1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.348 2006/04/25 14:02:29 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
36 PH/06 A failure to get a lock for a hints database would erroneously always say
37 "Failed to get write lock", even when it was really a read lock.
39 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
40 of 0600. This has been changed to use the value of the lockfile_mode
41 option (which defaults to 0600).
43 PH/08 Applied small patch from the Sieve maintainer.
45 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
46 folder from quota calculations, a direct delivery into this folder messed
47 up the contents of the maildirsize file. This was because the regex was
48 used only to exclude .Trash (or whatever) when the size of the mailbox
49 was calculated. There was no check that a delivery was happening into an
50 excluded directory. This bug has been fixed by ignoring all quota
51 processing for deliveries into excluded directories.
57 PH/01 The code for finding all the local interface addresses on a FreeBSD
58 system running IPv6 was broken. This may well have applied to all BSD
59 systems, as well as to others that have similar system calls. The broken
60 code found IPv4 interfaces correctly, but gave incorrect values for the
61 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
62 that it would not match correctly against @[] and not recognize the IPv6
65 PH/02 The ipliteral router was not recognizing addresses of the form user@
66 [ipv6:....] because it didn't know about the "ipv6:" prefix.
68 PH/03 Added disable_ipv6.
70 PH/04 Changed $reply_address to use the raw form of the headers instead of the
71 decoded form, because it is most often used to construct To: headers
72 lines in autoreplies, and the decoded form may well be syntactically
73 invalid. However, $reply_address has leading white space removed, and all
74 newlines turned into spaces so that the autoreply transport does not
77 PH/05 If group was specified without a user on a router, and no group or user
78 was specified on a transport, the group from the router was ignored.
80 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
81 for visible compile-time settings that can be used to change these
82 numbers, for those that want even more. Backwards compatibility with old
83 spool files has been maintained. However, going back to a previous Exim
84 release will lost any variables that are in spool files.
86 PH/07 Two small changes when running in the test harness: increase delay when
87 passing a TCP/IP connection to a new process, in case the original
88 process has to generate a bounce, and remove special handling of
89 127.0.0.2 (sic), which is no longer necessary.
91 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
92 be the same on different OS.
94 PH/09 Moved a debug statement in filter processing to avoid a race problem when
97 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
98 whether --show-vars was specified or not
100 JJ/02 exipick: Added support for new ACL variable spool format introduced
103 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
104 syntactically invalid From: or Reply-to: line, and a filter used this to
105 generate an autoreply, and therefore failed to obtain an address for the
106 autoreply, Exim could try to deliver to a non-existent relative file
107 name, causing unrelated and misleading errors. What now happens is that
108 it logs this as a hard delivery error, but does not attempt to create a
111 PH/11 The exinext utility has a -C option for testing purposes, but although
112 the given file was scanned by exinext itself; it wasn't being passed on
113 when Exim was called.
115 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
116 an end-of-file indication when reading a command response.
118 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
119 compiled. In many other places in Exim, IPv6 addresses are always
120 recognized, so I have changed this. It also means that IPv4 domain
121 literals of the form [IPV4:n.n.n.n] are now always recognized.
123 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
124 used if the router is not running as root, for example, when verifying at
125 ACL time, or when using -bh. The debugging output from this situation was
126 non-existent - all you got was a failure to exec. I have made two
129 (a) Failures to set uid/gid, the current directory, or a process leader
130 in a subprocess such as that created by queryprogram now generate
131 suitable debugging ouput when -d is set.
133 (b) The queryprogram router detects when it is not running as root,
134 outputs suitable debugging information if -d is set, and then runs
135 the subprocess without attempting to change uid/gid.
137 PH/15 Minor change to Makefile for building test_host (undocumented testing
140 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
141 additional section of a DNS packet that returns MX or SRV records.
142 Instead, it always explicitly searches for A/AAAA records. This avoids
143 major problems that occur when a DNS server includes only records of one
144 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
145 fixed another bug: if SRV records were looked up and the corresponding
146 address records were *not* found in the additional section, the port
147 values from the SRV records were lost.
149 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
150 using the correct key (the original address) when searching the retry
151 rules in order to find which one to use for generating the retry hint.
153 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
154 adding the default one. Similarly, if it contains a Reply-To: header, the
155 errors_reply_to option, if set, is not used.
157 PH/19 When calculating a retry time, Exim used to measure the "time since
158 failure" by looking at the "first failed" field in the retry record. Now
159 it does not use this if it is later than than the arrival time of the
160 message. Instead it uses the arrival time. This makes for better
161 behaviour in cases where some deliveries succeed, thus re-setting the
162 "first failed" field. An example is a quota failure for a huge message
163 when small messages continue to be delivered. Without this change, the
164 "time since failure" will always be short, possible causing more frequent
165 delivery attempts for the huge message than are intended.
166 [Note: This change was subsequently modified - see PH/04 for 4.62.]
168 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
169 $1, $2, $3) because the numerical variables can be reset during some
170 expansion items (e.g. "match"), thereby losing the authentication data.
172 PH/21 Make -bV show the size of off_t variables so that the test suite can
173 decide whether to run tests for quotas > 2G.
175 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
176 mailbox_size, and mailbox_filecount in the appendfile transport. If a
177 filecount value is greater than 2G or if a quota value is greater than 2G
178 on a system where the size of off_t is not greater than 4, a panic error
181 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
182 never match. The debug and -bh output now contains an explicit error
183 message indicating a malformed IPv4 address or mask.
185 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
186 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
189 PH/25 Do not write to syslog when running in the test harness. The only
190 occasion when this arises is a failure to open the main or panic logs
191 (for which there is an explicit test).
193 PH/26 Added the /no_tell option to "control=freeze".
195 PH/27 If a host name lookup failed very early in a connection, for example, if
196 the IP address matched host_lookup and the reverse lookup yielded a name
197 that did not have a forward lookup, an error message of the form "no IP
198 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
199 could be logged. Now it outputs the IP address instead of "NULL".
201 PH/28 An enabling patch from MH: add new function child_open_exim2() which
202 allows the sender and the authenticated sender to be set when
203 submitting a message from within Exim. Since child_open_exim() is
204 documented for local_scan(), the new function should be too.
206 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
207 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
208 results in an empty string is now treated as unset.
210 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
212 PH/31 Added sender_verify_fail logging option.
214 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
215 needed by RFC 822 but not by RFC 2822 was commented out. I have now
216 tidied the source and removed it altogether.
218 PH/33 When a queue run was abandoned because the load average was too high, a
219 log line was always written; now it is written only if the queue_run log
220 selector is set. In addition, the log line for abandonment now contains
221 information about the queue run such as the pid. This is always present
222 in "start" and "stop" lines but was omitted from the "abandon" line.
224 PH/34 Omit spaces between a header name and the colon in the error message that
225 is given when verify = headers_syntax fails (if there are lots of them,
226 the message gets confusing).
228 PH/35 Change the default for dns_check_names_pattern to allow slashes within
229 names, as there are now some PTR records that contain slashes. This check
230 is only to protect against broken name servers that fall over on strange
231 characters, so the fact that it applies to all lookups doesn't matter.
233 PH/36 Now that the new test suite is complete, we can remove some of the
234 special code in Exim that was needed for the old test suite. For example,
235 sorting DNS records because real resolvers return them in an arbitrary
236 order. The new test suite's fake resolver always returns records in the
239 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
240 bounces) except when queue_only is set, to avoid logging races between
241 the different processes.
243 PH/38 Panic-die if .include specifies a non-absolute path.
245 PH/39 A tweak to the "H" retry rule from its user.
247 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
248 a label. They prevented compilation on older perls.
250 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
251 a warning to be raised on newish perls.
253 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
254 on queue. Changes to match documented behaviour of showing count of
255 messages matching specified criteria.
257 PH/40 Changed the default ident timeout from 30s to 5s.
259 PH/41 Added support for the use of login_cap features, on those BSD systems
260 that have them, for controlling the resources used by pipe deliveries.
262 PH/42 The content-scanning code uses fopen() to create files in which to put
263 message data. Previously it was not paying any attention to the mode of
264 the files. Exim runs with umask(0) because the rest of the code creates
265 files with open(), and sets the required mode explicitly. Thus, these
266 files were ending up world-writeable. This was not a big issue, because,
267 being within the spool directory, they were not world-accessible. I have
268 created a function called modefopen, which takes an additional mode
269 argument. It sets umask(777), creates the file, chmods it to the required
270 mode, then resets the umask. All the relevant calls to fopen() in the
271 content scanning code have been changed to use this function.
273 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
274 to 24 hours. This avoids potential overflow problems when processing G
275 and H retry rules. I suspect nobody ever tinkers with this value.
277 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
279 PH/45 When the plaintext authenticator is running as a client, the server's
280 challenges are checked to ensure they are valid base64 strings. By
281 default, the authentication attempt is cancelled if an invalid string is
282 received. Setting client_ignore_invalid_base64 true ignores these errors.
283 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
284 they are received. Thus, the responses can be made to depend on the
285 challenges. If an invalid string is ignored, an empty string is placed in
288 PH/46 Messages that are created by the autoreply transport now contains a
289 References: header, in accordance with RFCs 2822 and 3834.
291 PH/47 Added authenticated_sender_force to the smtp transport.
293 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
295 PH/49 Installed latest patch from the Sieve maintainer.
297 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
298 was also set, the appendfile transport was unnecessarily scanning a
299 directory of message files (e.g. for maildir delivery) to find the count
300 of files (along with the size), even though it did not need this
301 information. It now does the scan only if it needs to find either the
302 size of the count of files.
304 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
306 PH/52 Two bugs concerned with error handling when the smtp transport is
309 (i) Exim was not creating retry information for temporary errors given
310 for individual recipients after the DATA command when the smtp transport
311 was used in LMTP mode. This meant that they could be retried too
312 frequently, and not timed out correctly.
314 (ii) Exim was setting the flag that allows error details to be returned
315 for LMTP errors on RCPT commands, but not for LMTP errors for individual
316 recipients that were returned after the DATA command.
318 PH/53 This is related to PH/52, but is more general: for any failing address,
319 when detailed error information was permitted to be returned to the
320 sender, but the error was temporary, then after the final timeout, only
321 "retry timeout exceeded" was returned. Now it returns the full error as
322 well as "retry timeout exceeded".
324 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
325 do this, and (what is worse) MTAs that accept it.
327 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
328 will now be deprecated.
330 PH/56 New os.c-cygwin from the Cygwin maintainer.
332 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
333 formats (previously only available in exim formats via -bpr, -bpru,
334 and -bpra. Now also available in native and exiqgrep formats)
336 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
337 with very large, slow to parse queues
339 JJ/08 exipick: added ! as generic prefix to negate any criteria format
341 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
343 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
344 responses to authentication challenges, though it was showing the
345 challenges; (ii) I've removed the CR characters from the debug output for
348 PH/58 Allow for the insertion of a newline as well as a space when a string
349 is turned into more than one encoded-word during RFC 2047 encoding. The
350 Sieve code now uses this.
352 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
353 data_4xx, lost_connection, tls_required.
355 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
356 message was being sent as an SMTP response.
358 PH/61 Add -l and -k options to exicyclog.
360 PH/62 When verifying, if an address was redirected to one new address, so that
361 verification continued, and the new address failed or deferred after
362 having set something in $address_data, the value of $address_data was not
363 passed back to the ACL. This was different to the case when no
364 redirection occurred. The value is now passed back in both cases.
366 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
367 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
368 use login_cap.h, so on its own it isn't the distinguishing feature. The
369 new name refers directly to the setclassresources() function.
371 PH/65 Added configuration files for NetBSD3.
373 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
375 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
376 is preferred over IPv4.
378 PH/68 The bounce_return_message and bounce_return_body options were not being
379 honoured for bounces generated during the reception of non-SMTP messages.
380 In particular, this applied to messages rejected by the ACL. This bug has
381 been fixed. However, if bounce_return_message is true and bounce_return_
382 body is false, the headers that are returned for a non-SMTP message
383 include only those that have been read before the error was detected.
384 (In the case of an ACL rejection, they have all been read.)
386 PH/69 The HTML version of the specification is now built in a directory called
387 spec_html instead of spec.html, because the latter looks like a path with
388 a MIME-type, and this confuses some software.
390 PH/70 Catch two compiler warnings in sieve.c.
392 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
393 function verify_get_ident() calls ip_connect() to connect a socket, but
394 if the "connect()" function timed out, ip_connect() used to close the
395 socket. However, verify_get_ident() also closes the socket later, and in
396 between Exim writes to the log, which may get opened at this point. When
397 the socket was closed in ip_connect(), the log could get the same file
398 descriptor number as the socket. This naturally causes chaos. The fix is
399 not to close the socket in ip_connect(); the socket should be closed by
400 the function that creates it. There was only one place in the code where
401 this was missing, in the iplookup router, which I don't think anybody now
402 uses, but I've fixed it anyway.
404 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
405 well as to direct DNS lookups. Otherwise the handling of names in host
406 lists is inconsistent and therefore confusing.
412 PH/01 Two changes to the default runtime configuration:
414 (1) Move the checks for relay_from_hosts and authenticated clients from
415 after to before the (commented out) DNS black list checks.
417 (2) Add control=submission to the relay_from_hosts and authenticated
418 clients checks, on the grounds that messages accepted by these
419 statements are most likely to be submissions.
421 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
423 (1) Generate an error if the third argument for the ${prvs expansion is
426 (2) Treat a missing third argument of ${prvscheck as if it were an empty
429 (3) Reset the variables that are obtained from the first argument of
430 ${prvscheck and used in the second argument before leaving the code,
431 because their memory is reclaimed, so using them afterwards may do
434 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
435 one (it's much easier than Tom thought :-).
437 (5) Because of (4), we can now allow for the use of $prvscheck_result
438 inside the third argument.
440 PH/03 For some reason, the default setting of PATH when running a command from
441 a pipe transport was just "/usr/bin". I have changed it to
444 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
445 anything to be listed in the output from -bV.
447 PH/05 When a filter generated an autoreply, the entire To: header line was
448 quoted in the delivery log line, like this:
450 => >A.N.Other <ano@some.domain> <original@ddress> ...
452 This has been changed so that it extracts the operative address. There
453 may be more than one such address. If so, they are comma-separated, like
456 => >ano@some.domain,ona@other.domain <original@ddress> ...
458 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
459 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
460 not being looked up in the rDNS to get a host name, Exim was showing the
461 IP address twice in Received: lines, even though the IP addresses were
462 identical. For example:
464 Received: from [1.2.3.4] (helo=[1.2.3.4])
466 However, if the real host name was known, it was omitting the HELO data
467 if it matched the actual IP address. This has been tidied up so that it
468 doesn't show the same IP address twice.
470 PH/07 When both +timestamp and +memory debugging was on, the value given by
471 $tod_xxx expansions could be wrong, because the tod_stamp() function was
472 called by the debug printing, thereby overwriting the timestamp buffer.
473 Debugging no longer uses the tod_stamp() function when +timestamp is set.
475 PH/08 When the original message was included in an autoreply transport, it
476 always said "this is a copy of the message, including all the headers",
477 even if body_only or headers_only was set. It now gives an appropriate
480 PH/09 Applied a patch from the Sieve maintainer which:
482 o fixes some comments
483 o adds the (disabled) notify extension core
484 o adds some debug output for the result of if/elsif tests
485 o points to the current vacation draft in the documentation
486 and documents the missing references header update
490 o fixes a bug in processing the envelope test (when testing
491 multiple envelope elements, the last element determinted the
494 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
495 Electronic Mail") by including:
497 Auto-submitted: auto-generated
499 in the messages that it generates (bounce messages and others, such as
500 warnings). In the case of bounce messages for non-SMTP mesages, there was
501 also a typo: it was using "Auto_submitted" (underscore instead of
502 hyphen). Since every message generated by Exim is necessarily in response
503 to another message, thes have all been changed to:
505 Auto-Submitted: auto-replied
507 in accordance with these statements in the RFC:
509 The auto-replied keyword:
511 - SHOULD be used on messages sent in direct response to another
512 message by an automatic process,
514 - MUST NOT be used on manually-generated messages,
516 - MAY be used on Delivery Status Notifications (DSNs) and Message
517 Disposition Notifications (MDNs),
519 - MUST NOT be used on messages generated by automatic or periodic
520 processes, except for messages which are automatic responses to
523 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
524 to the default Received: header definition.
526 PH/12 Added log selector acl_warn_skipped (default on).
528 PH/13 After a successful wildlsearch lookup, discard the values of numeric
529 variables because (a) they are in the wrong storage pool and (b) even if
530 they were copied, it wouldn't work properly because of the caching.
532 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
533 checking when decoding. Apparently there are clients that generate
534 overlong encoded strings. Why am I not surprised?
536 PH/15 If the first argument of "${if match_address" was not empty, but did not
537 contain an "@" character, Exim crashed. Now it writes a panic log message
538 and treats the condition as false.
540 PH/16 In autoreply, treat an empty string for "once" the same as unset.
542 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
543 extension "envelope-auth". The code is finished and in agreement with
544 other implementations, but there is no documentation so far and in fact,
545 nobody wrote the draft yet. This extension is currently #undef'ed, thus
546 not changing the active code.
548 Print executed "if" and "elsif" statements when debugging is used. This
549 helps a great deal to understand what a filter does.
551 Document more things not specified clearly in RFC3028. I had all this
552 sorted out, when out of a sudden new issues came to my mind. Oops."
554 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
557 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
558 canonical form (as documented). However, after a host name lookup from
559 the IP address, check_host() was doing a simple string comparison with
560 addresses acquired from the DNS when checking that the found name did
561 have the original IP as one of its addresses. Since any found IPv6
562 addresses are likely to be in abbreviated form, the comparison could
563 fail. Luckily, there already exists a function for doing the comparison
564 by converting both addresses to binary, so now that is used instead of
567 PH/20 There was another similar case to PH/19, when a complete host name was
568 given in a host list; looking up its IP address could give an abbreviated
569 form, whereas the current host's name might or might not be abbreviated.
570 The same fix has been applied.
576 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
577 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
580 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
581 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
583 PH/03 Typo: missing ".o" in src/pcre/Makefile.
585 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
586 header line, restrict the check to what is listed in RFCs 2369 and 2929.
587 Also, for "Auto-Submitted", treat anything other than "no" as
588 non-personal, in accordance with RFC 3834. (Previously it treated
589 anything starting "auto-" as non-personal.)
591 TF/01 The control=submission/name=... option had a problem with syntax
592 errors if the name included a slash character. The /name= option
593 now slurps the rest of the string, so it can include any characters
594 but it must come last in the list of options (after /sender_retain
597 PH/05 Some modifications to the interface to the fake nameserver for the new
605 TK/01 Added the "success_on_redirect" address verification option. See
606 NewStuff for rationale and an example.
608 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
610 PH/02 Patch to exigrep to allow it to work on syslog lines.
612 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
613 fread() to skip over the body file's header line, because in Cygwin the
614 header line is locked and is inaccessible.
616 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
617 co-exist for some time) to make it clear that it is the Exim ID that is
618 referenced, not the Message-ID: header line.
620 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
621 string_format() function, because snprintf() does not exist on all
624 PH/06 The use of forbid_filter_existstest now also locks out the use of the
625 ${stat: expansion item.
627 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
628 protocol synchronization error", to keep the pedants happy.
630 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
631 well as for IRIX systems, when gcc is being used. See the host.c source
634 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
636 PH/10 Named domain lists were not working if used in a queue_smtp_domains
639 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
640 transport and to the smtp transport in LMTP mode.
642 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
644 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
645 run a filter in a subprocess. This could lead to confusion in subsequent
646 lookups in the parent process. There should also be a search_tidyup() at
647 the end of the subprocess.
649 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
650 only if the host matched helo_try_verify_hosts, which caused the
651 verification to occur when the EHLO/HELO command was issued. The ACL just
652 tested the remembered result. Now, if a previous verification attempt has
653 not happened, "verify = helo" does it there and then.
655 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
657 TK/03 Fix log output including CR from clamd.
659 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
660 exist provoked a memory error which could cause a segfault.
662 PH/15 Installed PCRE 6.2
664 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
666 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
667 of the problem. Specifically, suggested +O2 rather than +O1 for the
670 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
672 PH/20 If a delivery was routed to a non-standard port by means of an SRV
673 record, the port was not correctly logged when the outgoing_port log
674 selector was set (it logged the transort's default port).
676 PH/21 Added support for host-specific ports to manualroute, queryprogram,
677 fallback_hosts, and "hosts" in the smtp transport.
679 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
680 host errors such as "Connection refused".
682 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
683 authentication with radiusclient 0.4.9:
685 - Error returned from rc_read_config was caught wrongly
686 - Username/password not passed on to radius server due to wrong length.
688 The presumption is that some radiusclient API changes for 4.51/PH/17
689 were not taken care of correctly. The code is still untested by me (my
690 Linux distribution still has 0.3.2 of radiusclient), but it was
691 contributed by a Radius user.
693 PH/24 When doing a callout, the value of $domain wasn't set correctly when
694 expanding the "port" option of the smtp transport.
696 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
697 while reading a MIME header. Thanks to Tom Hughes for a patch.
699 PH/24 Include config.h inside local_scan.h so that configuration settings are
702 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
703 that in an ACL for RCPT (for example), you can examine exactly what was
706 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
707 commands, but it was not correctly comparing the address with the actual
708 client host address. Thus, it would show the EHLO address in Received:
709 header lines when this was not necessary.
711 PH/27 Added the % operator to ${eval:}.
713 PH/28 Exim tries to create and chdir to its spool directory when it starts;
714 it should be ignoring failures (because with -C, for example, it has lost
715 privilege). It wasn't ignoring creation failures other than "already
718 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
721 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
722 because an input line was too long, either on its own, or by virtue of
723 too many continuations, the temporary file was not being removed, and the
724 return code was incorrect.
726 PH/31 Missing "BOOL" in function definition in filtertest.c.
728 PH/32 Applied Sieve patches from the maintainer.
730 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
732 PH/33 Added "verify = not_blind".
734 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
735 Local/Makefile (with some defaults set). These are used in built scripts
736 such as exicyclog, but they have never been used in the exim_install
737 script (though there are many overriding facilities there). I have
738 arranged that the exim_install script now takes note of these two
741 PH/35 Installed configuration files for Dragonfly.
743 PH/36 When a locally submitted message by a trusted user did not contain a
744 From: header, and the sender address was obtained from -f or from an SMTP
745 MAIL command, and the trusted user did not use -F to supply a sender
746 name, $originator_name was incorrectly used when constructing a From:
747 header. Furthermore, $originator_name was used for submission mode
748 messages from external hosts without From: headers in a similar way,
749 which is clearly wrong.
751 PH/37 Added control=suppress_local_fixups.
753 PH/38 When log_selector = +received_sender was set, and the addition of the
754 sender made the log line's construction buffer exactly full, or one byte
755 less than full, an overflow happened when the terminating "\n" was
758 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
759 when the result of a list match is failure because a DNS lookup failed.
761 PH/40 RM_COMMAND is now used in the building process.
763 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
764 the "build-* directories that it finds.
766 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
767 address in a domain literal was a prefix of an interface address.
769 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
770 when verifying a sender address, unless rewrite_headers is false.
772 PH/44 Wrote a long comment about why errors_to addresses are verified as
773 recipients, not senders.
775 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
776 the ratelimit ACL was added.
778 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
780 PH/46 Added extra information about PostgreSQL errors to the error string.
782 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
783 avoiding the need to install special zones in a real server. This is
784 backwards compatible; if it can't find the fake resolver, it drops back.
785 Thus, both old and new test suites can be run.
787 TF/02 Added util/ratelimit.pl
789 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
790 clock is set back in time.
792 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
793 Candler <B.Candler@pobox.com>.
795 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
796 OK for addresses that are the result of redirections.
798 PH/48 A number of further additions for the benefit of the new test suite,
799 including a fake gethostbyname() that interfaces to the fake DNS resolver
802 TF/06 The fix for widen_domains has also been applied to qualify_single and
803 search_parents which are the other dnslookup options that can cause
806 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
809 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
811 TF/07 Exim produced the error message "an SRV record indicated no SMTP
812 service" if it encountered an MX record with an empty target hostname.
813 The message is now "an MX or SRV record indicated no SMTP service".
815 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
816 if the DNS of the sending site is misconfigured. This is quite a
817 common situation. This change restores the behaviour of treating a
818 helo verification defer as a failure.
820 PH/51 If self=fail was set on a router, the bounce message did not include the
821 actual error message.
827 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
829 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
830 command itself ended in error, the underlying message about the transport
831 filter timeout was being overwritten with the pipe command error. Now the
832 underlying error message should be appended to the second error message.
834 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
836 PH/02 Reduce the amount of output that "make" produces by default. Full output
837 can still be requested.
839 PH/03 The warning log line about a condition test deferring for a "warn" verb
840 was being output only once per connection, rather than after each
841 occurrence (because it was using the same function as for successful
842 "warn" verbs). This seems wrong, so I have changed it.
844 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
845 it should not have, which might have caused a crash in the right
846 circumstances, but probably never did.
848 PH/04 Installed a modified version of Tony Finch's patch to make submission
849 mode fix the return path as well as the Sender: header line, and to
850 add a /name= option so that you can make the user's friendly name appear
853 TF/03 Added the control = fakedefer ACL modifier.
855 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
856 Mark Lowes for thorough testing.
858 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
860 TK/03 Merged latest SRS patch from Miles Wilton.
862 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
863 with the definition in sysexits.h (which is #included earlier).
864 Fortunately, Exim does not actually use EX_OK. The code used to try to
865 preserve the sysexits.h value, by assumimg that macro definitions were
866 scanned for macro replacements. I have been disabused of this notion,
867 so now the code just undefines EX_OK before #including unistd.h.
869 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
870 in the smtp transport. When a block could not be written in a single
871 write() function, the timeout was being re-applied to each part-write.
872 This seems wrong - if the receiver was accepting one byte at a time it
873 would take for ever. The timeout is now adjusted when this happens. It
874 doesn't have to be particularly precise.
876 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
877 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
879 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
880 without a domain if the check to <postmaster@domain> fails.
882 SC/01 Eximstats: added -xls and the ability to specify output files
883 (patch written by Frank Heydlauf).
885 SC/02 Eximstats: use FileHandles for outputing results.
887 SC/03 Eximstats: allow any combination of xls, txt, and html output.
889 SC/04 Eximstats: fixed display of large numbers with -nvr option
891 SC/05 Eximstats: fixed merging of reports with empty tables.
893 SC/06 Eximstats: added the -include_original_destination flag
895 SC/07 Eximstats: removed tabs and trailing whitespace.
897 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
899 TK/06 MBOX spool code: Add real "From " MBOX separator line
900 so the .eml file is really in mbox format (even though
901 most programs do not really care). Patch from Alex Miller.
903 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
904 The latter is generated from $received_to and is only set if the
905 message has one envelope recipient. SA can use these headers,
906 obviously out-of-the-box. Patch from Alex Miller.
908 PH/08 The ${def test on a variable was returning false if the variable's
909 value was "0", contrary to what the specification has always said!
910 The result should be true unless the variable is empty.
912 PH/09 The syntax error of a character other than { following "${if
913 def:variable_name" (after optional whitespace) was not being diagnosed.
914 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
915 accidental colon was present, for example, could give incorrect results.
917 PH/10 Tidied the code in a number of places where the st_size field of a stat()
918 result is used (not including appendfile, where other changes are about
921 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
922 This involved changing a lot of size variables from int to off_t. It
923 should work with maildirs and everything.
925 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
926 spamd dying while we are connected to it.
928 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
931 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
932 function for BDB 4.3.
934 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
936 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
937 the sources for PCRE have been re-arranged and more files are now
940 PH/15 The code I had for printing potentially long long variables in PH/11
941 above was not the best (it lost precision). The length of off_t variables
942 is now inspected at build time, and an appropriate printing format (%ld
943 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
944 to be "long long int" or "long int". This is needed for the internal
945 formatting function string_vformat().
947 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
948 the configuration file to be ":syslog", then the script "guesses" where
949 the logs files are, rather than using the compiled in default. In our
950 case the guess is not the same as the compiled default, so the script
951 suddenly stopped working when I started to use syslog. The patch checks
952 to see if log_file_path is "". If so, it attempts to read it from exim
953 with no configuration file to get the compiled in version, before it
954 falls back to the previous guessing code."
956 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
957 implementing BATV in an Exim configuration. See NewStuff for the gory
960 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
961 Makefile that are specific to HP-UX.
963 PH/18 If the "use_postmaster" option was set for a recipient callout together
964 with the "random" option, the postmaster address was used as the MAIL
965 FROM address for the random test, but not for the subsequent recipient
966 test. It is now used for both.
968 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
969 patch removes a few documentation additions to RFC 3028, because the
970 latest draft now contains them. It adds the new en;ascii-case comparator
971 and a new error check for 8bit text in MIME parts. Comparator and
972 require names are now matched exactly. I enabled the subaddress
973 extension, but it is not well tested yet (read: it works for me)."
975 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
976 rework some of the code of TK/09 above to avoid the hardwired use of
977 "%lld" and "long long". Replaced the call to snprintf() with a call to
980 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
981 records point to non-existent hosts", "retry timeout exceeded", and
982 "retry time not reached for any host after a long failure period".
984 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
985 experimental DomainKeys support:
987 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
988 (2) On an error such as an illegally used "control", the wrong name for
989 the control was given.
991 These problems did NOT occur unless DomainKeys support was compiled.
993 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
995 PH/24 Added ${if match_ip condition.
997 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
998 databases so that it will be absolutely obvious if a crash occurs in the
999 DB library. This is a regular occurrence (often caused by mis-matched
1002 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1003 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1004 picked up on a user's system that detects such things. There doesn't seem
1005 to be a gcc warning option for this - only an attribute that has to be
1006 put on the function's prototype. It seems that in Fedora Core 4 they have
1007 set this on a number of new functions. No doubt there will be more in due
1010 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1011 specify a transport. However, if an address that was verified by such a
1012 router was the subject of a callout, Exim crashed because it tried to
1013 read the rcpt_include_affixes from the non-existent transport. Now it
1014 just assumes that the setting of that option is false. This bug was
1015 introduced by 4.51/PH/31.
1017 PH/28 Changed -d+all to exclude +memory, because that information is very
1018 rarely of interest, but it makes the output a lot bigger. People tend to
1019 do -d+all out of habit.
1021 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1022 code in os-type was giving problems when libc.so lives in lib64, like on
1025 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1026 aren't the modern standard, and it seems that some systems' include files
1027 don't always have them. Exim was already checking for some of the newer
1028 ones like T_AAAA, and defining it itself. I've added checks for all the
1029 record types that Exim uses.
1031 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1032 not automatically generating a new one, as it is supposed to. This
1033 prevented TLS from working. If the file did exist, but contained invalid
1034 data, a new version was generated, as expected. It was only the case of a
1035 non-existent file that was broken.
1037 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1038 with a change in libdomainkeys > 0.64.
1040 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1041 from DNS. If the selector record carries the flag, it now has
1042 precedence over the domain-wide flag.
1044 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1046 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1047 the use of a port name that isn't defined in /etc/services) occurred, the
1048 message was deferred as in a normal delivery, and thus remained on the
1049 spool, instead of being failed because of the mua_wrapper setting. This
1050 is now fixed, and I tidied up some of the mua_wrapper messages at the
1053 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1054 the messages in a hash of arrays rather than using individual hashes.
1055 This is a bit cleaner and results in dramatic memory savings, albeit
1056 at a slight CPU cost.
1058 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1059 as requested by Marc Sherman.
1061 SC/10 Eximstats: added histograms for user specified patterns as requested
1064 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1066 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1067 fopen() in the content-scanning modules that did not already have it.
1073 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1074 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1076 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1078 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1080 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1081 to test Sieve filters that use "vacation".
1083 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1084 that changes the way the GnuTLS parameters are stored in the cache file.
1085 The new format can be generated externally. For backward compatibility,
1086 if the data in the cache doesn't make sense, Exim assumes it has read an
1087 old-format file, and it generates new data and writes a new file. This
1088 means that you can't go back to an older release without removing the
1091 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1092 work if there are any delivery delays because "one_time" forces the
1093 parent to be marked "delivered", so its unseen clone is never tried
1094 again. For this reason, Exim now forbids the simultaneous setting of
1097 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1098 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1099 entry if you want to know the details. The fix, however, neglected to
1100 consider the case when local delivery batching is involved. The test for
1101 "previously delivered" was not happening when checking to see if an
1102 address could be batched with a previous (undelivered) one; under
1103 certain circumstances this could lead to multiple deliveries to the same
1106 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1107 in its include files, and this causes problems building Exim.
1109 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1110 header_syntax) but Exim was just ignoring anything given after a slash.
1111 In particular, this caused confusion with an attempt to use "verify =
1112 reverse_host_lookup/defer_ok". An error is now given when options are
1113 supplied for verify items that do not have them. (Maybe reverse_host_
1114 lookup should have a defer_ok option, but that's a different point.)
1116 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1117 defined by RFC 821) to 2048, because there were problems with some AUTH
1118 commands, and RFC 1869 says the size should be increased for extended
1119 SMTP commands that take arguments.
1121 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1124 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1125 "unknown" error; now it says that the functionality isn't in the binary.
1127 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1128 an address' error message when a string expansion fails (syntax or
1129 whatever). Otherwise the password may appear in the log. Following change
1130 PH/42 below, there is no longer a chance of it appearing in a bounce
1133 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1135 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1136 of its addresses was ever tried. (Bugzilla bug #2.)
1138 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1139 the result incorrectly in the debug output. (It correctly added a newline
1140 to what was transported.)
1142 TF/01 Added $received_time.
1144 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1145 commented out examples of how to interface to a virus scanner and to
1146 SpamAssassin. Also added commented examples of av_scanner and
1147 spamd_address settings.
1149 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1150 and controls are allowed in which ACLs. There were a couple of minor
1151 errors. Some of the entries in the conditions table (which is a table of
1152 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1153 negation of where the condition IS allowed.
1155 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1157 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1158 header file does not have a version number, so I've had to invent a new
1159 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1160 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1161 radiusclient), but it was contributed by a Radius user.
1163 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1164 files or directories, for OpenSSL.
1166 PH/19 When an Exim process that is running as root has to create an Exim log
1167 file, it does so in a subprocess that runs as exim:exim so as to get the
1168 ownership right at creation (otherwise, other Exim processes might see
1169 the file with the wrong ownership). There was no test for failure of this
1170 fork() call, which would lead to the process getting stuck as it waited
1171 for a non-existent subprocess. Forks do occasionally fail when resources
1172 run out. I reviewed all the other calls to fork(); they all seem to check
1175 PH/20 When checking for unexpected SMTP input at connect time (before writing
1176 the banner), Exim was not dealing correctly with a non-positive return
1177 from the read() function. If the client had disconnected by this time,
1178 the result was a log entry for a synchronization error with an empty
1179 string after "input=" when read() returned zero. If read() returned -1
1180 (an event I could not check), uninitialized data bytes were printed.
1181 There were reports of junk text (parts of files, etc) appearing after
1184 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1186 PH/22 Added support for macro redefinition, and (re)definition in between
1187 driver and ACL definitions.
1189 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1190 forgetting to use the resulting value; it was using the unexpanded value.
1192 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1193 hadn't been configured. The fix is from Juergen Kreileder, who
1194 understands it better than I do:
1196 "Here's what I see happening with three configured cyrus_sasl
1197 authenticators configured (plain, login, cram-md5):
1199 On startup auth_cyrus_sasl_init() gets called for each of these.
1200 This means three calls to sasl_listmech() without a specified mech_list.
1201 => SASL tests which mechs of all available mechs actually work
1202 => three warnings about OTP not working
1203 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1205 With the patch, sasl_listmech() also gets called three times. But now
1206 SASL's mech_list option is set to the server_mech specified in the the
1207 authenticator. Or in other words, the answer from sasl_listmech()
1208 gets limited to just the mech you're testing for (which is different
1210 => the return list contains just 'plain' or 'login', 'cram-md5' or
1211 nothing depending on the value of ob->server_mech.
1213 I've just tested the patch: Authentication still works fine,
1214 unavailable mechs specified in the exim configuration are still
1215 caught, and the auth.log warnings about OTP are gone."
1217 PH/25 When debugging is enabled, the contents of the command line are added
1218 to the debugging output, even when log_selector=+arguments is not
1221 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1222 answer is "GNU", and only if the return is "GNU/something" is the answer
1225 PH/27 $acl_verify_message is now set immediately after the failure of a
1226 verification in an ACL, and so is available in subsequent modifiers. In
1227 particular, the message can be preserved by coding like this:
1229 warn !verify = sender
1230 set acl_m0 = $acl_verify_message
1232 Previously, $acl_verify_message was set only while expanding "message"
1233 and "log_message" when a very denied access.
1235 PH/28 Modified OS/os.c-Linux with
1237 -#ifndef OS_LOAD_AVERAGE
1238 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1240 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1241 nomenclature these days.)
1243 PH/29 Installed patch from the Sieve maintainer that adds the options
1244 sieve_useraddress and sieve_subaddress to the redirect router.
1246 PH/30 In these circumstances:
1247 . Two addresses routed to the same list of hosts;
1248 . First host does not offer TLS;
1249 . First host accepts first address;
1250 . First host gives temporary error to second address;
1251 . Second host offers TLS and a TLS session is established;
1252 . Second host accepts second address.
1253 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1254 and peerdn, if requested) that were in fact used only for the second
1257 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1258 attention to any local part prefix or suffix that was matched by the
1259 router that accepted the address. It now behaves in the same way as it
1260 does for delivery: the affixes are removed from the local part unless
1261 rcpt_include_affixes is set on the transport.
1263 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1264 timeout during the DATA phase of an incoming message.
1266 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1267 applied a patch sanctioned by the Sieve maintainer.
1269 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1270 the uid or gid is negative. A case of a negative gid caused this to be
1271 noticed. The fix allows for either to be negative.
1273 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1274 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1277 PH/36 The change PH/12 above was broken. Fixed it.
1279 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1280 the grounds that routing the same address twice would always produce the
1281 same answer. This might have been true once, but it is certainly no
1282 longer true now. Routing a child address may depend on the previous
1283 routing that produced that child. Some complicated redirection strategies
1284 went wrong when messages had multiple recipients, and made Exim's
1285 behaviour dependent on the order in which the addresses were given.
1287 I have moved the duplicate checking until after the routing is complete.
1288 Exim scans the addresses that are assigned to local and remote
1289 transports, and removes any duplicates. This means that more work will be
1290 done, as duplicates will always all be routed, but duplicates are
1291 presumably rare, so I don't expect this is of any significance.
1293 For deliveries to pipes, files, and autoreplies, the duplicate checking
1294 still happens during the routing process, since they are not going to be
1297 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1298 It corrects a timeout issue with spamd. This is Ian's comment: "The
1299 background is that sometimes spamd either never reads data from a
1300 connection it has accepted, or it never writes response data. The exiscan
1301 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1302 blindly assumes that writes won't block so it may never time out."
1304 PH/39 Allow G after quota size as well as K and M.
1306 PH/40 The value set for $authenticated_id in an authenticator may not contain
1307 binary zeroes or newlines because the value is written to log lines and
1308 to spool files. There was no check on this. Now the value is run through
1309 the string_printing() function so that such characters are converted to
1310 printable escape sequences.
1312 PH/41 $message_linecount is a new variable that contains the total number of
1313 lines in the message. Compare $body_linecount, which is the count for the
1316 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1317 bounce and delay warning messages, except in certain special cases, which
1320 (a) An SMTP error message from a remote host;
1321 (b) A message specified in a :fail: redirection;
1322 (c) A message specified in a "fail" command in a system filter;
1323 (d) A message specified in a FAIL return from the queryprogram router;
1324 (e) A message specified by the cannot_route_message router option.
1326 In these cases only, Exim does include the error details in bounce and
1327 warning messages. There are also a few cases where bland messages such
1328 as "unrouteable address" or "local delivery error" are given.
1330 PH/43 $value is now also set for the "else" part of a ${run expansion.
1332 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1333 being worked on, but at least Exim now implements the latest version to
1336 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1337 process to complete was treated as a delivery failure, a timeout while
1338 writing the message to the pipe was logged, but erroneously treated as a
1339 successful delivery. Such timeouts include transport filter timeouts. For
1340 consistency with the overall process timeout, these timeouts are now
1341 treated as errors, giving rise to delivery failures by default. However,
1342 there is now a new Boolean option for the pipe transport called
1343 timeout_defer, which, if set TRUE, converts the failures into defers for
1344 both kinds of timeout. A transport filter timeout is now identified in
1347 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1348 systems where "make" and "gmake" are different, calling "gmake" at top
1349 level broke things. I've arranged for the value of $(MAKE) to be passed
1350 from the Makefile to this script so that it can call the same version of
1354 A note about Exim versions 4.44 and 4.50
1355 ----------------------------------------
1357 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1358 changes of various kinds. As a consequence, a big documentation update was
1359 needed. This delayed the release for rather longer than seemed good, especially
1360 in the light of a couple of (minor) security issues. Therefore, the changes
1361 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1362 release. So 4.44 and 4.50 are in effect two different branches that both start
1365 I have left the 4.50 change log unchanged; it contains all the changes since
1366 4.43. The change log for 4.44 is below; many of its items are identical to
1367 those for 4.50. This seems to be the most sensible way to preserve the
1368 historical information.
1374 1. Minor wording change to the doc/README.SIEVE file.
1376 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1377 computation of the current number of files was incorrect.
1379 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1380 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1383 4. Give more explanation in the error message when the command for a transport
1384 filter fails to execute.
1386 5. There are several places where Exim runs a non-Exim command in a
1387 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1388 was being done only for the command run by the queryprogram router. It is
1389 now done for all such subprocesses. The other cases are: ${run, transport
1390 filters, and the commands run by the lmtp and pipe transports.
1392 6. Added CONFIGURE_GROUP build-time option.
1394 7. Some older OS have a limit of 256 on the maximum number of file
1395 descriptors. Exim was using setrlimit() to set 1000 as a large value
1396 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1397 systems. I've change it so that if it can't get 1000, it tries for 256.
1399 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1400 was an oversight, and furthermore, ever since the addition of extra
1401 controls (e.g. 4.43/32), the checks on when to allow different forms of
1402 "control" were broken. There should now be diagnostics for all cases when a
1403 control that does not make sense is encountered.
1405 9. Added the /retain_sender option to "control=submission".
1407 10. $recipients is now available in the predata ACL (oversight).
1409 11. Tidy the search cache before the fork to do a delivery from a message
1410 received from the command line. Otherwise the child will trigger a lookup
1411 failure and thereby defer the delivery if it tries to use (for example) a
1412 cached ldap connection that the parent has called unbind on.
1414 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1415 of $address_data from the recipient verification was clobbered by the
1416 sender verification.
1418 13. The value of address_data from a sender verification is now available in
1419 $sender_address_data in subsequent conditions in the ACL statement.
1421 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1423 15. Added a new option "connect=<time>" to callout options, to set a different
1426 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1427 was its contents. (It was OK if the option was not defined at all.)
1429 17. A "Completed" log line is now written for messages that are removed from
1430 the spool by the -Mrm option.
1432 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1433 information about exactly what failed.
1435 19. Added -dd to debug only the daemon process.
1437 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1438 handles timeouts, both on the server side and network timeouts. Renamed the
1439 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1441 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1442 It is now set to "smtps".
1444 22. $host_address is now set to the target address during the checking of
1445 ignore_target_hosts.
1447 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1448 being passed; this would have caused $sender_host_name to have been used if
1449 matching the list had actually called for a host name (not very likely,
1450 since this list is usually IP addresses). A host name is now passed as
1453 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1454 code that specifies a non-restarting handler (typically sigaction() in
1455 modern systems) in an attempt to fix a rare and obscure crash bug.
1457 25. Narrowed the window for a race in the daemon that could cause it to ignore
1458 SIGCHLD signals. This is not a major problem, because they are used only to
1459 wake it up if nothing else does.
1461 26. A malformed maildirsize file could cause Exim to calculate negative values
1462 for the mailbox size or file count. Odd effects could occur as a result.
1463 The maildirsize information is now recalculated if the size or filecount
1466 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1467 support for a long time. Removed HAVE_SYS_VFS_H.
1469 28. Installed the latest version of exipick from John Jetmore.
1471 29. In an address list, if the pattern was not a regular expression, an empty
1472 subject address (from a bounce message) matched only if the pattern was an
1473 empty string. Non-empty patterns were not even tested. This was the wrong
1474 because it is perfectly reasonable to use an empty address as part of a
1475 database query. An empty address is now tested by patterns that are
1476 lookups. However, all the other forms of pattern expect the subject to
1477 contain a local part and a domain, and therefore, for them, an empty
1478 address still always fails if the pattern is not itself empty.
1480 30. Exim went into a mad DNS loop when attempting to do a callout where the
1481 host was specified on an smtp transport, and looking it up yielded more
1482 than one IP address.
1484 31. Re-factored the code for checking spool and log partition space into a
1485 function that finds that data and another that does the check. The former
1486 is then used to implement four new variables: $spool_space, $log_space,
1487 $spool_inodes, and $log_inodes.
1489 32. The RFC2047 encoding function was originally intended for short strings
1490 such as real names; it was not keeping to the 75-character limit for
1491 encoded words that the RFC imposes. It now respects the limit, and
1492 generates multiple encoded words if necessary. To be on the safe side, I
1493 have increased the buffer size for the ${rfc2047: expansion operator from
1496 33. It is now permitted to omit both strings after an "if" condition; if the
1497 condition is true, the result is "true". As before, when the second string
1498 is omitted, a false condition yields an empty string. This makes it less
1499 cumbersome to write custom ACL and router conditions.
1501 34. Failure to deliver a bounce message always caused it to be frozen, even if
1502 there was an errors_to setting on the router. The errors_to setting is now
1505 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1506 canonical form (fully expanded) before being placed in
1507 $sender_host_address.
1509 36. The table in the code that translates DNS record types into text (T_A to
1510 "A" for instance) was missing entries for NS and CNAME. It is just possible
1511 that this could have caused confusion if both these types were looked up
1512 for the same domain, because the text type is used as part of Exim's
1513 per-process caching. But the chance of anyone hitting this buglet seems
1516 37. The dnsdb lookup has been extended in a number of ways.
1518 (1) There is a new type, "zns", which walks up the domain tree until it
1519 finds some nameserver records. It should be used with care.
1521 (2) There is a new type, "mxh", which is like "mx" except that it returns
1522 just the host names, not the priorities.
1524 (3) It is now possible to give a list of domains (or IP addresses) to be
1525 looked up. The behaviour when one of the lookups defers can be
1526 controlled by a keyword.
1528 (4) It is now possible to specify the separator character for use when
1529 multiple records are returned.
1531 38. The dnslists ACL condition has been extended: it is now possible to supply
1532 a list of IP addresses and/or domains to be looked up in a particular DNS
1535 39. Added log_selector=+queue_time_overall.
1537 40. When running the queue in the test harness, wait just a tad after forking a
1538 delivery process, to get repeatability of debugging output.
1540 41. Include certificate and key file names in error message when GnuTLS fails
1541 to set them up, because the GnuTLS error message doesn't include the name
1542 of the failing file when there is a problem reading it.
1544 42. Allow both -bf and -bF in the same test run.
1546 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1548 44. The "Exiscan patch" is now merged into the mainline Exim source.
1550 45. Sometimes the final signoff response after QUIT could fail to get
1551 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1552 < 0 before doing a fflush(). This bug looks as though it goes back to the
1553 introduction of TLS in release 3.20, but "sometimes" must have been rare
1554 because the tests only now provoked it.
1556 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1557 (this can affect the format of dates).
1559 47. exim_tidydb, when checking for the continued existence of a message for
1560 which it has found a message-specific retry record, was not finding
1561 messages that were in split spool directories. Consequently, it was
1562 deleting retry records that should have stayed in existence.
1564 48. Steve fixed some bugs in eximstats.
1566 49. The SPA authentication driver was not abandoning authentication and moving
1567 on to the next authenticator when an expansion was forced to fail,
1568 contradicting the general specification for all authenticators. Instead it
1569 was generating a temporary error. It now behaves as specified.
1571 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1572 (the order specifies the preference for clients). The order is now AES256,
1573 AES128, 3DES, ARCFOUR128.
1575 51. Small patch to Sieve code - explicitly set From: when generating an
1578 52. Exim crashed if a remote delivery caused a very long error message to be
1579 recorded - for instance if somebody sent an entire SpamAssassin report back
1580 as a large number of 550 error lines. This bug was coincidentally fixed by
1581 increasing the size of one of Exim's internal buffers (big_buffer) that
1582 happened as part of the Exiscan merge. However, to be on the safe side, I
1583 have made the code more robust (and fixed the comments that describe what
1586 53. Now that there can be additional text after "Completed" in log lines (if
1587 the queue_time_overall log selector is set), a one-byte patch to exigrep
1588 was needed to allow it to recognize "Completed" as not the last thing in
1591 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1592 patch that reportedly fixes this has been added. I am not expert enough to
1593 create a test for it. This is what the patch creator wrote:
1595 "I found a little strange behaviour of ldap code when working with
1596 Windows 2003 AD Domain, where users was placed in more than one
1597 Organization Units. When I tried to give exim partial DN, the exit code
1598 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1599 But simultaneously result of request was absolutely normal ldap result,
1600 so I produce this patch..."
1602 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1603 REFERENCE, so I have modified the code to exclude the patch when that macro
1606 55. Some experimental protocols are using DNS PTR records for new purposes. The
1607 keys for these records are domain names, not reversed IP addresses. The
1608 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1609 leaves it alone. Component reversal etc. now happens only for IP addresses.
1612 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1614 57. Double the size of the debug message buffer (to 2048) so that more of very
1615 long debug lines gets shown.
1617 58. The exicyclog utility now does better if the number of log files to keep
1618 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1620 59. Two changes related to the smtp_active_hostname option:
1622 (1) $smtp_active_hostname is now available as a variable.
1623 (2) The default for smtp_banner uses $smtp_active_hostname instead
1624 of $primary_hostname.
1626 60. The host_aton() function is supposed to be passed a string that is known
1627 to be a valid IP address. However, in the case of IPv6 addresses, it was
1628 not checking this. This is a hostage to fortune. Exim now panics and dies
1629 if the condition is not met. A case was found where this could be provoked
1630 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1631 components; fortuitously, this particular loophole had already been fixed
1632 by change 4.50/55 above.
1634 If there are any other similar loopholes, the new check in host_aton()
1635 itself should stop them being exploited. The report I received stated that
1636 data on the command line could provoke the exploit when Exim was running as
1637 exim, but did not say which command line option was involved. All I could
1638 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1639 running as the user.
1642 61. There was a buffer overflow vulnerability in the SPA authentication code
1643 (which came originally from the Samba project). I have added a test to the
1644 spa_base64_to_bits() function which I hope fixes it.
1647 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1648 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1650 63. The daemon start-up calls getloadavg() while still root for those OS that
1651 need the first call to be done as root, but it missed one case: when
1652 deliver_queue_load_max is set with deliver_drop_privilege. This is
1653 necessary for the benefit of the queue runner, because there is no re-exec
1654 when deliver_drop_privilege is set.
1656 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1657 This has been fixed.
1659 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1660 was in use, was not putting the data itself into the right store pool;
1661 consequently, it could be overwritten for a subsequent message in the same
1662 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1665 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1667 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1668 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1669 the function were treating the return as a boolean value, which happened to
1670 work because 0=false and not-0=true, but is not correct code.
1672 68. The host_aton() function was not handling scoped IPv6 addresses (those
1673 with, for example, "%eth0" on the end) correctly.
1675 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1676 negated items (that is, ~something) in unsigned ints. Some compilers
1677 apparently mutter when there is no cast.
1679 70. If an address verification called from an ACL failed, and did not produce a
1680 user-specific message (i.e. there was only a "system" message), nothing was
1681 put in $acl_verify_message. In this situation, it now puts the system
1684 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1685 session; change 4.31/43 added the unwanted input to the log line - except
1686 that it did not do this in the start of session case. It now does.
1688 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1689 This is wrong and can cause the other end to generate a synchronization
1690 error if it is another Exim or anything else that does the synchronization
1691 check. A QUIT command is no longer sent after a timeout.
1693 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1694 during host lookups.
1696 74. The defer_ok option of callout verification was not working if it was used
1697 when verifying addresses in header lines, that is, for this case:
1699 verify = header_sender/callout=defer_ok
1701 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1702 those file descriptors could be used for SMTP connections. If anything
1703 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1704 could be sent to the SMTP client, causing chaos. The daemon now opens
1705 stdin, stdout, and stderr to /dev/null when it puts itself into the
1708 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1709 log by default. The user can override this with suitable Perl magic.
1711 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1712 the log message when discard triggers, was not working for the DATA ACL or
1713 for the non-SMTP ACL.
1715 78. Error message wording change in sieve.c.
1717 79. If smtp_accept_max_per_host was set, the number of connections could be
1718 restricted to fewer than expected, because the daemon was trying to set up
1719 a new connection before checking whether the processes handling previous
1720 connections had finished. The check for completed processes is now done
1721 earlier. On busy systems, this bug wouldn't be noticed because something
1722 else would have woken the daemon, and it would have reaped the completed
1725 80. If a message was submitted locally by a user whose login name contained one
1726 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1727 It caused a spool format error. I have fixed the spool reading code. A
1728 related problem was that the "from" clause in the Received: line became
1729 illegal because of the space(s). It is now covered by ${quote_local_part.
1731 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1734 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1736 83. Patch from Sieve maintainer to fix unterminated string problem in
1737 "vacation" handling.
1739 84. Some minor changes to the Linux configuration files to help with other
1740 OS variants using glibc.
1742 85. One more patch for Sieve to update vacation handling to latest spec.
1745 ----------------------------------------------------
1746 See the note above about the 4.44 and 4.50 releases.
1747 ----------------------------------------------------
1753 1. Change 4.43/35 introduced a bug that caused file counts to be
1754 incorrectly computed when quota_filecount was set in an appendfile
1757 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1758 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1761 3. Give more explanation in the error message when the command for a transport
1762 filter fails to execute.
1764 4. There are several places where Exim runs a non-Exim command in a
1765 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1766 was being done only for the command run by the queryprogram router. It is
1767 now done for all such subprocesses. The other cases are: ${run, transport
1768 filters, and the commands run by the lmtp and pipe transports.
1770 5. Some older OS have a limit of 256 on the maximum number of file
1771 descriptors. Exim was using setrlimit() to set 1000 as a large value
1772 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1773 systems. I've change it so that if it can't get 1000, it tries for 256.
1775 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1776 was an oversight, and furthermore, ever since the addition of extra
1777 controls (e.g. 4.43/32), the checks on when to allow different forms of
1778 "control" were broken. There should now be diagnostics for all cases when a
1779 control that does not make sense is encountered.
1781 7. $recipients is now available in the predata ACL (oversight).
1783 8. Tidy the search cache before the fork to do a delivery from a message
1784 received from the command line. Otherwise the child will trigger a lookup
1785 failure and thereby defer the delivery if it tries to use (for example) a
1786 cached ldap connection that the parent has called unbind on.
1788 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1789 of $address_data from the recipient verification was clobbered by the
1790 sender verification.
1792 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1793 was its contents. (It was OK if the option was not defined at all.)
1795 11. A "Completed" log line is now written for messages that are removed from
1796 the spool by the -Mrm option.
1798 12. $host_address is now set to the target address during the checking of
1799 ignore_target_hosts.
1801 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1802 being passed; this would have caused $sender_host_name to have been used if
1803 matching the list had actually called for a host name (not very likely,
1804 since this list is usually IP addresses). A host name is now passed as
1807 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1808 code that specifies a non-restarting handler (typically sigaction() in
1809 modern systems) in an attempt to fix a rare and obscure crash bug.
1811 15. Narrowed the window for a race in the daemon that could cause it to ignore
1812 SIGCHLD signals. This is not a major problem, because they are used only to
1813 wake it up if nothing else does.
1815 16. A malformed maildirsize file could cause Exim to calculate negative values
1816 for the mailbox size or file count. Odd effects could occur as a result.
1817 The maildirsize information is now recalculated if the size or filecount
1820 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1821 support for a long time. Removed HAVE_SYS_VFS_H.
1823 18. Updated exipick to current release from John Jetmore.
1825 19. Allow an empty sender to be matched against a lookup in an address list.
1826 Previously the only cases considered were a regular expression, or an
1829 20. Exim went into a mad DNS lookup loop when doing a callout where the
1830 host was specified on the transport, if the DNS lookup yielded more than
1833 21. The RFC2047 encoding function was originally intended for short strings
1834 such as real names; it was not keeping to the 75-character limit for
1835 encoded words that the RFC imposes. It now respects the limit, and
1836 generates multiple encoded words if necessary. To be on the safe side, I
1837 have increased the buffer size for the ${rfc2047: expansion operator from
1840 22. Failure to deliver a bounce message always caused it to be frozen, even if
1841 there was an errors_to setting on the router. The errors_to setting is now
1844 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1845 canonical form (fully expanded) before being placed in
1846 $sender_host_address.
1848 24. Updated eximstats to version 1.33
1850 25. Include certificate and key file names in error message when GnuTLS fails
1851 to set them up, because the GnuTLS error message doesn't include the name
1852 of the failing file when there is a problem reading it.
1854 26. Expand error message when OpenSSL has problems setting up cert/key files.
1857 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1858 (this can affect the format of dates).
1860 28. exim_tidydb, when checking for the continued existence of a message for
1861 which it has found a message-specific retry record, was not finding
1862 messages that were in split spool directories. Consequently, it was
1863 deleting retry records that should have stayed in existence.
1865 29. eximstats updated to version 1.35
1866 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1867 1.35 - bugfix such that pie charts by volume are generated correctly
1869 30. The SPA authentication driver was not abandoning authentication and moving
1870 on to the next authenticator when an expansion was forced to fail,
1871 contradicting the general specification for all authenticators. Instead it
1872 was generating a temporary error. It now behaves as specified.
1874 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1875 (the order specifies the preference for clients). The order is now AES256,
1876 AES128, 3DES, ARCFOUR128.
1878 31. Small patch to Sieve code - explicitly set From: when generating an
1881 32. Exim crashed if a remote delivery caused a very long error message to be
1882 recorded - for instance if somebody sent an entire SpamAssassin report back
1883 as a large number of 550 error lines. This bug was coincidentally fixed by
1884 increasing the size of one of Exim's internal buffers (big_buffer) that
1885 happened as part of the Exiscan merge. However, to be on the safe side, I
1886 have made the code more robust (and fixed the comments that describe what
1889 33. Some experimental protocols are using DNS PTR records for new purposes. The
1890 keys for these records are domain names, not reversed IP addresses. The
1891 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1892 leaves it alone. Component reversal etc. now happens only for IP addresses.
1895 34. The host_aton() function is supposed to be passed a string that is known
1896 to be a valid IP address. However, in the case of IPv6 addresses, it was
1897 not checking this. This is a hostage to fortune. Exim now panics and dies
1898 if the condition is not met. A case was found where this could be provoked
1899 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1900 components; fortuitously, this particular loophole had already been fixed
1901 by change 4.50/55 or 4.44/33 above.
1903 If there are any other similar loopholes, the new check in host_aton()
1904 itself should stop them being exploited. The report I received stated that
1905 data on the command line could provoke the exploit when Exim was running as
1906 exim, but did not say which command line option was involved. All I could
1907 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1908 running as the user.
1911 35. There was a buffer overflow vulnerability in the SPA authentication code
1912 (which came originally from the Samba project). I have added a test to the
1913 spa_base64_to_bits() function which I hope fixes it.
1916 36. The daemon start-up calls getloadavg() while still root for those OS that
1917 need the first call to be done as root, but it missed one case: when
1918 deliver_queue_load_max is set with deliver_drop_privilege. This is
1919 necessary for the benefit of the queue runner, because there is no re-exec
1920 when deliver_drop_privilege is set.
1922 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1923 was in use, was not putting the data itself into the right store pool;
1924 consequently, it could be overwritten for a subsequent message in the same
1925 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1928 38. Sometimes the final signoff response after QUIT could fail to get
1929 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1930 < 0 before doing a fflush(). This bug looks as though it goes back to the
1931 introduction of TLS in release 3.20, but "sometimes" must have been rare
1932 because the tests only now provoked it.
1938 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1939 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1940 Now it returns an integer. A number of calls were still expecting a T/F
1941 return. Fortuitously, in all cases, the tests worked in OK situations,
1942 which is the norm. However, things would have gone wrong on any write
1943 failures on the smtp file descriptor. This function is used when sending
1944 messages over SMTP and also when doing verify callouts.
1946 2. When Exim is called to do synchronous delivery of a locally submitted
1947 message (the -odf or -odi options), it no longer closes stderr before doing
1950 3. Implemented the mua_wrapper option.
1952 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1954 5. Implemented the functions header_remove(), header_testname(),
1955 header_add_at_position(), and receive_remove_recipient(), and exported them
1958 6. If an ACL "warn" statement specified the addition of headers, Exim already
1959 inserted X-ACL-Warn: at the start if there was no header name. However, it
1960 was not making this test for the second and subsequent header lines if
1961 there were newlines in the string. This meant that an invalid header could
1962 be inserted if Exim was badly configured.
1964 7. Allow an ACL "warn" statement to add header lines at the start or after all
1965 the Received: headers, as well as at the end.
1967 8. Added the rcpt_4xx retry error code.
1969 9. Added postmaster_mailfrom=xxx to callout verification option.
1971 10. Added mailfrom=xxxx to the callout verification option, for verify=
1974 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1975 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1977 12. Inserted some casts to stop certain compilers warning when using pointer
1978 differences as field lengths or precisions in printf-type calls (mostly
1979 affecting debugging statements).
1981 13. Added optional readline() support for -be (dynamically loaded).
1983 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1984 same clock tick as a message's arrival, so that its received time was the
1985 same as the "first fail" time on the retry record, and that message
1986 remained on the queue past the ultimate address timeout, every queue runner
1987 would try a delivery (because it was past the ultimate address timeout) but
1988 after another failure, the ultimate address timeout, which should have then
1989 bounced the address, did not kick in. This was a "< instead of <=" error;
1990 in most cases the first failure would have been in the next clock tick
1991 after the received time, and all would be well.
1993 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1994 being recognized when the domain list was tested by the match_domain
1995 condition in an expansion string.
1997 16. Added the ${str2b64: operator.
1999 17. Exim was always calling setrlimit() to set a large limit for the number of
2000 processes, without checking whether the existing limit was already
2001 adequate. (It did check for the limit on file descriptors.) Furthermore,
2002 errors from getrlimit() and setrlimit() were being ignored. Now they are
2003 logged to the main and panic logs, but Exim does carry on, to try to do its
2004 job under whatever limits there are.
2006 18. Imported PCRE 5.0.
2008 19. Trivial typo in log message " temporarily refused connection" (the leading
2011 20. If the log selector return_path_on_delivery was set and an address was
2012 redirected to /dev/null, the delivery process crashed because it assumed
2013 that a return path would always be set for a "successful" delivery. In this
2014 case, the whole delivery is bypassed as an optimization, and therefore no
2017 21. Internal re-arrangement: the function for sending a challenge and reading
2018 a response while authentication was assuming a zero-terminated challenge
2019 string. It's now changed to take a pointer and a length, to allow for
2020 binary data in such strings.
2022 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2024 23. Exim was not respecting finduser_retries when seeking the login of the
2025 uid under which it was called; it was always trying 10 times. (The default
2026 setting of finduser_retries is zero.) Also, it was sleeping after the final
2027 failure, which is pointless.
2029 24. Implemented tls_on_connect_ports.
2031 25. Implemented acl_smtp_predata.
2033 26. If the domain in control=submission is set empty, Exim assumes that the
2034 authenticated id is a complete email address when it generates From: or
2035 Sender: header lines.
2037 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2038 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2039 chown and chgrp in /bin and hostname in /usr/bin.
2041 28. Exim was keeping the "process log" file open after each use, just as it
2042 does for the main log. This opens the possibility of it remaining open for
2043 long periods when the USR1 signal hits a daemon. Occasional processlog
2044 errors were reported, that could have been caused by this. Anyway, it seems
2045 much more sensible not to leave this file open at all, so that is what now
2048 29. The long-running daemon process does not normally write to the log once it
2049 has entered its main loop, and it closes the log before doing so. This is
2050 so that log files can straightforwardly be renamed and moved. However,
2051 there are a couple of unusual error situations where the daemon does write
2052 log entries, and I had neglected to close the log afterwards.
2054 30. The text of an SMTP error response that was received during a remote
2055 delivery was being truncated at 512 bytes. This is too short for some of
2056 the long messages that one sometimes sees. I've increased the limit to
2059 31. It is now possible to make retry rules that apply only when a message has a
2060 specific sender, in particular, an empty sender.
2062 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2063 it possible to be selective about when SMTP synchronization is enforced.
2065 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2067 32. Implemented hosts_connection_nolog.
2069 33. Added an ACL for QUIT.
2071 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2074 35. Added mailbox_size and mailbox_filecount to appendfile.
2076 36. Added control = no_multiline_responses to ACLs.
2078 37. There was a bug in the logic of the code that waits for the clock to tick
2079 in the case where the clock went backwards by a substantial amount such
2080 that the microsecond fraction of "now" was more than the microsecond
2081 fraction of "then" (but the whole seconds number was less).
2083 38. Added support for the libradius Radius client library this is found on
2084 FreeBSD (previously only the radiusclient library was supported).
2090 1. When certain lookups returned multiple values in the form name=value, the
2091 quoting of the values was not always being done properly. Specifically:
2092 (a) If the value started with a double quote, but contained no whitespace,
2094 (b) If the value contained whitespace other than a space character (i.e.
2095 tabs or newlines or carriage returns) it was not quoted.
2096 This fix has been applied to the mysql and pgsql lookups by writing a
2097 separate quoting function and calling it from the lookup code. The fix
2098 should probably also be applied to nisplus, ibase and oracle lookups, but
2099 since I cannot test any of those, I have not disturbed their existing code.
2101 2. A hit in the callout cache for a specific address caused a log line with no
2102 reason for rejecting RCPT. Now it says "Previous (cached) callout
2103 verification failure".
2105 3. There was an off-by-one bug in the queryprogram router. An over-long
2106 return line was truncated at 256 instead of 255 characters, thereby
2107 overflowing its buffer with the terminating zero. As well as fixing this, I
2108 have increased the buffer size to 1024 (and made a note to document this).
2110 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2111 when Exim is waiting for an SMTP response from a remote server, Exim
2112 restarts its select() call on the socket, thereby resetting its timeout.
2113 This is not a problem when such interrupts are rare. Somebody set up a cron
2114 job to run exiwhat every 2 minutes, which is less than the normal select()
2115 timeout (5 or 10 minutes). This meant that the select() timeout never
2116 kicked in because it was always reset. I have fixed this by comparing the
2117 time when an interrupt arrives with the time at the start of the first call
2118 to select(). If more time than the timeout has elapsed, the interrupt is
2119 treated as a timeout.
2121 5. Some internal re-factoring in preparation for the addition of Sieve
2122 extensions (by MH). In particular, the "personal" test is moved to a
2123 separate function, and given an option for scanning Cc: and Bcc: (which is
2124 not set for Exim filters).
2126 6. When Exim created an email address using the login of the caller as the
2127 local part (e.g. when creating a From: or Sender: header line), it was not
2128 quoting the local part when it contained special characters such as @.
2130 7. Installed new OpenBSD configuration files.
2132 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2133 try to make them clearer.
2135 9. Callout options, other than the timeout value, were being ignored when
2136 verifying sender addresses in header lines. For example, when using
2138 verify = header_sender/callout=no_cache
2140 the cache was (incorrectly) being used.
2142 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2143 only the Cygwin environment.
2145 11. When return_path_on_delivery was set as a log selector, if different remote
2146 addresses in the same message used different return paths and parallel
2147 remote delivery occurred, the wrong values would sometimes be logged.
2148 (Whenever a remote delivery process finished, the return path value from
2149 the most recently started remote delivery process was logged.)
2151 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2152 lines when AUTH and/or TLS are in use. This is the "received protocol"
2153 field. Exim used to use "asmtp" for authenticated SMTP, without any
2154 indication (in the protocol name) for TLS use. Now it follows the RFC and
2155 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2156 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2157 names appear in log lines as well as in Received: header lines.
2159 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2160 extensions, and comparison tests, and to fix some bugs.
2162 14. Changes to the "personal" filter test:
2164 (1) The test was buggy in that it was just doing the equivalent of
2165 "contains" tests on header lines. For example, if a user's address was
2166 anne@some.where, the "personal" test would incorrectly be true for
2168 To: susanne@some.where
2170 This test is now done by extracting each address from the header in turn,
2171 and checking the entire address. Other tests that are part of "personal"
2172 are now done using regular expressions (for example, to check local parts
2173 of addresses in From: header lines).
2175 (2) The list of non-personal local parts in From: addresses has been
2176 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2177 taken from the Sieve specification recommendations.
2179 (3) If the message contains any header line starting with "List-" it is
2180 treated as non-personal.
2182 (4) The test for "circular" in the Subject: header line has been removed
2183 because it now seems ill-conceived.
2185 15. Minor typos in src/EDITME comments corrected.
2187 16. Installed latest exipick from John Jetmore.
2189 17. If headers_add on a router specified a text string that was too long for
2190 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2191 of string_sprintf() is now avoided.
2193 18. $message_body_size was not set (it was always zero) when running the DATA
2194 ACL and the local_scan() function.
2196 19. For the "mail" command in an Exim filter, no default was being set for
2197 the once_repeat time, causing a random time value to be used if "once" was
2198 specified. (If the value happened to be <= 0, no repeat happened.) The
2199 default is now 0s, meaning "never repeat". The "vacation" command was OK
2200 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2201 (I found it when inspecting the code).
2203 20. There is now an overall timeout for performing a callout verification. It
2204 defaults to 4 times the callout timeout, which applies to individual SMTP
2205 commands during the callout. The overall timeout applies when there is more
2206 than one host that can be tried. The timeout is checked before trying the
2207 next host. This prevents very long delays if there are a large number of
2208 hosts and all are timing out (e.g. when the network connections are timing
2209 out). The value of the overall timeout can be changed by specifying an
2210 additional sub-option for "callout", called "maxwait". For example:
2212 verify = sender/callout=5s,maxwait=20s
2214 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2215 to the end before writing, but this should make it even safer).
2217 22. Exim was forgetting that it had advertised PIPELINING for the second and
2218 subsequent messages on an SMTP connection. It was also not resetting its
2219 memory on STARTTLS and an internal HELO.
2221 23. When Exim logs an SMTP synchronization error within a session, it now
2222 records whether PIPELINING has been advertised or not.
2224 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2225 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2226 rather than long int.
2228 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2230 26. Added the never_mail option to autoreply.
2236 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2237 crash if the getsockname() call failed; this can happen if a connection is
2238 closed very soon after it is established. The problem was simply in the
2239 order in which certain operations were done, causing Exim to try to write
2240 to the SMTP stream before it had set up the file descriptor. The bug has
2241 been fixed by making things happen in the correct order.
2247 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2248 before the connection was closed, thus losing the rejection response.
2250 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2251 some early Solaris releases, but causes trouble in current releases where
2252 socklen_t is defined.
2254 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2257 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2260 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2261 that releasing the top of it at the end releases what was used for sub-
2262 expansions (unless the block got too big). However, discard this block if
2263 the first thing is a variable or header, so that we can use its block when
2264 it is dynamic (useful for very large $message_headers, for example).
2266 6. Lookups now cache *every* query, not just the most recent. A new, separate
2267 store pool is used for this. It can be recovered when all lookup caches are
2268 flushed. Lookups now release memory at the end of their result strings.
2269 This has involved some general refactoring of the lookup sources.
2271 7. Some code has been added to the store_xxx() functions to reduce the amount
2272 of flapping under certain conditions.
2274 8. log_incoming_interface used to affect only the <= reception log lines. Now
2275 it causes the local interface and port to be added to several more SMTP log
2276 lines, for example "SMTP connection from", and rejection lines.
2278 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2280 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2282 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2283 could be overwritten at the end of the current message (or the start of a
2284 new message if it was set in a HELO ACL). The value is now preserved for
2285 the duration of the SMTP connection.
2287 12. If a transport had a headers_rewrite setting, and a matching header line
2288 contained an unqualified address, that address was qualified, even if it
2289 did not match any rewriting rules. The underlying bug was that the values
2290 of the flags that permit the existence of unqualified sender and recipient
2291 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2292 non-local messages, and by -bnq for local messages) were not being
2293 preserved with the message after it was received.
2295 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2296 "next input=" as part of the text comprising the host identity instead of
2297 the correct text. The code was using the same buffer for two different
2298 strings. However, depending on which order the printing function evaluated
2299 its arguments, the bug did not always show up. Under Linux, for example, my
2300 test suite worked just fine.
2302 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2303 doesn't work with some older versions of Perl. It has been changed to "my",
2304 which in any case is probably the better facility to use.
2306 15. A really picky compiler found some instances of statements for creating
2307 error messages that either had too many or two few arguments for the format
2310 16. The size of the buffer for calls to the DNS resolver has been increased
2311 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2312 for addresses that have a lot of PTR records. This alleviates a problem; it
2313 does not fully solve it.
2315 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2316 buffer now truncates the list and logs the incident, which is the same
2317 action as happens when Exim is looking up a host name and its aliases.
2318 Previously in this situation something unpredictable would happen;
2319 sometimes it was "internal error: store_reset failed".
2321 18. If a server dropped the connection unexpectedly when an Exim client was
2322 using GnuTLS and trying to read a response, the client delivery process
2323 crashed while trying to generate an error log message.
2325 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2326 string, for example:
2328 warn message = H1: something\nH2: something
2330 the text was added as a single header line from Exim's point of view
2331 though it ended up OK in the delivered message. However, searching for the
2332 second and subsequent header lines using $h_h2: did not work. This has been
2333 fixed. Similarly, if a system filter added multiple headers in this way,
2334 the routers could not see them.
2336 20. Expanded the error message when iplsearch is called with an invalid key to
2337 suggest using net-iplsearch in a host list.
2339 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2340 ACLs are no longer actually imposed (and a message to that effect is
2343 22. If a "gecos" field in a passwd entry contained escaped characters, in
2344 particular, if it contained a \" sequence, Exim got it wrong when building
2345 a From: or a Sender: header from that name. A second bug also caused
2346 incorrect handling when an unquoted " was present following a character
2347 that needed quoting.
2349 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2350 was not being matched caselessly.
2352 24. Arranged for all hyphens in the exim.8 source to be escaped with
2355 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2356 properly. Recipient callout cache records were still being keyed to include
2357 the sender, even when use_sender was set false. This led to far more
2358 callouts that were necessary. The sender is no longer included in the key
2359 when use_sender is false.
2361 26. Added "control = submission" modifier to ACLs.
2363 27. Added the ${base62d: operator to decode base 62 numbers.
2365 28. dnsdb lookups can now access SRV records.
2367 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2368 the configuration file.
2370 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2371 (-v) mode. This makes the output for a verbose queue run more intelligible.
2373 31. Added a use_postmaster feature to recipient callouts.
2375 32. Added the $body_zerocount variable, containing the number of binary zero
2376 bytes in the message body.
2378 33. The time of last modification of the "new" subdirectory is now used as the
2379 "mailbox time last read" when there is a quota error for a maildir
2382 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2384 35. Added +ignore_unknown as a special item in host lists.
2386 36. Code for decoding IPv6 addresses in host lists is now included, even if
2387 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2388 address was recognized as an IP address, but was then not correctly decoded
2389 into binary, causing unexpected and incorrect effects when compared with
2396 1. Very minor rewording of debugging text in manualroute to say "list of
2397 hosts" instead of "hostlist".
2399 2. If verify=header_syntax was set, and a header line with an unqualified
2400 address (no domain) and a large number of spaces between the end of the
2401 name and the colon was received, the reception process suffered a buffer
2402 overflow, and (when I tested it) crashed. This was caused by some obsolete
2403 code that should have been removed. The fix is to remove it!
2405 3. When running in the test harness, delay a bit after writing a bounce
2406 message to get a bit more predictability in the log output.
2408 4. Added a call to search_tidyup() just before forking a reception process. In
2409 theory, someone could use a lookup in the expansion of smtp_accept_max_
2410 per_host which, without the tidyup, could leave open a database connection.
2412 5. Added the variables $recipient_data and $sender_data which get set from a
2413 lookup success in an ACL "recipients" or "senders" condition, or a router
2414 "senders" option, similar to $domain_data and $local_part_data.
2416 6. Moved the writing of debug_print from before to after the "senders" test
2419 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2420 problems for message scanning, either using a data ACL, or using
2421 local_scan() because the Received: header was not generated till after they
2422 were called (in order to set the time as the time of reception completion).
2423 I have revised the way this works. The header is now generated after the
2424 body is received, but before the ACL or local_scan() are called. After they
2425 are run, the timestamp in the header is updated.
2431 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2432 before starting a queue runner without re-exec. This happened only when
2433 deliver_drop_privilege was set or when the Exim user was set to root. The
2434 effect of the bug was that timeouts during subsequent deliveries caused
2435 crashes instead of being properly handled. The handler is now left at its
2436 default (and expected) setting.
2438 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2439 message, again when deliver_drop_privilege is set or Exim is run as root.
2440 The bug described in (1) was not present in this case, but the tidying up
2441 of the other signals was missing. I have made the two cases consistent.
2443 3. The ignore_target_hosts setting on a manualroute router was being ignored
2444 for hosts that were looked up using the /MX notation.
2446 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2449 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2450 operated on the sender address. After changing the $sender_address to <>
2451 for the sender address verify, Exim was re-instated it as the original
2452 (before rewriting) address, but remembering that it had rewritten it, so it
2453 wasn't rewriting it again. This bug also had the effect of breaking the
2454 sender address verification caching when the sender address was rewritten.
2456 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2457 This has been changed so that if the ip literal address matches
2458 ignore_target_hosts, the router declines.
2460 7. Added expansion conditions match_domain, match_address, and match_local_
2461 part (NOT match_host).
2463 8. The placeholder for the Received: header didn't have a length field set.
2465 9. Added code to Exim itself and to exim_lock to test for a specific race
2466 condition that could lead to file corruption when using MBX delivery. The
2467 issue is with the lockfile that is created in /tmp. If this file is removed
2468 after a process has opened it but before that process has acquired a lock,
2469 there is the potential for a second process to recreate the file and also
2470 acquire a lock. This could lead to two Exim processes writing to the file
2471 at the same time. The added code performs the same test as UW imapd; it
2472 checks after acquiring the lock that its file descriptor still refers to
2473 the same named file.
2475 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2476 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2479 11. Added the smtp_active_hostname option. If used, this will typically be made
2480 to depend on the incoming interface address. Because $interface_address is
2481 not set up until the daemon has forked a reception process, error responses
2482 that can happen earlier (such as "too many connections") no longer contain
2485 12. If an expansion in a condition on a "warn" statement fails because a lookup
2486 defers, the "warn" statement is abandoned, and the next ACL statement is
2487 processed. Previously this caused the whole ACL to be aborted.
2489 13. Added the iplsearch lookup type.
2491 14. Added ident_timeout as a log selector.
2493 15. Added tls_certificate_verified as a log selector.
2495 16. Added a global option tls_require_ciphers (compare the smtp transport
2496 option of the same name). This controls incoming TLS connections.
2498 17. I finally figured out how to make tls_require_ciphers do a similar thing
2499 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2500 before starting the TLS session.
2502 18. Tabs are now shown as \t in -bP output.
2504 19. If the log selector return_path_on_delivery was set, Exim crashed when
2505 bouncing a message because it had too many Received: header lines.
2507 20. If two routers both had headers_remove settings, and the first one included
2508 a superfluous trailing colon, the final name in the first list and the
2509 first name in the second list were incorrectly joined into one item (with a
2510 colon in the middle).
2516 1. Added -C and -D options to the exinext utility, mainly to make it easier
2517 to include in the automated testing, but these could be helpful when
2518 multiple configurations are in use.
2520 2. The exinext utility was not formatting the output nicely when there was
2521 an alternate port involved in the retry record key, nor when there was a
2522 message id as well (for retries that were specific to a specific message
2523 and a specific host). It was also confused by IPv6 addresses, because of
2524 the additional colons they contain. I have fixed the IPv4 problem, and
2525 patched it up to do a reasonable job for IPv6.
2527 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2528 delivery, the log line now contains "pipelined" if PIPELINING was used.
2530 4. An SMTP transport process used to panic and die if the bind() call to set
2531 an explicit outgoing interface failed. This has been changed; it is now
2532 treated in the same way as a connect() failure.
2534 5. A reference to $sender_host_name in the part of a conditional expansion
2535 that was being skipped was still causing a DNS lookup. This no longer
2538 6. The def: expansion condition was not recognizing references to header lines
2539 that used bh_ and bheader_.
2541 7. Added the _cache feature to named lists.
2543 8. The code for checking quota_filecount in the appendfile transport was
2544 allowing one more file than it should have been.
2546 9. For compatibility with Sendmail, the command line option
2554 and sets the incoming protocol and host name (for trusted callers). The
2555 host name and its colon can be omitted when only the protocol is to be set.
2556 Note the Exim already has two private options, -pd and -ps, that refer to
2557 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2558 "s", but I don't think that's a major issue.
2560 10. A number of refactoring changes to the code, none of which should affect
2563 (a) The number of logging options was getting close to filling up the
2564 32-bit word that was used as a bit map. I have split them into two classes:
2565 those that are passed in the argument to log_write(), and those that are
2566 only ever tested independently outside of that function. These are now in
2567 separate 32-bit words, so there is plenty of room for expansion again.
2568 There is no change in the user interface or the logging behaviour.
2570 (b) When building, for example, log lines, the code previously used a
2571 macro that called string_cat() twice, in order to add two strings. This is
2572 not really sufficiently general. Furthermore, there was one instance where
2573 it was actually wrong because one of the argument was used twice, and in
2574 one call a function was used. (As it happened, calling the function twice
2575 did not affect the overall behaviour.) The macro has been replaced by a
2576 function that can join an arbitrary number of extra strings onto a growing
2579 (c) The code for expansion conditions now uses a table and a binary chop
2580 instead of a serial search (which was left over from when there were very
2581 few conditions). Also, it now recognizes conditions like "pam" even when
2582 the relevant support is not compiled in: a suitably worded error message is
2583 given if an attempt is made to use such a condition.
2585 11. Added ${time_interval:xxxxx}.
2587 12. A bug was causing one of the ddress fields not to be passed back correctly
2588 from remote delivery subprocesses. The field in question was not being
2589 subsequently used, so this caused to problems in practice.
2591 13. Added new log selectors queue_time and deliver_time.
2593 14. Might have fixed a bug in maildirsizefile handling that threw up
2594 "unexpected character" debug warnings, and recalculated the data
2595 unnecessarily. In any case, I expanded the warning message to give more
2598 15. Added the message "Restricted characters in address" to the statements in
2599 the default ACL that block characters like @ and % in local parts.
2601 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2602 Three changes have been made:
2604 (a) There was a serious bug; a negative response to MAIL caused the whole
2605 recipient domain to be cached as invalid, thereby blocking all messages
2606 to all local parts at the same domain, from all senders. This bug has
2607 been fixed. The domain is no longer cached after a negative response to
2608 MAIL if the sender used is not empty.
2610 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2613 (c) A new callout option, "use_sender" has been added for people who want
2614 the modified behaviour.
2620 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2623 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2624 indeed breaks things for older releases.
2626 3. Added additional logging to the case where there is a problem reading data
2627 from a filter that is running in a subprocess using a pipe, in order to
2628 try to track down a specific problem.
2630 4. Testing facility fudge: when running in the test harness and attempting
2631 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2632 getting "No route to host". Convert this to a timeout.
2634 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2637 6. Some OS don't have socklen_t but use size_t instead. This affects the
2638 fifth argument of getsockopt() amongst other things. This is now
2639 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2640 can be set for individual OS. I have set it for SunOS5, OSF1, and
2641 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2642 some earlier ones do not.
2644 7. Change 4.30/15 was not doing the test caselessly.
2646 8. The standard form for an IPv6 address literal was being rejected by address
2647 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2648 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2649 this, as well as the form without the "IPv6" on the front (but only when
2650 address literals are enabled, of course).
2652 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2654 10. Exim crashed if a message with an empty sender address specified by -f
2655 encountered a router with an errors_to setting. This could be provoked only
2656 by a command such as
2660 where an empty string was supplied; "<>" did not hit this bug.
2662 11. Installed PCRE release 4.5.
2664 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2665 remained set. It is now erased.
2667 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2668 times from message ids (which are base 36 rather than the normal 62).
2670 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2671 were being counted as actual protocol errors, and logged if the log
2672 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2673 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2674 and DATA following a set of rejected RCPTs do not count as protocol errors.
2675 In other words, Exim assumes they were pipelined, though this may not
2676 actually be the case. Of course, in all cases the client gets an
2677 appropriate error code.
2679 15. If a lookup fails in an ACL condition, a message about the failure may
2680 be available; it is used if testing the ACL cannot continue, because most
2681 such messages specify what the cause of the deferral is. However, some
2682 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2683 that caused an old message to be retained and used if a later statement
2684 caused a defer, replacing the real cause of the deferral.
2686 16. If an IP address had so many PTR records that the DNS lookup buffer
2687 was not large enough to hold them, Exim could crash while trying to process
2688 the truncated data. It now detects and logs this case.
2690 17. Further to 4.21/58, another change has been made: if (and only if) the
2691 first line of a message (the first header line) ends with CRLF, a bare LF
2692 in a subsequent header line has a space inserted after it, so as not to
2693 terminate the header.
2695 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2696 unnecessarily, used atoi() instead of strtol(), and didn't check the
2697 termination when getting file sizes from file names by regex.
2699 19. Completely re-implemented the support for maildirsize files, in the light
2700 of a number of problems with the previous contributed implementation
2701 (4.30/29). In particular:
2703 . If the quota is zero, the maildirsize file is maintained, but no quota is
2706 . If the maildir directory does not exist, it is created before any attempt
2707 to write a maildirsize file.
2709 . The quota value in the file is just a cache; if the quota is changed in
2710 the transport, the new value overrides.
2712 . A regular expression is available for excluding directories from the
2715 20. The autoreply transport checks the characters in options that define the
2716 message's headers; it allows continued headers, but it was checking with
2717 isspace() after an embedded newline instead of explicitly looking for a
2720 21. If all the "regular" hosts to which an address was routed had passed their
2721 expiry times, and had not reached their retry times, the address was
2722 bounced, even if fallback hosts were defined. Now Exim should go on to try
2725 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2726 equivalent code in the SMTP transport. Some hosts send humungous responses
2727 to HELO/EHLO, more than 1024 it seems.
2729 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2730 gives compiler warnings in some environments. I've now done it "properly",
2733 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2734 (because of problems with the built-in one) was declared to return uschar *
2735 instead of char *, causing compiler failure.
2737 25. Fixed a file descriptor leak when processing alias/forward files.
2739 26. Fixed a minor format string issue in dbfn.c.
2741 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2743 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2744 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2747 29. When a sender address is verified, it is cached, to save repeating the test
2748 when there is more than one recipient in a message. However, when the
2749 verification involves a callout, it is possible for different callout
2750 options to be set for different recipients. It is too complicated to keep
2751 track of this in the cache, so now Exim always runs a verification when a
2752 callout is required, relying on the callout cache for the optimization.
2753 The overhead is duplication of the address routing, but this should not be
2756 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2757 to be verified with callout=postmaster, and the main callout worked but the
2758 postmaster check failed, the verification correctly failed. However, if a
2759 subsequent RCPT command asked for sender verification *without* the
2760 postmaster check, incorrect caching caused this verification also to fail,
2763 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2764 it was not caching the DNS options (qualify_single, search_parents) that
2765 were used when the lookup failed. A subsequent lookup with different
2766 options therefore always gave the same answer, though there were cases
2767 where it should not have. (Example: a "domains = !$mx_any" option on a
2768 dnslookup router: the "domains" option is always processed without any
2769 widening, but the router might have qualify_single set.) Now Exim uses the
2770 cached value only when the same options are set.
2772 32. Added John Jetmore's "exipick" utility to the distribution.
2774 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2775 than a timeout (e.g. a certificate is required, and is not provided), an
2776 Exim server now closes the connection immediately. Previously it waited for
2777 the client to close - but if the client is SSL, it seems that they each
2778 wait for each other, leading to a delay before one of them times out.
2780 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2781 maintained 0.8.x compatibility because I don't think many are using it, and
2782 it is clearly obsolete.
2784 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2787 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2788 client certificate was expired. A simple patch fixes this, though I don't
2789 understand the full logic of why the verify callback is called multiple
2792 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2793 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2794 which causes problems with some clients (such as the Certicom SSL Plus
2795 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2796 disables the coutermeasure allowing Eudora to connect."
2798 38. Exim was not checking that a write() to a log file succeeded. This could
2799 lead to Bad Things if a log got too big, in particular if it hit a file
2800 size limit. Exim now panics and dies if it cannot write to a log file, just
2801 as it does if it cannot open a log file.
2803 39. Modified OS/Makefile-Linux so that it now contains
2805 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2807 The two -D definitions ensure that Exim is compiled with large file
2808 support, which makes it possible to handle log files that are bigger than
2811 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2812 instance) a domain was checked against a named list that involved a lookup,
2813 causing $domain_data to be set, then another domain was checked against the
2814 same list, then the first domain was re-checked, the value of $domain_data
2815 after the final check could be wrong. In particular, if the second check
2816 failed, it could be set empty. This bug probably also applied to
2819 41. The strip_trailing_dot option was not being applied to the address given
2820 with the -f command-line option.
2822 42. The code for reading a message's header from the spool was incrementing
2823 $received_count, but never initializing it. This meant that the value was
2824 incorrect (doubled) while delivering a message in the same process in which
2825 it was received. In the most common configuration of Exim, this never
2826 happens - a fresh exec is done - but it can happen when
2827 deliver_drop_privilege is set.
2829 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2830 it now includes up to 150 characters of the unexpected data in the log
2833 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2834 and building data strings. The size of both of these buffers was 10 000
2835 bytes - far larger than anybody would *ever* want, thought I. Needless to
2836 say, somebody hit the limit. I have increased the maximum line length to
2837 20 000 and the maximum data length of concatenated lines to 100 000. I have
2838 also fixed two bugs, because there was no checking on these buffers. Tsk,
2839 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2840 buffer is too small.
2842 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2843 lsearch lookups. Now it does.
2845 46. When parsing a route_list item in a manualroute router, a fixed-length
2846 buffer was used for the list of hosts. I made this 1024 bytes long,
2847 thinking that nobody would ever have a list of hosts that long. Wrong.
2848 Somebody had a whole pile of complicated expansion conditions, and the
2849 string was silently truncated, leading to an expansion error. It turns out
2850 that it is easier to change to an unlimited length (owing to other changes
2851 that have happened since this code was originally written) than to build
2852 structure for giving a limitation error. The length of the item that
2853 expands into the list of hosts is now unlimited.
2855 47. The lsearch lookup could not handle data where the length of text line was
2856 more than 4095 characters. Such lines were truncated, leading to shortened
2857 data being returned. It should now handle lines of any length.
2859 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2860 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2863 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2864 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2865 no longer happens in comment lines. A list of replacements is now placed
2866 at the head of all of the source files, except those whose only change is
2867 to replace PERL_COMMAND in the very first #! line.
2869 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2870 messages on the queue, with a bottom-up merge sort, using code contributed
2871 by Michael Haardt. This should make operations like -bp somewhat faster on
2872 large queues. It won't affect queue runners, except when queue_run_in_order
2875 51. Installed eximstats 1.31 in the distribution.
2877 52. Added support for SRV lookups to the dnslookup router.
2879 53. If an ACL referred to $message_body or $message_body_end, the value was not
2880 reset for any messages that followed in the same SMTP session.
2882 54. The store-handling optimization for building very long strings was not
2883 differentiating between the different store pools. I don't think this
2884 actually made any difference in practice, but I've tidied it.
2886 55. While running the routers to verify a sender address, $sender_address
2887 was still set to the sender address. This is wrong, because when routing to
2888 send a bounce to the sender, it would be empty. Therefore, I have changed
2889 it so that, while verifying a sender address, $sender_address is set to <>.
2890 (There is no change to what happens when verifying a recipient address.)
2892 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2893 target A or AAAA records (if not already returned) without resetting the
2894 qualify_single or search_parents options of the DNS resolver. These are
2895 inappropriate in this case because the targets of MX and SRV records must
2896 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2897 target that, when qualified, matched something in the local domain. These
2898 two options are now turned off when doing these lookups.
2900 57. It seems that at least some releases of Reiserfs (which does not have the
2901 concept of a fixed number of inodes) returns zero and not -1 for the
2902 number of available inodes. This interacted badly with check_spool_inodes,
2903 which assumed that -1 was the "no such thing" setting. What I have done is
2904 to check that the total number of inodes is greater than zero before doing
2905 the test of how many are available.
2907 58. When a "warn" ACL statement has a log_message modifier, the message is
2908 remembered, and not repeated. This is to avoid a lot of repetition when a
2909 message has many recipients that cause the same warning to be written.
2910 Howewer, Exim was preserving the list of already written lines for an
2911 entire SMTP session, which doesn't seem right. The memory is now reset if a
2912 new message is started.
2914 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2915 debugging output unless log_rewrite was also set.
2917 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2918 of (int)(handle) when we know that handle contains (void *)(-1).
2920 61. The Exim daemon panic-logs an error return when it closes the incoming
2921 connection. However "connection reset by peer" seems to be common, and
2922 isn't really an error worthy of noting specially, so that particular error
2925 62. When Exim is trying to find all the local interfaces, it used to panic and
2926 die if the ioctl to get the interface flags failed. However, it seems that
2927 on at least one OS (Solaris 9) it is possible to have an interface that is
2928 included in the list of interfaces, but for which you get a failure error
2929 for this call. This happens when the interface is not "plumbed" into a
2930 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2931 failure of the "get flags" call assumes that the interface is down.
2933 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2934 makes life easier for people who are doing arithmetic on fields extracted
2935 from dates, where you often get leading zeros that should not be
2936 interpreted as octal.
2938 64. Added qualify_domain to the redirect router, to override the global
2941 65. If a pathologically long header line contained very many addresses (the
2942 report of this problem mentioned 10 000) and each of them was rewritten,
2943 Exim could use up a very large amount of memory. (It kept on making new
2944 copies of the header line as it rewrote, and never released the old ones.)
2945 At the expense of a bit more processing, the header rewriting function has
2946 been changed so that it no longer eats memory in this way.
2948 66. The generation of the Received: header has been moved from the time that a
2949 message starts to be received, to the time that it finishes. The timestamp
2950 in the Received: header should now be very close to that of the <= log
2951 line. There are two side-effects of this change:
2953 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2954 logged header lines no longer include the local Received: line, because
2955 it has not yet been created. The same applies to a copy of the message
2956 that is returned to a non-SMTP sender when a message is rejected.
2958 (b) When a filter file is tested using -bf, no additional Received: header
2959 is added to the test message. After some thought, I decided that this
2962 This change does not affect the value of $received_for. It is still set
2963 after address rewriting, but before local_scan() is called.
2965 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2967 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2968 gave an unhelpful panic error message, and a defer error. I have managed to
2969 change this behaviour so that it now rejects any supplied certificate,
2970 which seems right, as the list of acceptable certificates is empty.
2972 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2973 gave an unhelpful defer error. I have not managed to make this reject any
2974 supplied certificates, but the error message it gives is "no certificate
2975 supplied", which is not helpful.
2977 70. exigrep's output now also includes lines that are not associated with any
2978 message, but which match the given pattern. Implemented by a patch from
2979 Martin Sluka, which also tidied up the Perl a bit.
2981 71. Recipient callout verification, like sender verification, was using <> in
2982 the MAIL FROM command. This isn't really the right thing, since the actual
2983 sender may affect whether the remote host accepts the recipient or not. I
2984 have changed it to use the actual sender in the callout; this means that
2985 the cache record is now keyed on a recipient/sender pair, not just the
2986 recipient address. There doesn't seem to be a real danger of callout loops,
2987 since a callout by the remote host to check the sender would use <>.
2988 [SEE ABOVE: changed after hitting problems.]
2990 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2991 temporary errors. However, in the case of such a code being given after
2992 the end of a data transmission (i.e. after ".") Exim was failing to write
2993 a retry record for the message. (Yes, there was some broken host that was
2994 actually sending 8xx at this point.)
2996 73. An unknown lookup type in a host list could cause Exim to panic-die when
2997 the list was checked. (An example that provoked this was putting <; in the
2998 middle of a list instead of at the start.) If this happened during a DATA
2999 ACL check, a -D file could be left lying around. This kind of configuration
3000 error no longer causes Exim to die; instead it causes a defer errror. The
3001 incident is still logged to the main and panic logs.
3003 74. Buglet left over from Exim 3 conversion. The message "too many messages
3004 in one connection" was written to the rejectlog but not the mainlog, except
3005 when address rewriting (yes!) was being logged.
3007 75. Added write_rejectlog option.
3009 76. When a system filter was run not as root (that is, when system_filter_user
3010 was set), the values of the $n variables were not being returned to the
3011 main process; thus, they were not subsequently available in the $sn
3014 77. Added +return_path_on_delivery log selector.
3016 78. A connection timeout was being treated differently from recipients deferred
3017 when testing hosts_max_try with a message that was older than the host's
3018 retry timeout. (The host should not be counted, thus allowing all hosts to
3019 be tried at least once before bouncing.) This may have been the cause of an
3020 occasionally reported bug whereby a message would remain on the queue
3021 longer than the retry timeout, but would be bounced if a delivery was
3022 forced. I say "may" because I never totally pinned down the problem;
3023 setting up timeout/retry tests is difficult. See also the next item.
3025 79. The ultimate address timeout was not being applied to errors that involved
3026 a combination of host plus message (for example, a timeout on a MAIL
3027 command). When an address resolved to a number of possible hosts, and they
3028 were not all tried for each delivery (e.g. because of hosts_max_try), a
3029 message could remain on the queue longer than the retry timeout.
3031 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3034 81. Fixed an obscure SMTP outgoing bug which required at least the following
3035 conditions: (a) there was another message waiting for the same server;
3036 (b) the server returned 5xx to all RCPT commands in the first message so
3037 that the message was not completed; (c) the server dropped the connection
3038 or gave a negative response to the RSET that Exim sends to abort the
3039 transaction. The observed case was a dropped connection after DATA that had
3040 been sent in pipelining mode. That is, the server had advertised PIPELINING
3041 but was not implementing it correctly. The effect of the bug was incorrect
3042 behaviour, such as trying another host, and this could lead to a crash.
3048 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3049 and daemon.c were passed as pointers to ints; they should have been
3050 pointers to socklen_t variables (which are typically unsigned ints).
3052 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3055 3. Fixed a really odd bug that affected only the testing scheme; patching a
3056 certain fixed string in the binary changed the value of another string that
3057 happened to be identical to the end of the original first string.
3059 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3060 name", it returns that address as the IP address. On some operating
3061 systems (e.g. Solaris), it also passes back the IP address string as the
3062 "host name". However, on others (e.g. Linux), it passes back an empty
3063 string. Exim wasn't checking for this, and was changing the host name to an
3064 empty string, assuming it had been canonicized.
3066 5. Although rare, it is permitted to have more than one PTR record for a given
3067 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3068 all the names associated with an address, because they do in Solaris.
3069 However, it seems that they do not in Linux for data that comes from the
3070 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3071 I found this out when I moved to a new Linux workstation and tried to run
3072 the Exim test suite.
3074 To get round this problem I have changed the code so that it now does its
3075 own call to the DNS to look up PTR records when searching for a host name.
3076 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3077 addresses that are only in /etc/hosts are still found.
3079 This behaviour is, however, controlled by an option called host_lookup_
3080 order, which defaults to "bydns:byaddr". If people want to use the other
3081 order, or indeed, just use one or the other means of lookup, they can
3082 specify it in this variable.
3084 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3085 some operating systems, this comes back from gethostbyaddr() as an empty
3086 string, and this is what Exim used to test for. However, it seems that in
3087 other systems, "." is yielded. Exim now tests for this case too.
3089 7. The values of check_spool_space and check_log_space are now held internally
3090 as a number of kilobytes instead of an absolute number of bytes. If a
3091 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3092 kilobyte. This means that much larger values can be stored.
3094 8. Exim monitor: an attempt to get the action menu when not actually pointing
3095 at a message produces an empty menu entitled "No message selected". This
3096 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3097 no entries in it ("Shell widget menu has zero width and/or height"). So I
3098 have added a single, blank menu entry in this case.
3100 9. Added ${quote_local_part.
3102 10. MIME decoding is now applied to the contents of Subject: header lines when
3105 11. Now that a reference to $sender_host_address automatically causes a reverse
3106 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3107 host lookup before query-style lookups in lists that might use this
3108 variable. This has therefore been abolished, and the "net-" prefix is no
3109 longer necessary for query-style lookups.
3111 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3112 have been a typo for LFLAGS, so it has been changed.
3114 13. The install script calls Exim with "-C /dev/null" in order to find the
3115 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3116 to be output. Howeve, since Exim outputs its version number before the
3117 error, it didn't break the script. It just looked ugly. I fixed this by
3118 always allowing "-C /dev/null" if the caller is root.
3120 14. Ignore overlarge ACL variable number when reading spool file - insurance
3121 against a later release with more variables having written the file.
3123 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3124 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3125 this, as well as the form without the "IPv6" on the front.
3127 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3128 OS/Makefile-Darwin file.
3130 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3131 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3133 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3134 variable. However, some OS use ints for inodes. I've added cast to long int
3135 to get rid of the compiler warning.
3137 19. I had forgotten to lock out "/../" in configuration file names when
3138 ALT_CONFIG_PREFIX was set.
3140 20. Routers used for verification do not need to specify transports. However,
3141 if such a router generated a host list, and callout was configured, Exim
3142 crashed, because it could not find a port number from the (non-existent)
3143 transport. It now assumes port 25 in this circumstance.
3145 21. Added the -t option to exigrep.
3147 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3148 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3149 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3150 them is set, LOOKUP_LSEARCH is forced.
3152 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3154 24. Added sender and host information to the "rejected by local_scan()" log
3155 line; previously there was no indication of these.
3157 25. Added .include_if_exists.
3159 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3160 Linux. It turns out that not all file systems support this. Apparently some
3161 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3162 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3163 not supported on the file descriptor, is now ignored when Exim is trying to
3164 sync a directory. This applies only to Linux.
3166 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3168 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3169 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3172 29. Added support for maildirsize files from supplied patch (modified a bit).
3174 30. The use of :fail: followed by an empty string could lead Exim to respond to
3175 sender verification failures with (e.g.):
3177 550 Verification failed for <xxx>
3178 550 Sender verify failed
3180 where the first response line was missing the '-' that indicates it is not
3181 the final line of the response.
3183 31. The loop for finding the name of the user that called Exim had a hardwired
3184 limit of 10; it now uses the value of finduser_retries, which is used for
3185 all other user lookups.
3187 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3190 33. Exim was neglecting to zero errno before one call of strtol() when
3191 expanding a string and expecting an integer value. On some systems this
3192 resulted in spurious "integer overflow" errors. Also, it was casting the
3193 result into an int without checking.
3195 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3196 did not work. The code looks as if it has *never* worked, though it appears
3197 to have been documented since at least releast 1.62. I have made it work.
3199 35. The "timeout_DNS" error in retry rules, also documented since at least
3200 1.62, also never worked. As it isn't clear exactly what this means, and
3201 clearly it isn't a major issue, I have abolished the feature by treating it
3202 as "timeout", and writing a warning to the main and panic logs.
3204 36. The display of retry rules for -brt wasn't always showing the error code
3207 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3208 timeout_connect_A, timeout_connect_MX.
3210 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3211 to the empty sender.
3213 39. The daemon was not analysing the content of -oX till after it had closed
3214 stderr and disconnected from the controlling terminal. This meant that any
3215 syntax errors were only noted on the panic log, and the return code from
3216 the command was 0. By re-arranging the code a little, I've made the
3217 decoding happen first, so such errors now appear on stderr, and the return
3218 code is 1. However, the actual setting up of the sockets still happens in
3219 the disconnected process, so errors there are still only recorded on the
3222 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3223 connections (as happens on some IP stacks) was logged at start up time as
3224 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3225 it from "IPv6 and IPv4", which means that two separate sockets are being
3228 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3229 says whether AF_INET or AF_INET6 was passed as an argument.
3231 42. Exiwhat output was messed up when time zones were included in log
3234 43. Exiwhat now gives more information about the daemon's listening ports,
3235 and whether -tls-on-connect was used.
3237 44. The "port" option of the smtp transport is now expanded.
3239 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3240 silently ignored. Now an error message is written to the main and panic
3243 46. There's a new ACL modifier called "logwrite" which writes to a log file
3244 as soon as it is encountered.
3246 47. Added $local_user_uid and $local_user_gid at routing time.
3248 48. Exim crashed when trying to verify a sender address that was being
3251 49. Exim was recognizing only a space character after ".include". It now also
3252 recognizes a tab character.
3254 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3255 extracting the relevant information from the specification. The man page no
3256 longer contains scrambled data for the -d option, and I've added a section
3257 at the front about calling Exim under different names.
3259 51. Added "extra_headers" argument to the "mail" command in filter files.
3261 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3264 53. Installed eximstats 1.29.
3266 54. Added transport_filter_timeout as a generic transport option.
3268 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3269 Cc: header lines. This was required by RFC 822, but it not required by RFC
3272 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3273 incoming messages that did not have them. Now it does so only if the
3274 message originates locally, that is, if there is no associated remote host
3275 address. When Resent- header lines are present, this applies to the Resent-
3276 lines rather than the non-Resent- lines.
3278 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3279 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3281 58. Messages for configuration errors now include the name of the main
3282 configuration file - useful now that there may be more than one file in a
3283 list (.included file names were always shown).
3285 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3286 for those rare installations that do not start the daemon as root or run it
3287 setuid root. I've cut out the call to initgroups() if the daemon is not
3290 60. The Exim user and group can now be bound into the binary as text strings
3291 that are looked up at the start of Exim's processing.
3293 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3295 62. Added $mailstore_basename variable.
3297 63. Installed patch to sieve.c from Michael Haardt.
3299 64. When Exim failed to open the panic log after failing to open the main log,
3300 the original message it was trying to log was written to stderr and debug
3301 output, but if they were not available (the usual case in production), it
3302 was lost. Now it is written to syslog before the two lines that record the
3303 failures to open the logs.
3305 65. Users' Exim filters run in subprocesses under the user's uid. It is
3306 possible for a "deliver" command or an alias in a "personal" command to
3307 provoke an address rewrite. If logging of address rewriting is configured,
3308 this fails because the process is not running as root or exim. There may be
3309 a better way of dealing with this, but for the moment (because 4.30 needs
3310 to be released), I have disabled address rewrite logging when running a
3311 filter in a non-root, non-exim process.
3317 1. The buildconfig auxiliary program wasn't quoting the value set for
3318 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3319 not defined. This bug was masked in 4.22 by the effect that was fixed in
3322 2. Some messages that were rejected after a message id was allocated were
3323 shown as "incomplete" by exigrep. It no longer does this for messages that
3324 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3326 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3327 have allow_domain_literals set, the ID did not get logged in the <= line.
3328 Domain literals are now always recognized in Message-ID: header lines.
3330 4. The first argument for a ${extract expansion item is the key name or field
3331 number. Leading and trailing spaces in this item were not being ignored,
3332 causing some misleading effects.
3334 5. When deliver_drop_privilege was set, single queue runner processes started
3335 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3336 same command in the process it spins off) were not dropping privilege.
3338 6. When the daemon running as "exim" started a queue runner, it always
3339 re-executed Exim in the spun-off process. This is a waste of effort when
3340 deliver_drop_privilege is set. The new process now just calls the
3341 queue-runner function directly.
3347 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3350 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3351 ignored. Though the use of -oP was forcing the writing of a pid file, it
3352 was always written to the default place.
3354 3. If the message "no IP address found for host xxxx" is generated during
3355 incoming verification, it is now followed by identification of the incoming
3356 connection (so you can more easily find what provoked it).
3358 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3360 5. Added some features to "harden" Exim a bit more against certain attacks:
3362 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3363 be put in Local/Makefile. This is like the never_users runtime option,
3364 but it cannot be overridden. The default setting is "root".
3366 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3367 prefix string with which any file named in a -C command line option
3370 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3371 is retained for -C and -D only if the caller of Exim is root. Without
3372 it, the exim user may also use -C and -D and retain privilege.
3374 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3375 command line option is disabled.
3377 6. Macro names set by the -D option must start with an upper case letter, just
3378 like macro names defined in the configuration file.
3380 7. Added "dereference=" facility to LDAP.
3382 8. Two instances of the typo "uknown" in the source files are fixed.
3384 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3385 the Configure-Makefile script screwed up while processing it.
3387 10. Incorporated PCRE 4.4.
3389 11. The SMTP synchronization check was not operating right at the start of an
3390 SMTP session. For example, it could not catch a HELO sent before the client
3391 waited for the greeting. There is now a check for outstanding input at the
3392 point when the greeting is written. Because of the duplex, asynchronous
3393 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3394 way, but not yet received, when the check is performed.
3396 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3397 on TCP/IP sockets, because this apparently causes some broken clients to
3400 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3401 unchanged) from the Cygwin maintainer.
3403 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3404 was supported instead of testing for mbx. Effectively a typo.
3406 15. The spa authenticator server code was not checking that the input it
3407 received was valid base64.
3409 16. The debug output line for the "set" modifier in ACLs was not showing the
3410 name of the variable that was being set.
3412 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3414 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3415 lookup on its own. Something else was needed to trigger the lookup. For
3416 example, a match in host_lookup or the need for a host name in a host list.
3417 Now, if $sender_host_name is referenced and the host name has not yet been
3418 looked up, a lookup is performed. If the lookup fails, the variable remains
3419 empty, and $host_lookup_failed is set to "1".
3421 19. Added "eqi" as a case-independent comparison operator.
3423 20. The saslauthd authentication condition could segfault if neither service
3424 nor realm was specified.
3426 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3427 error message that was logged was misleading, and incoming SMTP
3428 connections were dropped. The message is now more accurate, and temporary
3429 errors are given to SMTP connections.
3431 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3432 (except RSET) with a 421 error, until QUIT is received. However, it was
3433 failing to send a response to QUIT.
3435 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3436 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3437 ACL is now run after the helo_try_verify_hosts code.
3439 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3440 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3441 case-independent, so other case variants are also recognized). Apparently
3442 some systems use these upper case variants.
3444 25. If more than two messages were waiting for the same host, and a transport
3445 filter was specified for the transport, Exim sent two messages over the
3446 same TCP/IP connection, and then failed with "socket operation on non-
3447 socket" when it tried to send the third.
3449 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3451 27. The extern definition of crypt16() in expand.c was not being excluded when
3452 the OS had its own crypt16() function.
3454 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3455 as a preferred synonym for return_size_limit, both as an option and as an
3458 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3460 30. Changed the default configuration ACL to relax the local part checking rule
3461 for addresses that are not in any local domains. For these addresses,
3462 slashes and pipe symbols are allowed within local parts, but the sequence
3463 /../ is explicitly forbidden.
3465 31. SPA server authentication was not clearing the challenge buffer before
3468 32. log_message in a "warn" ACL statement was writing to the reject log as
3469 well as to the main log, which contradicts the documentation and doesn't
3470 seem right (because no rejection is happening). So I have stopped it.
3472 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3473 However, I am unable to do any testing of this.
3475 34. Fixed an infelicity in the appendfile transport. When checking directories
3476 for a mailbox, to see if any needed to be created, it was accidentally
3477 using path names with one or more superfluous leading slashes; tracing
3478 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3480 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3481 contents are added to the log line that is written for every discarded
3482 recipient. (Previously a log_message setting was ignored.)
3484 36. The ${quote: operator now quotes the string if it is empty.
3486 37. The install script runs exim in order to find its version number. If for
3487 some reason other than non-existence or emptiness, which it checks, it
3488 could not run './exim', it was installing it with an empty version number,
3489 i.e. as "exim-". This error state is now caught, and the installation is
3492 38. An argument was missing from the function that creates an error message
3493 when Exim fails to connect to the socket for saslauthd authentication.
3494 This could cause Exim to crash, or give a corrupted message.
3496 39. Added isip, isip4, and isip6 to ${if conditions.
3498 40. The ACL variables $acl_xx are now saved with the message, and can be
3499 accessed later in routers, transports, and filters.
3501 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3502 strings in the file are not string-expanded.
3504 42. If a MAIL command specified a SIZE value that was too large to fit into an
3505 int variable, the check against message_size_limit failed. Such values are
3506 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3507 day this will have to be increased, but I don't think I want to be around
3508 when emails are that large.
3515 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3516 iconv() is not standard in FreeBSD.
3518 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3519 IPv6 enabled. The observed symptom was a segmentation fault on return from
3520 the function os_common_find_running_interfaces() in src/os.c.
3522 3. In the check_special_case() function in daemon.c I had used "errno" as an
3523 argument name, which causes warnings on some systems. This was basically a
3524 typo, since it was named "eno" in the comments!
3526 4. The code that waits for the clock to tick (at a resolution of some fraction
3527 of a second) so as to ensure message-id uniqueness was always waiting for
3528 at least one whole tick, when it could have waited for less. [This is
3529 almost certainly not relevant at current processor speeds, where it is
3530 unlikely to ever wait at all. But we try to future-proof.]
3532 5. The function that sleeps for a time interval that includes fractions of a
3533 second contained a race. It did not block SIGALRM between setting the
3534 timer, and suspending (a couple of lines later). If the interval was short
3535 and the sigsuspend() was delayed until after it had expired, the suspension
3536 never ended. On busy systems this could lead to processes getting stuck for
3539 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3540 process, before it forks any delivery processes. The open lookup caching
3541 mechanism meant that the open file or database connection was passed into
3542 the delivery process. The problem was that delivery processes always tidy
3543 up cached lookup data. This could cause a problem for the next delivery
3544 process started by the queue runner, because the external queue runner
3545 process does not know about the closure. So the next delivery process
3546 still has data in the lookup cache. In the case of a file lookup, there was
3547 no problem because closing a file descriptor in a subprocess doesn't affect
3548 the parent. However, if the lookup was caching a connection to a database,
3549 the connection was closed, and the second delivery process was likely to
3550 see errors such as "PGSQL: query failed: server closed the connection
3551 unexpectedly". The problem has been fixed by closing all cached lookups
3552 in a queue runner before running a delivery process.
3554 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3555 seem to have the "const" qualifier which it has on other OS. I've
3558 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3559 *of the same type* (client or server) with the same public name that an
3560 error should be diagnosed.
3562 9. When Exim looked up a host name for an IP address, but failed to find the
3563 original IP address when looking up the host name (a safety check), it
3564 output the message "<ip address> does not match any IP for NULL", which was
3565 confusing, to say the least. The bug was that the host name should have
3566 appeared instead of "NULL".
3568 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3569 user that is built into the binary, and the -C or -D options is used, root
3570 privilege is dropped before the configuration file is read. In addition,
3571 logging is switched to stderr instead of the normal log files. If the
3572 configuration then re-defines the Exim user, the unprivileged environment
3573 is probably not what is expected, so Exim logs a panic warning message (but
3576 However, if deliver_drop_privilege is set, the unprivileged state may well
3577 be exactly what is intended, so the warning has been cut out in that case,
3578 and Exim is allowed to try to write to its normal log files.
3584 1. smtp_return_error_details was not giving details for temporary sender
3585 or receiver verification errors.
3587 2. Diagnose a configuration error if two authenticators have the same public
3590 3. Exim used not to create the message log file for a message until the first
3591 delivery attempt. This could be confusing when incoming messages were held
3592 for policy or load reasons. The message log file is now created at the time
3593 the message is received, and an initial "Received" line is written to it.
3595 4. The automatically generated man page for command line options had a minor
3596 bug that caused no ill effects; however, a more serious problem was that
3597 the procedure for building the man page automatically didn't always
3598 operate. Consequently, release 4.20 contains an out-of-date version. This
3599 shouldn't happen again.
3601 5. When building Exim with embedded Perl support, the script that builds the
3602 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3603 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3605 6. The freeze_tell option was not being used for messages that were frozen on
3606 arrival, either by an ACL or by local_scan().
3608 7. Added the smtp_incomplete_transaction log selector.
3610 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3611 was accepting AUTH without a new EHLO.
3613 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3614 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3617 10. Logging of TCP/IP connections (when configured) now happens in the main
3618 daemon process instead of the child process, so that the TCP/IP connection
3619 count is more accurate (but it can never be perfect).
3621 11. The use of "drop" in a nested ACL was not being handled correctly in the
3622 outer ACL. Now, if condition failure induced by the nested "drop" causes
3623 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3624 or "require"), the connection is dropped.
3626 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3627 that yield "discard" can now be used with an "accept" or a "discard" verb,
3628 but an error is generated for any others (because I can't see a useful way
3629 to define what should happen).
3631 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3632 are now processed in the same way as lines in the Exim configuration file.
3633 In particular, continuation lines are supported.
3635 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3637 15. Added -ti meaning -t -i.
3639 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3640 domains, and warn by logging if others are found.
3642 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3643 returned from GIFCONF when Exim is trying to find the list of interfaces on
3644 a host. The code in os.c has been modified to copy each ifreq to an aligned
3645 structure in all cases.
3647 Also, in some cases, the returned ifreq's were being copied to a 'struct
3648 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3649 means the last couple of bytes of an IPv6 address could be chopped if the
3650 ifreq contained only a normal sockaddr (14 bytes storage).
3652 18. Named domain lists were not supported in the hosts_treat_as_local option.
3653 An entry such as +xxxx was not recognized, and was treated as a literal
3656 19. Ensure that header lines added by a DATA ACL are included in the reject log
3657 if the ACL subsequently rejects the message.
3659 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3660 MD5 (which is deprecated).
3662 21. When testing a filter file using -bf, Exim was writing a message when it
3663 took the sender from a "From " line in the message, but it was not doing so
3664 when it took $return_path from a Return-Path: header line. It now does.
3666 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3667 with a valid header line field name (a series of printing characters
3668 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3670 23. Changed "disc" in the source to "disk" to conform to the documentation and
3671 the book and for uniformity.
3673 24. Ignore Sendmail's -Ooption=value command line item.
3675 25. When execve() failed while trying to run a command in a pipe transport,
3676 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3677 could be confused with a return value of 69 from the command itself. This
3678 has been changed to 127, the value the shell returns if it is asked to run
3679 a non-existent command. The wording for the related log line suggests a
3680 non-existent command as the problem.
3682 26. If received_header_text expands to an empty string, do not add a Received:
3683 header line to the message. (Well, it adds a token one on the spool, but
3684 marks it "old" so that it doesn't get used or transmitted.)
3686 27. Installed eximstats 1.28 (addition of -nt option).
3688 28. There was no check for failure on the call to getsockname() in the daemon
3689 code. This can fail if there is a shortage of resources on the system, with
3690 ENOMEM, for example. A temporary error is now given on failure.
3692 29. Contrary to the C standard, it seems that in some environments, the
3693 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3694 program. Exim now does this explicitly; it affects the formatting of
3695 timestamps using strftime().
3697 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3698 complaints. I've now initialized all the variables, to avoid this.
3700 32. Header lines added by a system filter were not being "seen" during
3701 transport-time rewrites.
3703 33. The info_callback() function passed to OpenSSL is set up with type void
3704 (*)(SSL *, int, int), as described somewhere. However, when calling the
3705 function (actually a macro) that sets it up, the type void(*)() is
3706 expected. I've put in a cast to prevent warnings from picky compilers.
3708 34. If a DNS black list lookup found a CNAME record, but there were no A
3709 records associated with the domain it pointed at, Exim crashed.
3711 35. If a DNS black list lookup returned more than one A record, Exim ignored
3712 all but the first. It now scans all returned addresses if a particular IP
3713 value is being sought. In this situation, the contents of the
3714 $dnslist_value variable are a list of all the addresses, separated by a
3717 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3718 to accept a host name and all its aliases if the forward lookup for any of
3719 them yielded the IP address of the incoming connection. Now it accepts only
3720 those names whose forward lookup yields the correct IP address. Any other
3721 names are discarded. This closes a loophole whereby a rogue DNS
3722 administrator could create reverse DNS records to break through a
3723 wildcarded host restriction in an ACL.
3725 37. If a user filter or a system filter that ran in a subprocess used any of
3726 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3727 the wrong values were passed to the pipe command ($thisaddress had the
3728 value of $0, $0 had the value of $1, etc). This bug was introduced by
3729 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3731 38. Improved the line breaking for long SMTP error messages from ACLs.
3732 Previously, if there was no break point between 40 and 75 characters, Exim
3733 left the rest of the message alone. Two changes have been made: (a) I've
3734 reduced the minimum length to 35 characters; (b) if it can't find a break
3735 point between 35 and 75 characters, it looks ahead and uses the first one
3736 that it finds. This may give the occasional overlong line, but at least the
3737 remaining text gets split now.
3739 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3740 file descriptors might be low, and that setting 1000 would always raise it.
3741 It turns out that in some environments, the limit is already over 1000 and
3742 that lowering it causes trouble. So now Exim takes care not to decrease it.
3744 40. When delivering a message, the value of $return_path is set to $sender_
3745 address at the start of routing (routers may change the value). By an
3746 oversight, this default was not being set up when an address was tested by
3747 -bt or -bv, which affected the outcome if any router or filter referred to
3750 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3751 log only when "message" or "log_message" are set. However, if one of the
3752 conditions was an address verification, or a call to a nested ACL, the
3753 messages generated by the underlying test were being passed through. This
3754 no longer happens. The underlying message is available in $acl_verify_
3755 message for both "message" and "log_message" expansions, so it can be
3756 passed through if needed.
3758 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3759 new expansion $bh_ to give the encoded byte string without charset
3760 translation. Translation happens only if iconv() is available; HAVE_ICONV
3761 indicates this at build time. HEADERS_CHARSET gives the charset to
3762 translate to; headers_charset can change it in the configuration, and
3763 "headers charset" can change it in an individual filter file.
3765 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3766 that creates RFC 2047 encoded "words" labels them as that charset instead
3767 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3768 expansion operator; (ii) when Exim creates a From: line for a local
3769 message; (iii) when a header line is rewritten to include a "phrase" part.
3771 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3772 buggy, causing it to skip the first lines of messages whose message ID
3773 ended in 'D'. This would not have bitten before Exim release 4.14, because
3774 message IDs were unlikely to end in 'D' before then. The effect was to have
3775 incorrect size information for certain domains.
3777 45. #include "config.h" was missing at the start of the crypt16.c module. This
3778 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3781 46. If there was a timeout during a "random" callout check, Exim treated it as
3782 a failure of the random address, and carried on sending RSET and the real
3783 address. If the delay was just some slowness somewhere, the response to the
3784 original RCPT would be taken as a response to RSET and so on, causing
3785 mayhem of various kinds.
3787 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3788 when I implemented it. It didn't allow for the fact that some option values
3789 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3790 the right test for positive values.
3792 48. Domain names in DNS records are case-independent. Exim always looks them up
3793 in lower case. Some resolvers return domain names in exactly the case they
3794 appear in the zone file, that is, they may contain uppercase letters. Not
3795 all resolvers do this - some return always lower case. Exim was treating a
3796 change of case by a resolver as a change of domain, similar to a widening
3797 of a domain abbreviation. This triggered its re-routing code and so it was
3798 trying to route what was effectively the same domain again. This normally
3799 caused routing to fail (because the router wouldn't handle the domain
3800 twice). Now Exim checks for this case specially, and just changes the
3801 casing of the domain that it ultimately uses when it transmits the message
3804 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3807 50. If a filter generated a file delivery with a non-absolute name (possible if
3808 no home directory exists for the router), the forbid_file option was not
3811 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3812 the existing equality matching.
3814 52. Exim was using ints instead of ino_t variables in some places where it was
3815 dealing with inode numbers.
3817 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3818 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3819 called TMPDIR, and if it finds it is different, it changes its value.
3821 54. The smtp_printf() function is now made available to local_scan() so
3822 additional output lines can be written before returning. There is also an
3823 smtp_fflush() function to enable the detection of a dropped connection.
3824 The variables smtp_input and smtp_batched_input are exported to
3827 55. Changed the default runtime configuration: the message "Unknown user"
3828 has been removed from the ACL, and instead placed on the localuser router,
3829 using the cannot_route_message feature. This means that any verification
3830 failures that generate their own messages won't get overridden. Similarly,
3831 the "Unrouteable address" message that was in the ACL for unverifiable
3832 relay addresses has also been removed.
3834 56. Added hosts_avoid_esmtp to the smtp transport.
3836 57. The exicyclog script was not checking for the esoteric option
3837 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3838 will work only if exicyclog is run under the appropriate euid.
3840 58. Following a discussion on the list, the rules by which Exim recognises line
3841 endings on incoming messages have been changed. The -dropcr and drop_cr
3842 options are now no-ops, retained only for backwards compatibility. The
3843 following line terminators are recognized: LF CRLF CR. However, special
3844 processing applies to CR:
3846 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3847 nor a local message in the state where . is a terminator.
3849 (ii) If a bare CR is encountered in a header line, an extra space is added
3850 after the line terminator so as not to end the header. The reasoning
3851 behind this is that bare CRs in header lines are most likely either
3852 to be mistakes, or people trying to play silly games.
3854 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3855 was being incorrectly given as 18 bytes larger than it should have been.
3856 This is a VOB (very old bug).
3858 60. This may never have affected anything current, but just in case it has:
3859 When the local host is found other than at the start of a list of hosts,
3860 the local host, those with the same MX, and any that follow, are discarded.
3861 When the list in question was part of a longer list of hosts, the following
3862 hosts (not currently being processed) were also being discarded. This no
3863 longer happens. I'm not sure if this situation could ever has previously
3866 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3869 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3870 header. This is something that is recommended in a new Internet Draft, and
3871 is something that is documented as being done by Sendmail. There are two
3872 possible values. For messages generated by the autoreply transport, Exim
3875 Auto-Submitted: auto-replied
3877 whereas for all other generated messages (e.g. bounces) it adds
3879 Auto-Submitted: auto-generated
3881 63. The "personal" condition in filters now includes a test for the
3882 Auto-Submitted: header. If it contains the string "auto-" the message it
3883 not considered personal.
3885 64. Added rcpt_include_affixes as a generic transport option.
3887 65. Added queue_only_override (default true).
3889 66. Added the syslog_duplication option.
3891 67. If what should have been the first header line of a message consisted of
3892 a space followed by a colon, Exim was mis-interpreting it as a header line.
3893 It isn't of course - it is syntactically invalid and should therefore be
3894 treated as the start of the message body. The misbehaviour could have
3895 caused a number of strange effects, including loss of data in subsequent
3896 header lines, and spool format errors.
3898 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3899 client host had authenticated. This control can now be exercised by an ACL
3900 for more flexibility.
3902 69. By default, callouts do not happen when testing with -bh. There is now a
3903 variant, -bhc, which does actually run the callout code, including
3904 consulting and updating the callout cache.
3906 70. Added support for saslauthd authentication, courtesy of Alexander
3909 71. If statvfs() failed on the spool or log directories while checking their
3910 size for availability, Exim confusingly gave the error "space shortage".
3911 Furthermore, in debugging mode it crashed with a floating point exception.
3912 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3913 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3914 serious problem, Exim now writes to the main and panic logs when this
3915 happens, with details of the failure. It then refuses to accept the
3916 incoming message, giving the message "spool directory problem" or "log
3917 directory problem" with a 421 code for SMTP messages.
3919 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3920 0, 1, and 2 exist, because some OS complain for execs without them (see
3921 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3922 descriptors. However, the code omitted to check that the open succeeded,
3923 causing mysterious errors if for some reason the permissions on /dev/null
3924 got screwed. Now Exim writes a message to the main and panic logs, and
3925 bombs out if it can't open /dev/null.
3927 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3928 interact so that it is all more flexible. It is supposed to remain
3929 backwards compatible. Also added extra_local_interfaces.
3931 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3932 to bomb out with an assertion failure - to the client this appears as a
3933 connection drop. This problem occurs in the part of the code that was taken
3934 from the Samba project. Fortunately, the assertion is in a very simple
3935 function, so I have fixed this by reproducing the function inline in the
3936 one place where it is called, and arranging for authentication to fail
3937 instead of killing the process with assert().
3939 75. The SPA client code was not working when the server requested OEM rather
3940 than Unicode encoding.
3942 76. Added code to make require_files with a specific uid setting more usable in
3943 the case where statting the file as root fails - usually a non-root-mounted
3944 NFS file system. When this happens and the failure is EACCES, Exim now
3945 forks a subprocess and does the per-uid checking as the relevant uid.
3947 77. Added process_log_path.
3949 78. If log_file_path was not explicitly set, a setting of check_log_space or
3950 check_log_inodes was ignored.
3952 79. If a space check for the spool or log partitions fails, the incident is now
3953 logged. Of course, in the latter case the data may get lost...
3955 80. Added the %p formatting code to string_format() so that it can be used to
3956 print addresses in debug_print(). Adjusted all the address printing in the
3957 debugging in store.c to use %p rather than %d.
3959 81. There was a concern that a line of code in smtp_in.c could overflow a
3960 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3961 initially expressed, the concern was not well-founded, because trailing
3962 spaces are removed early. However, if the trailing spaces were followed by
3963 a NULL, they did not get removed, so the overflow was possible. Two fixes
3966 (a) I re-wrote the offending code in a cleaner fashion.
3967 (b) If an incoming SMTP command contains a NULL character, it is rejected
3970 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3971 runs initgroups(), so that if the Exim user is in any additional groups,
3972 they will be used during message reception.
3978 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob