yields the correct IP address. Any other
3704 names are discarded. This closes a loophole whereby a rogue DNS
3705 administrator could create reverse DNS records to break through a
3706 wildcarded host restriction in an ACL.
3708 37. If a user filter or a system filter that ran in a subprocess used any of
3709 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3710 the wrong values were passed to the pipe command ($thisaddress had the
3711 value of $0, $0 had the value of $1, etc). This bug was introduced by
3712 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3714 38. Improved the line breaking for long SMTP error messages from ACLs.
3715 Previously, if there was no break point between 40 and 75 characters, Exim
3716 left the rest of the message alone. Two changes have been made: (a) I've
3717 reduced the minimum length to 35 characters; (b) if it can't find a break
3718 point between 35 and 75 characters, it looks ahead and uses the first one
3719 that it finds. This may give the occasional overlong line, but at least the
3720 remaining text gets split now.
3722 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3723 file descriptors might be low, and that setting 1000 would always raise it.
3724 It turns out that in some environments, the limit is already over 1000 and
3725 that lowering it causes trouble. So now Exim takes care not to decrease it.
3727 40. When delivering a message, the value of $return_path is set to $sender_
3728 address at the start of routing (routers may change the value). By an
3729 oversight, this default was not being set up when an address was tested by
3730 -bt or -bv, which affected the outcome if any router or filter referred to
3733 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3734 log only when "message" or "log_message" are set. However, if one of the
3735 conditions was an address verification, or a call to a nested ACL, the
3736 messages generated by the underlying test were being passed through. This
3737 no longer happens. The underlying message is available in $acl_verify_
3738 message for both "message" and "log_message" expansions, so it can be
3739 passed through if needed.
3741 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3742 new expansion $bh_ to give the encoded byte string without charset
3743 translation. Translation happens only if iconv() is available; HAVE_ICONV
3744 indicates this at build time. HEADERS_CHARSET gives the charset to
3745 translate to; headers_charset can change it in the configuration, and
3746 "headers charset" can change it in an individual filter file.
3748 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3749 that creates RFC 2047 encoded "words" labels them as that charset instead
3750 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3751 expansion operator; (ii) when Exim creates a From: line for a local
3752 message; (iii) when a header line is rewritten to include a "phrase" part.
3754 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3755 buggy, causing it to skip the first lines of messages whose message ID
3756 ended in 'D'. This would not have bitten before Exim release 4.14, because
3757 message IDs were unlikely to end in 'D' before then. The effect was to have
3758 incorrect size information for certain domains.
3760 45. #include "config.h" was missing at the start of the crypt16.c module. This
3761 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3764 46. If there was a timeout during a "random" callout check, Exim treated it as
3765 a failure of the random address, and carried on sending RSET and the real
3766 address. If the delay was just some slowness somewhere, the response to the
3767 original RCPT would be taken as a response to RSET and so on, causing
3768 mayhem of various kinds.
3770 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3771 when I implemented it. It didn't allow for the fact that some option values
3772 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3773 the right test for positive values.
3775 48. Domain names in DNS records are case-independent. Exim always looks them up
3776 in lower case. Some resolvers return domain names in exactly the case they
3777 appear in the zone file, that is, they may contain uppercase letters. Not
3778 all resolvers do this - some return always lower case. Exim was treating a
3779 change of case by a resolver as a change of domain, similar to a widening
3780 of a domain abbreviation. This triggered its re-routing code and so it was
3781 trying to route what was effectively the same domain again. This normally
3782 caused routing to fail (because the router wouldn't handle the domain
3783 twice). Now Exim checks for this case specially, and just changes the
3784 casing of the domain that it ultimately uses when it transmits the message
3787 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3790 50. If a filter generated a file delivery with a non-absolute name (possible if
3791 no home directory exists for the router), the forbid_file option was not
3794 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3795 the existing equality matching.
3797 52. Exim was using ints instead of ino_t variables in some places where it was
3798 dealing with inode numbers.
3800 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3801 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3802 called TMPDIR, and if it finds it is different, it changes its value.
3804 54. The smtp_printf() function is now made available to local_scan() so
3805 additional output lines can be written before returning. There is also an
3806 smtp_fflush() function to enable the detection of a dropped connection.
3807 The variables smtp_input and smtp_batched_input are exported to
3810 55. Changed the default runtime configuration: the message "Unknown user"
3811 has been removed from the ACL, and instead placed on the localuser router,
3812 using the cannot_route_message feature. This means that any verification
3813 failures that generate their own messages won't get overridden. Similarly,
3814 the "Unrouteable address" message that was in the ACL for unverifiable
3815 relay addresses has also been removed.
3817 56. Added hosts_avoid_esmtp to the smtp transport.
3819 57. The exicyclog script was not checking for the esoteric option
3820 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3821 will work only if exicyclog is run under the appropriate euid.
3823 58. Following a discussion on the list, the rules by which Exim recognises line
3824 endings on incoming messages have been changed. The -dropcr and drop_cr
3825 options are now no-ops, retained only for backwards compatibility. The
3826 following line terminators are recognized: LF CRLF CR. However, special
3827 processing applies to CR:
3829 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3830 nor a local message in the state where . is a terminator.
3832 (ii) If a bare CR is encountered in a header line, an extra space is added
3833 after the line terminator so as not to end the header. The reasoning
3834 behind this is that bare CRs in header lines are most likely either
3835 to be mistakes, or people trying to play silly games.
3837 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3838 was being incorrectly given as 18 bytes larger than it should have been.
3839 This is a VOB (very old bug).
3841 60. This may never have affected anything current, but just in case it has:
3842 When the local host is found other than at the start of a list of hosts,
3843 the local host, those with the same MX, and any that follow, are discarded.
3844 When the list in question was part of a longer list of hosts, the following
3845 hosts (not currently being processed) were also being discarded. This no
3846 longer happens. I'm not sure if this situation could ever has previously
3849 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3852 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3853 header. This is something that is recommended in a new Internet Draft, and
3854 is something that is documented as being done by Sendmail. There are two
3855 possible values. For messages generated by the autoreply transport, Exim
3858 Auto-Submitted: auto-replied
3860 whereas for all other generated messages (e.g. bounces) it adds
3862 Auto-Submitted: auto-generated
3864 63. The "personal" condition in filters now includes a test for the
3865 Auto-Submitted: header. If it contains the string "auto-" the message it
3866 not considered personal.
3868 64. Added rcpt_include_affixes as a generic transport option.
3870 65. Added queue_only_override (default true).
3872 66. Added the syslog_duplication option.
3874 67. If what should have been the first header line of a message consisted of
3875 a space followed by a colon, Exim was mis-interpreting it as a header line.
3876 It isn't of course - it is syntactically invalid and should therefore be
3877 treated as the start of the message body. The misbehaviour could have
3878 caused a number of strange effects, including loss of data in subsequent
3879 header lines, and spool format errors.
3881 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3882 client host had authenticated. This control can now be exercised by an ACL
3883 for more flexibility.
3885 69. By default, callouts do not happen when testing with -bh. There is now a
3886 variant, -bhc, which does actually run the callout code, including
3887 consulting and updating the callout cache.
3889 70. Added support for saslauthd authentication, courtesy of Alexander
3892 71. If statvfs() failed on the spool or log directories while checking their
3893 size for availability, Exim confusingly gave the error "space shortage".
3894 Furthermore, in debugging mode it crashed with a floating point exception.
3895 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3896 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3897 serious problem, Exim now writes to the main and panic logs when this
3898 happens, with details of the failure. It then refuses to accept the
3899 incoming message, giving the message "spool directory problem" or "log
3900 directory problem" with a 421 code for SMTP messages.
3902 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3903 0, 1, and 2 exist, because some OS complain for execs without them (see
3904 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3905 descriptors. However, the code omitted to check that the open succeeded,
3906 causing mysterious errors if for some reason the permissions on /dev/null
3907 got screwed. Now Exim writes a message to the main and panic logs, and
3908 bombs out if it can't open /dev/null.
3910 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3911 interact so that it is all more flexible. It is supposed to remain
3912 backwards compatible. Also added extra_local_interfaces.
3914 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3915 to bomb out with an assertion failure - to the client this appears as a
3916 connection drop. This problem occurs in the part of the code that was taken
3917 from the Samba project. Fortunately, the assertion is in a very simple
3918 function, so I have fixed this by reproducing the function inline in the
3919 one place where it is called, and arranging for authentication to fail
3920 instead of killing the process with assert().
3922 75. The SPA client code was not working when the server requested OEM rather
3923 than Unicode encoding.
3925 76. Added code to make require_files with a specific uid setting more usable in
3926 the case where statting the file as root fails - usually a non-root-mounted
3927 NFS file system. When this happens and the failure is EACCES, Exim now
3928 forks a subprocess and does the per-uid checking as the relevant uid.
3930 77. Added process_log_path.
3932 78. If log_file_path was not explicitly set, a setting of check_log_space or
3933 check_log_inodes was ignored.
3935 79. If a space check for the spool or log partitions fails, the incident is now
3936 logged. Of course, in the latter case the data may get lost...
3938 80. Added the %p formatting code to string_format() so that it can be used to
3939 print addresses in debug_print(). Adjusted all the address printing in the
3940 debugging in store.c to use %p rather than %d.
3942 81. There was a concern that a line of code in smtp_in.c could overflow a
3943 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3944 initially expressed, the concern was not well-founded, because trailing
3945 spaces are removed early. However, if the trailing spaces were followed by
3946 a NULL, they did not get removed, so the overflow was possible. Two fixes
3949 (a) I re-wrote the offending code in a cleaner fashion.
3950 (b) If an incoming SMTP command contains a NULL character, it is rejected
3953 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3954 runs initgroups(), so that if the Exim user is in any additional groups,
3955 they will be used during message reception.
3961 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob