1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.346 2006/04/25 10:06:30 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).
47 PH/01 The code for finding all the local interface addresses on a FreeBSD
48 system running IPv6 was broken. This may well have applied to all BSD
49 systems, as well as to others that have similar system calls. The broken
50 code found IPv4 interfaces correctly, but gave incorrect values for the
51 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
52 that it would not match correctly against @[] and not recognize the IPv6
55 PH/02 The ipliteral router was not recognizing addresses of the form user@
56 [ipv6:....] because it didn't know about the "ipv6:" prefix.
58 PH/03 Added disable_ipv6.
60 PH/04 Changed $reply_address to use the raw form of the headers instead of the
61 decoded form, because it is most often used to construct To: headers
62 lines in autoreplies, and the decoded form may well be syntactically
63 invalid. However, $reply_address has leading white space removed, and all
64 newlines turned into spaces so that the autoreply transport does not
67 PH/05 If group was specified without a user on a router, and no group or user
68 was specified on a transport, the group from the router was ignored.
70 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
71 for visible compile-time settings that can be used to change these
72 numbers, for those that want even more. Backwards compatibility with old
73 spool files has been maintained. However, going back to a previous Exim
74 release will lost any variables that are in spool files.
76 PH/07 Two small changes when running in the test harness: increase delay when
77 passing a TCP/IP connection to a new process, in case the original
78 process has to generate a bounce, and remove special handling of
79 127.0.0.2 (sic), which is no longer necessary.
81 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
82 be the same on different OS.
84 PH/09 Moved a debug statement in filter processing to avoid a race problem when
87 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
88 whether --show-vars was specified or not
90 JJ/02 exipick: Added support for new ACL variable spool format introduced
93 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
94 syntactically invalid From: or Reply-to: line, and a filter used this to
95 generate an autoreply, and therefore failed to obtain an address for the
96 autoreply, Exim could try to deliver to a non-existent relative file
97 name, causing unrelated and misleading errors. What now happens is that
98 it logs this as a hard delivery error, but does not attempt to create a
101 PH/11 The exinext utility has a -C option for testing purposes, but although
102 the given file was scanned by exinext itself; it wasn't being passed on
103 when Exim was called.
105 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
106 an end-of-file indication when reading a command response.
108 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
109 compiled. In many other places in Exim, IPv6 addresses are always
110 recognized, so I have changed this. It also means that IPv4 domain
111 literals of the form [IPV4:n.n.n.n] are now always recognized.
113 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
114 used if the router is not running as root, for example, when verifying at
115 ACL time, or when using -bh. The debugging output from this situation was
116 non-existent - all you got was a failure to exec. I have made two
119 (a) Failures to set uid/gid, the current directory, or a process leader
120 in a subprocess such as that created by queryprogram now generate
121 suitable debugging ouput when -d is set.
123 (b) The queryprogram router detects when it is not running as root,
124 outputs suitable debugging information if -d is set, and then runs
125 the subprocess without attempting to change uid/gid.
127 PH/15 Minor change to Makefile for building test_host (undocumented testing
130 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
131 additional section of a DNS packet that returns MX or SRV records.
132 Instead, it always explicitly searches for A/AAAA records. This avoids
133 major problems that occur when a DNS server includes only records of one
134 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
135 fixed another bug: if SRV records were looked up and the corresponding
136 address records were *not* found in the additional section, the port
137 values from the SRV records were lost.
139 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
140 using the correct key (the original address) when searching the retry
141 rules in order to find which one to use for generating the retry hint.
143 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
144 adding the default one. Similarly, if it contains a Reply-To: header, the
145 errors_reply_to option, if set, is not used.
147 PH/19 When calculating a retry time, Exim used to measure the "time since
148 failure" by looking at the "first failed" field in the retry record. Now
149 it does not use this if it is later than than the arrival time of the
150 message. Instead it uses the arrival time. This makes for better
151 behaviour in cases where some deliveries succeed, thus re-setting the
152 "first failed" field. An example is a quota failure for a huge message
153 when small messages continue to be delivered. Without this change, the
154 "time since failure" will always be short, possible causing more frequent
155 delivery attempts for the huge message than are intended.
156 [Note: This change was subsequently modified - see PH/04 for 4.62.]
158 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
159 $1, $2, $3) because the numerical variables can be reset during some
160 expansion items (e.g. "match"), thereby losing the authentication data.
162 PH/21 Make -bV show the size of off_t variables so that the test suite can
163 decide whether to run tests for quotas > 2G.
165 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
166 mailbox_size, and mailbox_filecount in the appendfile transport. If a
167 filecount value is greater than 2G or if a quota value is greater than 2G
168 on a system where the size of off_t is not greater than 4, a panic error
171 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
172 never match. The debug and -bh output now contains an explicit error
173 message indicating a malformed IPv4 address or mask.
175 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
176 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
179 PH/25 Do not write to syslog when running in the test harness. The only
180 occasion when this arises is a failure to open the main or panic logs
181 (for which there is an explicit test).
183 PH/26 Added the /no_tell option to "control=freeze".
185 PH/27 If a host name lookup failed very early in a connection, for example, if
186 the IP address matched host_lookup and the reverse lookup yielded a name
187 that did not have a forward lookup, an error message of the form "no IP
188 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
189 could be logged. Now it outputs the IP address instead of "NULL".
191 PH/28 An enabling patch from MH: add new function child_open_exim2() which
192 allows the sender and the authenticated sender to be set when
193 submitting a message from within Exim. Since child_open_exim() is
194 documented for local_scan(), the new function should be too.
196 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
197 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
198 results in an empty string is now treated as unset.
200 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
202 PH/31 Added sender_verify_fail logging option.
204 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
205 needed by RFC 822 but not by RFC 2822 was commented out. I have now
206 tidied the source and removed it altogether.
208 PH/33 When a queue run was abandoned because the load average was too high, a
209 log line was always written; now it is written only if the queue_run log
210 selector is set. In addition, the log line for abandonment now contains
211 information about the queue run such as the pid. This is always present
212 in "start" and "stop" lines but was omitted from the "abandon" line.
214 PH/34 Omit spaces between a header name and the colon in the error message that
215 is given when verify = headers_syntax fails (if there are lots of them,
216 the message gets confusing).
218 PH/35 Change the default for dns_check_names_pattern to allow slashes within
219 names, as there are now some PTR records that contain slashes. This check
220 is only to protect against broken name servers that fall over on strange
221 characters, so the fact that it applies to all lookups doesn't matter.
223 PH/36 Now that the new test suite is complete, we can remove some of the
224 special code in Exim that was needed for the old test suite. For example,
225 sorting DNS records because real resolvers return them in an arbitrary
226 order. The new test suite's fake resolver always returns records in the
229 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
230 bounces) except when queue_only is set, to avoid logging races between
231 the different processes.
233 PH/38 Panic-die if .include specifies a non-absolute path.
235 PH/39 A tweak to the "H" retry rule from its user.
237 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
238 a label. They prevented compilation on older perls.
240 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
241 a warning to be raised on newish perls.
243 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
244 on queue. Changes to match documented behaviour of showing count of
245 messages matching specified criteria.
247 PH/40 Changed the default ident timeout from 30s to 5s.
249 PH/41 Added support for the use of login_cap features, on those BSD systems
250 that have them, for controlling the resources used by pipe deliveries.
252 PH/42 The content-scanning code uses fopen() to create files in which to put
253 message data. Previously it was not paying any attention to the mode of
254 the files. Exim runs with umask(0) because the rest of the code creates
255 files with open(), and sets the required mode explicitly. Thus, these
256 files were ending up world-writeable. This was not a big issue, because,
257 being within the spool directory, they were not world-accessible. I have
258 created a function called modefopen, which takes an additional mode
259 argument. It sets umask(777), creates the file, chmods it to the required
260 mode, then resets the umask. All the relevant calls to fopen() in the
261 content scanning code have been changed to use this function.
263 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
264 to 24 hours. This avoids potential overflow problems when processing G
265 and H retry rules. I suspect nobody ever tinkers with this value.
267 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
269 PH/45 When the plaintext authenticator is running as a client, the server's
270 challenges are checked to ensure they are valid base64 strings. By
271 default, the authentication attempt is cancelled if an invalid string is
272 received. Setting client_ignore_invalid_base64 true ignores these errors.
273 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
274 they are received. Thus, the responses can be made to depend on the
275 challenges. If an invalid string is ignored, an empty string is placed in
278 PH/46 Messages that are created by the autoreply transport now contains a
279 References: header, in accordance with RFCs 2822 and 3834.
281 PH/47 Added authenticated_sender_force to the smtp transport.
283 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
285 PH/49 Installed latest patch from the Sieve maintainer.
287 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
288 was also set, the appendfile transport was unnecessarily scanning a
289 directory of message files (e.g. for maildir delivery) to find the count
290 of files (along with the size), even though it did not need this
291 information. It now does the scan only if it needs to find either the
292 size of the count of files.
294 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
296 PH/52 Two bugs concerned with error handling when the smtp transport is
299 (i) Exim was not creating retry information for temporary errors given
300 for individual recipients after the DATA command when the smtp transport
301 was used in LMTP mode. This meant that they could be retried too
302 frequently, and not timed out correctly.
304 (ii) Exim was setting the flag that allows error details to be returned
305 for LMTP errors on RCPT commands, but not for LMTP errors for individual
306 recipients that were returned after the DATA command.
308 PH/53 This is related to PH/52, but is more general: for any failing address,
309 when detailed error information was permitted to be returned to the
310 sender, but the error was temporary, then after the final timeout, only
311 "retry timeout exceeded" was returned. Now it returns the full error as
312 well as "retry timeout exceeded".
314 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
315 do this, and (what is worse) MTAs that accept it.
317 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
318 will now be deprecated.
320 PH/56 New os.c-cygwin from the Cygwin maintainer.
322 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
323 formats (previously only available in exim formats via -bpr, -bpru,
324 and -bpra. Now also available in native and exiqgrep formats)
326 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
327 with very large, slow to parse queues
329 JJ/08 exipick: added ! as generic prefix to negate any criteria format
331 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
333 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
334 responses to authentication challenges, though it was showing the
335 challenges; (ii) I've removed the CR characters from the debug output for
338 PH/58 Allow for the insertion of a newline as well as a space when a string
339 is turned into more than one encoded-word during RFC 2047 encoding. The
340 Sieve code now uses this.
342 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
343 data_4xx, lost_connection, tls_required.
345 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
346 message was being sent as an SMTP response.
348 PH/61 Add -l and -k options to exicyclog.
350 PH/62 When verifying, if an address was redirected to one new address, so that
351 verification continued, and the new address failed or deferred after
352 having set something in $address_data, the value of $address_data was not
353 passed back to the ACL. This was different to the case when no
354 redirection occurred. The value is now passed back in both cases.
356 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
357 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
358 use login_cap.h, so on its own it isn't the distinguishing feature. The
359 new name refers directly to the setclassresources() function.
361 PH/65 Added configuration files for NetBSD3.
363 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
365 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
366 is preferred over IPv4.
368 PH/68 The bounce_return_message and bounce_return_body options were not being
369 honoured for bounces generated during the reception of non-SMTP messages.
370 In particular, this applied to messages rejected by the ACL. This bug has
371 been fixed. However, if bounce_return_message is true and bounce_return_
372 body is false, the headers that are returned for a non-SMTP message
373 include only those that have been read before the error was detected.
374 (In the case of an ACL rejection, they have all been read.)
376 PH/69 The HTML version of the specification is now built in a directory called
377 spec_html instead of spec.html, because the latter looks like a path with
378 a MIME-type, and this confuses some software.
380 PH/70 Catch two compiler warnings in sieve.c.
382 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
383 function verify_get_ident() calls ip_connect() to connect a socket, but
384 if the "connect()" function timed out, ip_connect() used to close the
385 socket. However, verify_get_ident() also closes the socket later, and in
386 between Exim writes to the log, which may get opened at this point. When
387 the socket was closed in ip_connect(), the log could get the same file
388 descriptor number as the socket. This naturally causes chaos. The fix is
389 not to close the socket in ip_connect(); the socket should be closed by
390 the function that creates it. There was only one place in the code where
391 this was missing, in the iplookup router, which I don't think anybody now
392 uses, but I've fixed it anyway.
394 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
395 well as to direct DNS lookups. Otherwise the handling of names in host
396 lists is inconsistent and therefore confusing.
402 PH/01 Two changes to the default runtime configuration:
404 (1) Move the checks for relay_from_hosts and authenticated clients from
405 after to before the (commented out) DNS black list checks.
407 (2) Add control=submission to the relay_from_hosts and authenticated
408 clients checks, on the grounds that messages accepted by these
409 statements are most likely to be submissions.
411 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
413 (1) Generate an error if the third argument for the ${prvs expansion is
416 (2) Treat a missing third argument of ${prvscheck as if it were an empty
419 (3) Reset the variables that are obtained from the first argument of
420 ${prvscheck and used in the second argument before leaving the code,
421 because their memory is reclaimed, so using them afterwards may do
424 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
425 one (it's much easier than Tom thought :-).
427 (5) Because of (4), we can now allow for the use of $prvscheck_result
428 inside the third argument.
430 PH/03 For some reason, the default setting of PATH when running a command from
431 a pipe transport was just "/usr/bin". I have changed it to
434 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
435 anything to be listed in the output from -bV.
437 PH/05 When a filter generated an autoreply, the entire To: header line was
438 quoted in the delivery log line, like this:
440 => >A.N.Other <ano@some.domain> <original@ddress> ...
442 This has been changed so that it extracts the operative address. There
443 may be more than one such address. If so, they are comma-separated, like
446 => >ano@some.domain,ona@other.domain <original@ddress> ...
448 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
449 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
450 not being looked up in the rDNS to get a host name, Exim was showing the
451 IP address twice in Received: lines, even though the IP addresses were
452 identical. For example:
454 Received: from [1.2.3.4] (helo=[1.2.3.4])
456 However, if the real host name was known, it was omitting the HELO data
457 if it matched the actual IP address. This has been tidied up so that it
458 doesn't show the same IP address twice.
460 PH/07 When both +timestamp and +memory debugging was on, the value given by
461 $tod_xxx expansions could be wrong, because the tod_stamp() function was
462 called by the debug printing, thereby overwriting the timestamp buffer.
463 Debugging no longer uses the tod_stamp() function when +timestamp is set.
465 PH/08 When the original message was included in an autoreply transport, it
466 always said "this is a copy of the message, including all the headers",
467 even if body_only or headers_only was set. It now gives an appropriate
470 PH/09 Applied a patch from the Sieve maintainer which:
472 o fixes some comments
473 o adds the (disabled) notify extension core
474 o adds some debug output for the result of if/elsif tests
475 o points to the current vacation draft in the documentation
476 and documents the missing references header update
480 o fixes a bug in processing the envelope test (when testing
481 multiple envelope elements, the last element determinted the
484 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
485 Electronic Mail") by including:
487 Auto-submitted: auto-generated
489 in the messages that it generates (bounce messages and others, such as
490 warnings). In the case of bounce messages for non-SMTP mesages, there was
491 also a typo: it was using "Auto_submitted" (underscore instead of
492 hyphen). Since every message generated by Exim is necessarily in response
493 to another message, thes have all been changed to:
495 Auto-Submitted: auto-replied
497 in accordance with these statements in the RFC:
499 The auto-replied keyword:
501 - SHOULD be used on messages sent in direct response to another
502 message by an automatic process,
504 - MUST NOT be used on manually-generated messages,
506 - MAY be used on Delivery Status Notifications (DSNs) and Message
507 Disposition Notifications (MDNs),
509 - MUST NOT be used on messages generated by automatic or periodic
510 processes, except for messages which are automatic responses to
513 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
514 to the default Received: header definition.
516 PH/12 Added log selector acl_warn_skipped (default on).
518 PH/13 After a successful wildlsearch lookup, discard the values of numeric
519 variables because (a) they are in the wrong storage pool and (b) even if
520 they were copied, it wouldn't work properly because of the caching.
522 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
523 checking when decoding. Apparently there are clients that generate
524 overlong encoded strings. Why am I not surprised?
526 PH/15 If the first argument of "${if match_address" was not empty, but did not
527 contain an "@" character, Exim crashed. Now it writes a panic log message
528 and treats the condition as false.
530 PH/16 In autoreply, treat an empty string for "once" the same as unset.
532 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
533 extension "envelope-auth". The code is finished and in agreement with
534 other implementations, but there is no documentation so far and in fact,
535 nobody wrote the draft yet. This extension is currently #undef'ed, thus
536 not changing the active code.
538 Print executed "if" and "elsif" statements when debugging is used. This
539 helps a great deal to understand what a filter does.
541 Document more things not specified clearly in RFC3028. I had all this
542 sorted out, when out of a sudden new issues came to my mind. Oops."
544 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
547 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
548 canonical form (as documented). However, after a host name lookup from
549 the IP address, check_host() was doing a simple string comparison with
550 addresses acquired from the DNS when checking that the found name did
551 have the original IP as one of its addresses. Since any found IPv6
552 addresses are likely to be in abbreviated form, the comparison could
553 fail. Luckily, there already exists a function for doing the comparison
554 by converting both addresses to binary, so now that is used instead of
557 PH/20 There was another similar case to PH/19, when a complete host name was
558 given in a host list; looking up its IP address could give an abbreviated
559 form, whereas the current host's name might or might not be abbreviated.
560 The same fix has been applied.
566 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
567 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
570 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
571 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
573 PH/03 Typo: missing ".o" in src/pcre/Makefile.
575 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
576 header line, restrict the check to what is listed in RFCs 2369 and 2929.
577 Also, for "Auto-Submitted", treat anything other than "no" as
578 non-personal, in accordance with RFC 3834. (Previously it treated
579 anything starting "auto-" as non-personal.)
581 TF/01 The control=submission/name=... option had a problem with syntax
582 errors if the name included a slash character. The /name= option
583 now slurps the rest of the string, so it can include any characters
584 but it must come last in the list of options (after /sender_retain
587 PH/05 Some modifications to the interface to the fake nameserver for the new
595 TK/01 Added the "success_on_redirect" address verification option. See
596 NewStuff for rationale and an example.
598 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
600 PH/02 Patch to exigrep to allow it to work on syslog lines.
602 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
603 fread() to skip over the body file's header line, because in Cygwin the
604 header line is locked and is inaccessible.
606 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
607 co-exist for some time) to make it clear that it is the Exim ID that is
608 referenced, not the Message-ID: header line.
610 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
611 string_format() function, because snprintf() does not exist on all
614 PH/06 The use of forbid_filter_existstest now also locks out the use of the
615 ${stat: expansion item.
617 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
618 protocol synchronization error", to keep the pedants happy.
620 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
621 well as for IRIX systems, when gcc is being used. See the host.c source
624 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
626 PH/10 Named domain lists were not working if used in a queue_smtp_domains
629 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
630 transport and to the smtp transport in LMTP mode.
632 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
634 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
635 run a filter in a subprocess. This could lead to confusion in subsequent
636 lookups in the parent process. There should also be a search_tidyup() at
637 the end of the subprocess.
639 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
640 only if the host matched helo_try_verify_hosts, which caused the
641 verification to occur when the EHLO/HELO command was issued. The ACL just
642 tested the remembered result. Now, if a previous verification attempt has
643 not happened, "verify = helo" does it there and then.
645 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
647 TK/03 Fix log output including CR from clamd.
649 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
650 exist provoked a memory error which could cause a segfault.
652 PH/15 Installed PCRE 6.2
654 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
656 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
657 of the problem. Specifically, suggested +O2 rather than +O1 for the
660 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
662 PH/20 If a delivery was routed to a non-standard port by means of an SRV
663 record, the port was not correctly logged when the outgoing_port log
664 selector was set (it logged the transort's default port).
666 PH/21 Added support for host-specific ports to manualroute, queryprogram,
667 fallback_hosts, and "hosts" in the smtp transport.
669 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
670 host errors such as "Connection refused".
672 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
673 authentication with radiusclient 0.4.9:
675 - Error returned from rc_read_config was caught wrongly
676 - Username/password not passed on to radius server due to wrong length.
678 The presumption is that some radiusclient API changes for 4.51/PH/17
679 were not taken care of correctly. The code is still untested by me (my
680 Linux distribution still has 0.3.2 of radiusclient), but it was
681 contributed by a Radius user.
683 PH/24 When doing a callout, the value of $domain wasn't set correctly when
684 expanding the "port" option of the smtp transport.
686 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
687 while reading a MIME header. Thanks to Tom Hughes for a patch.
689 PH/24 Include config.h inside local_scan.h so that configuration settings are
692 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
693 that in an ACL for RCPT (for example), you can examine exactly what was
696 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
697 commands, but it was not correctly comparing the address with the actual
698 client host address. Thus, it would show the EHLO address in Received:
699 header lines when this was not necessary.
701 PH/27 Added the % operator to ${eval:}.
703 PH/28 Exim tries to create and chdir to its spool directory when it starts;
704 it should be ignoring failures (because with -C, for example, it has lost
705 privilege). It wasn't ignoring creation failures other than "already
708 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
711 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
712 because an input line was too long, either on its own, or by virtue of
713 too many continuations, the temporary file was not being removed, and the
714 return code was incorrect.
716 PH/31 Missing "BOOL" in function definition in filtertest.c.
718 PH/32 Applied Sieve patches from the maintainer.
720 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
722 PH/33 Added "verify = not_blind".
724 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
725 Local/Makefile (with some defaults set). These are used in built scripts
726 such as exicyclog, but they have never been used in the exim_install
727 script (though there are many overriding facilities there). I have
728 arranged that the exim_install script now takes note of these two
731 PH/35 Installed configuration files for Dragonfly.
733 PH/36 When a locally submitted message by a trusted user did not contain a
734 From: header, and the sender address was obtained from -f or from an SMTP
735 MAIL command, and the trusted user did not use -F to supply a sender
736 name, $originator_name was incorrectly used when constructing a From:
737 header. Furthermore, $originator_name was used for submission mode
738 messages from external hosts without From: headers in a similar way,
739 which is clearly wrong.
741 PH/37 Added control=suppress_local_fixups.
743 PH/38 When log_selector = +received_sender was set, and the addition of the
744 sender made the log line's construction buffer exactly full, or one byte
745 less than full, an overflow happened when the terminating "\n" was
748 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
749 when the result of a list match is failure because a DNS lookup failed.
751 PH/40 RM_COMMAND is now used in the building process.
753 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
754 the "build-* directories that it finds.
756 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
757 address in a domain literal was a prefix of an interface address.
759 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
760 when verifying a sender address, unless rewrite_headers is false.
762 PH/44 Wrote a long comment about why errors_to addresses are verified as
763 recipients, not senders.
765 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
766 the ratelimit ACL was added.
768 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
770 PH/46 Added extra information about PostgreSQL errors to the error string.
772 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
773 avoiding the need to install special zones in a real server. This is
774 backwards compatible; if it can't find the fake resolver, it drops back.
775 Thus, both old and new test suites can be run.
777 TF/02 Added util/ratelimit.pl
779 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
780 clock is set back in time.
782 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
783 Candler <B.Candler@pobox.com>.
785 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
786 OK for addresses that are the result of redirections.
788 PH/48 A number of further additions for the benefit of the new test suite,
789 including a fake gethostbyname() that interfaces to the fake DNS resolver
792 TF/06 The fix for widen_domains has also been applied to qualify_single and
793 search_parents which are the other dnslookup options that can cause
796 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
799 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
801 TF/07 Exim produced the error message "an SRV record indicated no SMTP
802 service" if it encountered an MX record with an empty target hostname.
803 The message is now "an MX or SRV record indicated no SMTP service".
805 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
806 if the DNS of the sending site is misconfigured. This is quite a
807 common situation. This change restores the behaviour of treating a
808 helo verification defer as a failure.
810 PH/51 If self=fail was set on a router, the bounce message did not include the
811 actual error message.
817 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
819 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
820 command itself ended in error, the underlying message about the transport
821 filter timeout was being overwritten with the pipe command error. Now the
822 underlying error message should be appended to the second error message.
824 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
826 PH/02 Reduce the amount of output that "make" produces by default. Full output
827 can still be requested.
829 PH/03 The warning log line about a condition test deferring for a "warn" verb
830 was being output only once per connection, rather than after each
831 occurrence (because it was using the same function as for successful
832 "warn" verbs). This seems wrong, so I have changed it.
834 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
835 it should not have, which might have caused a crash in the right
836 circumstances, but probably never did.
838 PH/04 Installed a modified version of Tony Finch's patch to make submission
839 mode fix the return path as well as the Sender: header line, and to
840 add a /name= option so that you can make the user's friendly name appear
843 TF/03 Added the control = fakedefer ACL modifier.
845 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
846 Mark Lowes for thorough testing.
848 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
850 TK/03 Merged latest SRS patch from Miles Wilton.
852 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
853 with the definition in sysexits.h (which is #included earlier).
854 Fortunately, Exim does not actually use EX_OK. The code used to try to
855 preserve the sysexits.h value, by assumimg that macro definitions were
856 scanned for macro replacements. I have been disabused of this notion,
857 so now the code just undefines EX_OK before #including unistd.h.
859 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
860 in the smtp transport. When a block could not be written in a single
861 write() function, the timeout was being re-applied to each part-write.
862 This seems wrong - if the receiver was accepting one byte at a time it
863 would take for ever. The timeout is now adjusted when this happens. It
864 doesn't have to be particularly precise.
866 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
867 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
869 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
870 without a domain if the check to <postmaster@domain> fails.
872 SC/01 Eximstats: added -xls and the ability to specify output files
873 (patch written by Frank Heydlauf).
875 SC/02 Eximstats: use FileHandles for outputing results.
877 SC/03 Eximstats: allow any combination of xls, txt, and html output.
879 SC/04 Eximstats: fixed display of large numbers with -nvr option
881 SC/05 Eximstats: fixed merging of reports with empty tables.
883 SC/06 Eximstats: added the -include_original_destination flag
885 SC/07 Eximstats: removed tabs and trailing whitespace.
887 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
889 TK/06 MBOX spool code: Add real "From " MBOX separator line
890 so the .eml file is really in mbox format (even though
891 most programs do not really care). Patch from Alex Miller.
893 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
894 The latter is generated from $received_to and is only set if the
895 message has one envelope recipient. SA can use these headers,
896 obviously out-of-the-box. Patch from Alex Miller.
898 PH/08 The ${def test on a variable was returning false if the variable's
899 value was "0", contrary to what the specification has always said!
900 The result should be true unless the variable is empty.
902 PH/09 The syntax error of a character other than { following "${if
903 def:variable_name" (after optional whitespace) was not being diagnosed.
904 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
905 accidental colon was present, for example, could give incorrect results.
907 PH/10 Tidied the code in a number of places where the st_size field of a stat()
908 result is used (not including appendfile, where other changes are about
911 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
912 This involved changing a lot of size variables from int to off_t. It
913 should work with maildirs and everything.
915 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
916 spamd dying while we are connected to it.
918 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
921 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
922 function for BDB 4.3.
924 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
926 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
927 the sources for PCRE have been re-arranged and more files are now
930 PH/15 The code I had for printing potentially long long variables in PH/11
931 above was not the best (it lost precision). The length of off_t variables
932 is now inspected at build time, and an appropriate printing format (%ld
933 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
934 to be "long long int" or "long int". This is needed for the internal
935 formatting function string_vformat().
937 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
938 the configuration file to be ":syslog", then the script "guesses" where
939 the logs files are, rather than using the compiled in default. In our
940 case the guess is not the same as the compiled default, so the script
941 suddenly stopped working when I started to use syslog. The patch checks
942 to see if log_file_path is "". If so, it attempts to read it from exim
943 with no configuration file to get the compiled in version, before it
944 falls back to the previous guessing code."
946 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
947 implementing BATV in an Exim configuration. See NewStuff for the gory
950 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
951 Makefile that are specific to HP-UX.
953 PH/18 If the "use_postmaster" option was set for a recipient callout together
954 with the "random" option, the postmaster address was used as the MAIL
955 FROM address for the random test, but not for the subsequent recipient
956 test. It is now used for both.
958 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
959 patch removes a few documentation additions to RFC 3028, because the
960 latest draft now contains them. It adds the new en;ascii-case comparator
961 and a new error check for 8bit text in MIME parts. Comparator and
962 require names are now matched exactly. I enabled the subaddress
963 extension, but it is not well tested yet (read: it works for me)."
965 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
966 rework some of the code of TK/09 above to avoid the hardwired use of
967 "%lld" and "long long". Replaced the call to snprintf() with a call to
970 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
971 records point to non-existent hosts", "retry timeout exceeded", and
972 "retry time not reached for any host after a long failure period".
974 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
975 experimental DomainKeys support:
977 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
978 (2) On an error such as an illegally used "control", the wrong name for
979 the control was given.
981 These problems did NOT occur unless DomainKeys support was compiled.
983 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
985 PH/24 Added ${if match_ip condition.
987 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
988 databases so that it will be absolutely obvious if a crash occurs in the
989 DB library. This is a regular occurrence (often caused by mis-matched
992 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
993 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
994 picked up on a user's system that detects such things. There doesn't seem
995 to be a gcc warning option for this - only an attribute that has to be
996 put on the function's prototype. It seems that in Fedora Core 4 they have
997 set this on a number of new functions. No doubt there will be more in due
1000 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1001 specify a transport. However, if an address that was verified by such a
1002 router was the subject of a callout, Exim crashed because it tried to
1003 read the rcpt_include_affixes from the non-existent transport. Now it
1004 just assumes that the setting of that option is false. This bug was
1005 introduced by 4.51/PH/31.
1007 PH/28 Changed -d+all to exclude +memory, because that information is very
1008 rarely of interest, but it makes the output a lot bigger. People tend to
1009 do -d+all out of habit.
1011 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1012 code in os-type was giving problems when libc.so lives in lib64, like on
1015 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1016 aren't the modern standard, and it seems that some systems' include files
1017 don't always have them. Exim was already checking for some of the newer
1018 ones like T_AAAA, and defining it itself. I've added checks for all the
1019 record types that Exim uses.
1021 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1022 not automatically generating a new one, as it is supposed to. This
1023 prevented TLS from working. If the file did exist, but contained invalid
1024 data, a new version was generated, as expected. It was only the case of a
1025 non-existent file that was broken.
1027 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1028 with a change in libdomainkeys > 0.64.
1030 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1031 from DNS. If the selector record carries the flag, it now has
1032 precedence over the domain-wide flag.
1034 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1036 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1037 the use of a port name that isn't defined in /etc/services) occurred, the
1038 message was deferred as in a normal delivery, and thus remained on the
1039 spool, instead of being failed because of the mua_wrapper setting. This
1040 is now fixed, and I tidied up some of the mua_wrapper messages at the
1043 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1044 the messages in a hash of arrays rather than using individual hashes.
1045 This is a bit cleaner and results in dramatic memory savings, albeit
1046 at a slight CPU cost.
1048 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1049 as requested by Marc Sherman.
1051 SC/10 Eximstats: added histograms for user specified patterns as requested
1054 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1056 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1057 fopen() in the content-scanning modules that did not already have it.
1063 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1064 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1066 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1068 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1070 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1071 to test Sieve filters that use "vacation".
1073 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1074 that changes the way the GnuTLS parameters are stored in the cache file.
1075 The new format can be generated externally. For backward compatibility,
1076 if the data in the cache doesn't make sense, Exim assumes it has read an
1077 old-format file, and it generates new data and writes a new file. This
1078 means that you can't go back to an older release without removing the
1081 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1082 work if there are any delivery delays because "one_time" forces the
1083 parent to be marked "delivered", so its unseen clone is never tried
1084 again. For this reason, Exim now forbids the simultaneous setting of
1087 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1088 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1089 entry if you want to know the details. The fix, however, neglected to
1090 consider the case when local delivery batching is involved. The test for
1091 "previously delivered" was not happening when checking to see if an
1092 address could be batched with a previous (undelivered) one; under
1093 certain circumstances this could lead to multiple deliveries to the same
1096 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1097 in its include files, and this causes problems building Exim.
1099 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1100 header_syntax) but Exim was just ignoring anything given after a slash.
1101 In particular, this caused confusion with an attempt to use "verify =
1102 reverse_host_lookup/defer_ok". An error is now given when options are
1103 supplied for verify items that do not have them. (Maybe reverse_host_
1104 lookup should have a defer_ok option, but that's a different point.)
1106 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1107 defined by RFC 821) to 2048, because there were problems with some AUTH
1108 commands, and RFC 1869 says the size should be increased for extended
1109 SMTP commands that take arguments.
1111 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1114 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1115 "unknown" error; now it says that the functionality isn't in the binary.
1117 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1118 an address' error message when a string expansion fails (syntax or
1119 whatever). Otherwise the password may appear in the log. Following change
1120 PH/42 below, there is no longer a chance of it appearing in a bounce
1123 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1125 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1126 of its addresses was ever tried. (Bugzilla bug #2.)
1128 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1129 the result incorrectly in the debug output. (It correctly added a newline
1130 to what was transported.)
1132 TF/01 Added $received_time.
1134 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1135 commented out examples of how to interface to a virus scanner and to
1136 SpamAssassin. Also added commented examples of av_scanner and
1137 spamd_address settings.
1139 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1140 and controls are allowed in which ACLs. There were a couple of minor
1141 errors. Some of the entries in the conditions table (which is a table of
1142 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1143 negation of where the condition IS allowed.
1145 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1147 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1148 header file does not have a version number, so I've had to invent a new
1149 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1150 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1151 radiusclient), but it was contributed by a Radius user.
1153 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1154 files or directories, for OpenSSL.
1156 PH/19 When an Exim process that is running as root has to create an Exim log
1157 file, it does so in a subprocess that runs as exim:exim so as to get the
1158 ownership right at creation (otherwise, other Exim processes might see
1159 the file with the wrong ownership). There was no test for failure of this
1160 fork() call, which would lead to the process getting stuck as it waited
1161 for a non-existent subprocess. Forks do occasionally fail when resources
1162 run out. I reviewed all the other calls to fork(); they all seem to check
1165 PH/20 When checking for unexpected SMTP input at connect time (before writing
1166 the banner), Exim was not dealing correctly with a non-positive return
1167 from the read() function. If the client had disconnected by this time,
1168 the result was a log entry for a synchronization error with an empty
1169 string after "input=" when read() returned zero. If read() returned -1
1170 (an event I could not check), uninitialized data bytes were printed.
1171 There were reports of junk text (parts of files, etc) appearing after
1174 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1176 PH/22 Added support for macro redefinition, and (re)definition in between
1177 driver and ACL definitions.
1179 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1180 forgetting to use the resulting value; it was using the unexpanded value.
1182 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1183 hadn't been configured. The fix is from Juergen Kreileder, who
1184 understands it better than I do:
1186 "Here's what I see happening with three configured cyrus_sasl
1187 authenticators configured (plain, login, cram-md5):
1189 On startup auth_cyrus_sasl_init() gets called for each of these.
1190 This means three calls to sasl_listmech() without a specified mech_list.
1191 => SASL tests which mechs of all available mechs actually work
1192 => three warnings about OTP not working
1193 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1195 With the patch, sasl_listmech() also gets called three times. But now
1196 SASL's mech_list option is set to the server_mech specified in the the
1197 authenticator. Or in other words, the answer from sasl_listmech()
1198 gets limited to just the mech you're testing for (which is different
1200 => the return list contains just 'plain' or 'login', 'cram-md5' or
1201 nothing depending on the value of ob->server_mech.
1203 I've just tested the patch: Authentication still works fine,
1204 unavailable mechs specified in the exim configuration are still
1205 caught, and the auth.log warnings about OTP are gone."
1207 PH/25 When debugging is enabled, the contents of the command line are added
1208 to the debugging output, even when log_selector=+arguments is not
1211 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1212 answer is "GNU", and only if the return is "GNU/something" is the answer
1215 PH/27 $acl_verify_message is now set immediately after the failure of a
1216 verification in an ACL, and so is available in subsequent modifiers. In
1217 particular, the message can be preserved by coding like this:
1219 warn !verify = sender
1220 set acl_m0 = $acl_verify_message
1222 Previously, $acl_verify_message was set only while expanding "message"
1223 and "log_message" when a very denied access.
1225 PH/28 Modified OS/os.c-Linux with
1227 -#ifndef OS_LOAD_AVERAGE
1228 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1230 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1231 nomenclature these days.)
1233 PH/29 Installed patch from the Sieve maintainer that adds the options
1234 sieve_useraddress and sieve_subaddress to the redirect router.
1236 PH/30 In these circumstances:
1237 . Two addresses routed to the same list of hosts;
1238 . First host does not offer TLS;
1239 . First host accepts first address;
1240 . First host gives temporary error to second address;
1241 . Second host offers TLS and a TLS session is established;
1242 . Second host accepts second address.
1243 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1244 and peerdn, if requested) that were in fact used only for the second
1247 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1248 attention to any local part prefix or suffix that was matched by the
1249 router that accepted the address. It now behaves in the same way as it
1250 does for delivery: the affixes are removed from the local part unless
1251 rcpt_include_affixes is set on the transport.
1253 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1254 timeout during the DATA phase of an incoming message.
1256 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1257 applied a patch sanctioned by the Sieve maintainer.
1259 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1260 the uid or gid is negative. A case of a negative gid caused this to be
1261 noticed. The fix allows for either to be negative.
1263 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1264 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1267 PH/36 The change PH/12 above was broken. Fixed it.
1269 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1270 the grounds that routing the same address twice would always produce the
1271 same answer. This might have been true once, but it is certainly no
1272 longer true now. Routing a child address may depend on the previous
1273 routing that produced that child. Some complicated redirection strategies
1274 went wrong when messages had multiple recipients, and made Exim's
1275 behaviour dependent on the order in which the addresses were given.
1277 I have moved the duplicate checking until after the routing is complete.
1278 Exim scans the addresses that are assigned to local and remote
1279 transports, and removes any duplicates. This means that more work will be
1280 done, as duplicates will always all be routed, but duplicates are
1281 presumably rare, so I don't expect this is of any significance.
1283 For deliveries to pipes, files, and autoreplies, the duplicate checking
1284 still happens during the routing process, since they are not going to be
1287 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1288 It corrects a timeout issue with spamd. This is Ian's comment: "The
1289 background is that sometimes spamd either never reads data from a
1290 connection it has accepted, or it never writes response data. The exiscan
1291 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1292 blindly assumes that writes won't block so it may never time out."
1294 PH/39 Allow G after quota size as well as K and M.
1296 PH/40 The value set for $authenticated_id in an authenticator may not contain
1297 binary zeroes or newlines because the value is written to log lines and
1298 to spool files. There was no check on this. Now the value is run through
1299 the string_printing() function so that such characters are converted to
1300 printable escape sequences.
1302 PH/41 $message_linecount is a new variable that contains the total number of
1303 lines in the message. Compare $body_linecount, which is the count for the
1306 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1307 bounce and delay warning messages, except in certain special cases, which
1310 (a) An SMTP error message from a remote host;
1311 (b) A message specified in a :fail: redirection;
1312 (c) A message specified in a "fail" command in a system filter;
1313 (d) A message specified in a FAIL return from the queryprogram router;
1314 (e) A message specified by the cannot_route_message router option.
1316 In these cases only, Exim does include the error details in bounce and
1317 warning messages. There are also a few cases where bland messages such
1318 as "unrouteable address" or "local delivery error" are given.
1320 PH/43 $value is now also set for the "else" part of a ${run expansion.
1322 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1323 being worked on, but at least Exim now implements the latest version to
1326 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1327 process to complete was treated as a delivery failure, a timeout while
1328 writing the message to the pipe was logged, but erroneously treated as a
1329 successful delivery. Such timeouts include transport filter timeouts. For
1330 consistency with the overall process timeout, these timeouts are now
1331 treated as errors, giving rise to delivery failures by default. However,
1332 there is now a new Boolean option for the pipe transport called
1333 timeout_defer, which, if set TRUE, converts the failures into defers for
1334 both kinds of timeout. A transport filter timeout is now identified in
1337 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1338 systems where "make" and "gmake" are different, calling "gmake" at top
1339 level broke things. I've arranged for the value of $(MAKE) to be passed
1340 from the Makefile to this script so that it can call the same version of
1344 A note about Exim versions 4.44 and 4.50
1345 ----------------------------------------
1347 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1348 changes of various kinds. As a consequence, a big documentation update was
1349 needed. This delayed the release for rather longer than seemed good, especially
1350 in the light of a couple of (minor) security issues. Therefore, the changes
1351 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1352 release. So 4.44 and 4.50 are in effect two different branches that both start
1355 I have left the 4.50 change log unchanged; it contains all the changes since
1356 4.43. The change log for 4.44 is below; many of its items are identical to
1357 those for 4.50. This seems to be the most sensible way to preserve the
1358 historical information.
1364 1. Minor wording change to the doc/README.SIEVE file.
1366 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1367 computation of the current number of files was incorrect.
1369 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1370 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1373 4. Give more explanation in the error message when the command for a transport
1374 filter fails to execute.
1376 5. There are several places where Exim runs a non-Exim command in a
1377 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1378 was being done only for the command run by the queryprogram router. It is
1379 now done for all such subprocesses. The other cases are: ${run, transport
1380 filters, and the commands run by the lmtp and pipe transports.
1382 6. Added CONFIGURE_GROUP build-time option.
1384 7. Some older OS have a limit of 256 on the maximum number of file
1385 descriptors. Exim was using setrlimit() to set 1000 as a large value
1386 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1387 systems. I've change it so that if it can't get 1000, it tries for 256.
1389 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1390 was an oversight, and furthermore, ever since the addition of extra
1391 controls (e.g. 4.43/32), the checks on when to allow different forms of
1392 "control" were broken. There should now be diagnostics for all cases when a
1393 control that does not make sense is encountered.
1395 9. Added the /retain_sender option to "control=submission".
1397 10. $recipients is now available in the predata ACL (oversight).
1399 11. Tidy the search cache before the fork to do a delivery from a message
1400 received from the command line. Otherwise the child will trigger a lookup
1401 failure and thereby defer the delivery if it tries to use (for example) a
1402 cached ldap connection that the parent has called unbind on.
1404 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1405 of $address_data from the recipient verification was clobbered by the
1406 sender verification.
1408 13. The value of address_data from a sender verification is now available in
1409 $sender_address_data in subsequent conditions in the ACL statement.
1411 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1413 15. Added a new option "connect=<time>" to callout options, to set a different
1416 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1417 was its contents. (It was OK if the option was not defined at all.)
1419 17. A "Completed" log line is now written for messages that are removed from
1420 the spool by the -Mrm option.
1422 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1423 information about exactly what failed.
1425 19. Added -dd to debug only the daemon process.
1427 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1428 handles timeouts, both on the server side and network timeouts. Renamed the
1429 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1431 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1432 It is now set to "smtps".
1434 22. $host_address is now set to the target address during the checking of
1435 ignore_target_hosts.
1437 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1438 being passed; this would have caused $sender_host_name to have been used if
1439 matching the list had actually called for a host name (not very likely,
1440 since this list is usually IP addresses). A host name is now passed as
1443 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1444 code that specifies a non-restarting handler (typically sigaction() in
1445 modern systems) in an attempt to fix a rare and obscure crash bug.
1447 25. Narrowed the window for a race in the daemon that could cause it to ignore
1448 SIGCHLD signals. This is not a major problem, because they are used only to
1449 wake it up if nothing else does.
1451 26. A malformed maildirsize file could cause Exim to calculate negative values
1452 for the mailbox size or file count. Odd effects could occur as a result.
1453 The maildirsize information is now recalculated if the size or filecount
1456 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1457 support for a long time. Removed HAVE_SYS_VFS_H.
1459 28. Installed the latest version of exipick from John Jetmore.
1461 29. In an address list, if the pattern was not a regular expression, an empty
1462 subject address (from a bounce message) matched only if the pattern was an
1463 empty string. Non-empty patterns were not even tested. This was the wrong
1464 because it is perfectly reasonable to use an empty address as part of a
1465 database query. An empty address is now tested by patterns that are
1466 lookups. However, all the other forms of pattern expect the subject to
1467 contain a local part and a domain, and therefore, for them, an empty
1468 address still always fails if the pattern is not itself empty.
1470 30. Exim went into a mad DNS loop when attempting to do a callout where the
1471 host was specified on an smtp transport, and looking it up yielded more
1472 than one IP address.
1474 31. Re-factored the code for checking spool and log partition space into a
1475 function that finds that data and another that does the check. The former
1476 is then used to implement four new variables: $spool_space, $log_space,
1477 $spool_inodes, and $log_inodes.
1479 32. The RFC2047 encoding function was originally intended for short strings
1480 such as real names; it was not keeping to the 75-character limit for
1481 encoded words that the RFC imposes. It now respects the limit, and
1482 generates multiple encoded words if necessary. To be on the safe side, I
1483 have increased the buffer size for the ${rfc2047: expansion operator from
1486 33. It is now permitted to omit both strings after an "if" condition; if the
1487 condition is true, the result is "true". As before, when the second string
1488 is omitted, a false condition yields an empty string. This makes it less
1489 cumbersome to write custom ACL and router conditions.
1491 34. Failure to deliver a bounce message always caused it to be frozen, even if
1492 there was an errors_to setting on the router. The errors_to setting is now
1495 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1496 canonical form (fully expanded) before being placed in
1497 $sender_host_address.
1499 36. The table in the code that translates DNS record types into text (T_A to
1500 "A" for instance) was missing entries for NS and CNAME. It is just possible
1501 that this could have caused confusion if both these types were looked up
1502 for the same domain, because the text type is used as part of Exim's
1503 per-process caching. But the chance of anyone hitting this buglet seems
1506 37. The dnsdb lookup has been extended in a number of ways.
1508 (1) There is a new type, "zns", which walks up the domain tree until it
1509 finds some nameserver records. It should be used with care.
1511 (2) There is a new type, "mxh", which is like "mx" except that it returns
1512 just the host names, not the priorities.
1514 (3) It is now possible to give a list of domains (or IP addresses) to be
1515 looked up. The behaviour when one of the lookups defers can be
1516 controlled by a keyword.
1518 (4) It is now possible to specify the separator character for use when
1519 multiple records are returned.
1521 38. The dnslists ACL condition has been extended: it is now possible to supply
1522 a list of IP addresses and/or domains to be looked up in a particular DNS
1525 39. Added log_selector=+queue_time_overall.
1527 40. When running the queue in the test harness, wait just a tad after forking a
1528 delivery process, to get repeatability of debugging output.
1530 41. Include certificate and key file names in error message when GnuTLS fails
1531 to set them up, because the GnuTLS error message doesn't include the name
1532 of the failing file when there is a problem reading it.
1534 42. Allow both -bf and -bF in the same test run.
1536 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1538 44. The "Exiscan patch" is now merged into the mainline Exim source.
1540 45. Sometimes the final signoff response after QUIT could fail to get
1541 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1542 < 0 before doing a fflush(). This bug looks as though it goes back to the
1543 introduction of TLS in release 3.20, but "sometimes" must have been rare
1544 because the tests only now provoked it.
1546 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1547 (this can affect the format of dates).
1549 47. exim_tidydb, when checking for the continued existence of a message for
1550 which it has found a message-specific retry record, was not finding
1551 messages that were in split spool directories. Consequently, it was
1552 deleting retry records that should have stayed in existence.
1554 48. Steve fixed some bugs in eximstats.
1556 49. The SPA authentication driver was not abandoning authentication and moving
1557 on to the next authenticator when an expansion was forced to fail,
1558 contradicting the general specification for all authenticators. Instead it
1559 was generating a temporary error. It now behaves as specified.
1561 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1562 (the order specifies the preference for clients). The order is now AES256,
1563 AES128, 3DES, ARCFOUR128.
1565 51. Small patch to Sieve code - explicitly set From: when generating an
1568 52. Exim crashed if a remote delivery caused a very long error message to be
1569 recorded - for instance if somebody sent an entire SpamAssassin report back
1570 as a large number of 550 error lines. This bug was coincidentally fixed by
1571 increasing the size of one of Exim's internal buffers (big_buffer) that
1572 happened as part of the Exiscan merge. However, to be on the safe side, I
1573 have made the code more robust (and fixed the comments that describe what
1576 53. Now that there can be additional text after "Completed" in log lines (if
1577 the queue_time_overall log selector is set), a one-byte patch to exigrep
1578 was needed to allow it to recognize "Completed" as not the last thing in
1581 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1582 patch that reportedly fixes this has been added. I am not expert enough to
1583 create a test for it. This is what the patch creator wrote:
1585 "I found a little strange behaviour of ldap code when working with
1586 Windows 2003 AD Domain, where users was placed in more than one
1587 Organization Units. When I tried to give exim partial DN, the exit code
1588 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1589 But simultaneously result of request was absolutely normal ldap result,
1590 so I produce this patch..."
1592 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1593 REFERENCE, so I have modified the code to exclude the patch when that macro
1596 55. Some experimental protocols are using DNS PTR records for new purposes. The
1597 keys for these records are domain names, not reversed IP addresses. The
1598 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1599 leaves it alone. Component reversal etc. now happens only for IP addresses.
1602 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1604 57. Double the size of the debug message buffer (to 2048) so that more of very
1605 long debug lines gets shown.
1607 58. The exicyclog utility now does better if the number of log files to keep
1608 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1610 59. Two changes related to the smtp_active_hostname option:
1612 (1) $smtp_active_hostname is now available as a variable.
1613 (2) The default for smtp_banner uses $smtp_active_hostname instead
1614 of $primary_hostname.
1616 60. The host_aton() function is supposed to be passed a string that is known
1617 to be a valid IP address. However, in the case of IPv6 addresses, it was
1618 not checking this. This is a hostage to fortune. Exim now panics and dies
1619 if the condition is not met. A case was found where this could be provoked
1620 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1621 components; fortuitously, this particular loophole had already been fixed
1622 by change 4.50/55 above.
1624 If there are any other similar loopholes, the new check in host_aton()
1625 itself should stop them being exploited. The report I received stated that
1626 data on the command line could provoke the exploit when Exim was running as
1627 exim, but did not say which command line option was involved. All I could
1628 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1629 running as the user.
1632 61. There was a buffer overflow vulnerability in the SPA authentication code
1633 (which came originally from the Samba project). I have added a test to the
1634 spa_base64_to_bits() function which I hope fixes it.
1637 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1638 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1640 63. The daemon start-up calls getloadavg() while still root for those OS that
1641 need the first call to be done as root, but it missed one case: when
1642 deliver_queue_load_max is set with deliver_drop_privilege. This is
1643 necessary for the benefit of the queue runner, because there is no re-exec
1644 when deliver_drop_privilege is set.
1646 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1647 This has been fixed.
1649 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1650 was in use, was not putting the data itself into the right store pool;
1651 consequently, it could be overwritten for a subsequent message in the same
1652 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1655 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1657 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1658 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1659 the function were treating the return as a boolean value, which happened to
1660 work because 0=false and not-0=true, but is not correct code.
1662 68. The host_aton() function was not handling scoped IPv6 addresses (those
1663 with, for example, "%eth0" on the end) correctly.
1665 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1666 negated items (that is, ~something) in unsigned ints. Some compilers
1667 apparently mutter when there is no cast.
1669 70. If an address verification called from an ACL failed, and did not produce a
1670 user-specific message (i.e. there was only a "system" message), nothing was
1671 put in $acl_verify_message. In this situation, it now puts the system
1674 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1675 session; change 4.31/43 added the unwanted input to the log line - except
1676 that it did not do this in the start of session case. It now does.
1678 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1679 This is wrong and can cause the other end to generate a synchronization
1680 error if it is another Exim or anything else that does the synchronization
1681 check. A QUIT command is no longer sent after a timeout.
1683 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1684 during host lookups.
1686 74. The defer_ok option of callout verification was not working if it was used
1687 when verifying addresses in header lines, that is, for this case:
1689 verify = header_sender/callout=defer_ok
1691 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1692 those file descriptors could be used for SMTP connections. If anything
1693 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1694 could be sent to the SMTP client, causing chaos. The daemon now opens
1695 stdin, stdout, and stderr to /dev/null when it puts itself into the
1698 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1699 log by default. The user can override this with suitable Perl magic.
1701 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1702 the log message when discard triggers, was not working for the DATA ACL or
1703 for the non-SMTP ACL.
1705 78. Error message wording change in sieve.c.
1707 79. If smtp_accept_max_per_host was set, the number of connections could be
1708 restricted to fewer than expected, because the daemon was trying to set up
1709 a new connection before checking whether the processes handling previous
1710 connections had finished. The check for completed processes is now done
1711 earlier. On busy systems, this bug wouldn't be noticed because something
1712 else would have woken the daemon, and it would have reaped the completed
1715 80. If a message was submitted locally by a user whose login name contained one
1716 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1717 It caused a spool format error. I have fixed the spool reading code. A
1718 related problem was that the "from" clause in the Received: line became
1719 illegal because of the space(s). It is now covered by ${quote_local_part.
1721 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1724 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1726 83. Patch from Sieve maintainer to fix unterminated string problem in
1727 "vacation" handling.
1729 84. Some minor changes to the Linux configuration files to help with other
1730 OS variants using glibc.
1732 85. One more patch for Sieve to update vacation handling to latest spec.
1735 ----------------------------------------------------
1736 See the note above about the 4.44 and 4.50 releases.
1737 ----------------------------------------------------
1743 1. Change 4.43/35 introduced a bug that caused file counts to be
1744 incorrectly computed when quota_filecount was set in an appendfile
1747 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1748 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1751 3. Give more explanation in the error message when the command for a transport
1752 filter fails to execute.
1754 4. There are several places where Exim runs a non-Exim command in a
1755 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1756 was being done only for the command run by the queryprogram router. It is
1757 now done for all such subprocesses. The other cases are: ${run, transport
1758 filters, and the commands run by the lmtp and pipe transports.
1760 5. Some older OS have a limit of 256 on the maximum number of file
1761 descriptors. Exim was using setrlimit() to set 1000 as a large value
1762 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1763 systems. I've change it so that if it can't get 1000, it tries for 256.
1765 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1766 was an oversight, and furthermore, ever since the addition of extra
1767 controls (e.g. 4.43/32), the checks on when to allow different forms of
1768 "control" were broken. There should now be diagnostics for all cases when a
1769 control that does not make sense is encountered.
1771 7. $recipients is now available in the predata ACL (oversight).
1773 8. Tidy the search cache before the fork to do a delivery from a message
1774 received from the command line. Otherwise the child will trigger a lookup
1775 failure and thereby defer the delivery if it tries to use (for example) a
1776 cached ldap connection that the parent has called unbind on.
1778 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1779 of $address_data from the recipient verification was clobbered by the
1780 sender verification.
1782 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1783 was its contents. (It was OK if the option was not defined at all.)
1785 11. A "Completed" log line is now written for messages that are removed from
1786 the spool by the -Mrm option.
1788 12. $host_address is now set to the target address during the checking of
1789 ignore_target_hosts.
1791 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1792 being passed; this would have caused $sender_host_name to have been used if
1793 matching the list had actually called for a host name (not very likely,
1794 since this list is usually IP addresses). A host name is now passed as
1797 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1798 code that specifies a non-restarting handler (typically sigaction() in
1799 modern systems) in an attempt to fix a rare and obscure crash bug.
1801 15. Narrowed the window for a race in the daemon that could cause it to ignore
1802 SIGCHLD signals. This is not a major problem, because they are used only to
1803 wake it up if nothing else does.
1805 16. A malformed maildirsize file could cause Exim to calculate negative values
1806 for the mailbox size or file count. Odd effects could occur as a result.
1807 The maildirsize information is now recalculated if the size or filecount
1810 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1811 support for a long time. Removed HAVE_SYS_VFS_H.
1813 18. Updated exipick to current release from John Jetmore.
1815 19. Allow an empty sender to be matched against a lookup in an address list.
1816 Previously the only cases considered were a regular expression, or an
1819 20. Exim went into a mad DNS lookup loop when doing a callout where the
1820 host was specified on the transport, if the DNS lookup yielded more than
1823 21. The RFC2047 encoding function was originally intended for short strings
1824 such as real names; it was not keeping to the 75-character limit for
1825 encoded words that the RFC imposes. It now respects the limit, and
1826 generates multiple encoded words if necessary. To be on the safe side, I
1827 have increased the buffer size for the ${rfc2047: expansion operator from
1830 22. Failure to deliver a bounce message always caused it to be frozen, even if
1831 there was an errors_to setting on the router. The errors_to setting is now
1834 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1835 canonical form (fully expanded) before being placed in
1836 $sender_host_address.
1838 24. Updated eximstats to version 1.33
1840 25. Include certificate and key file names in error message when GnuTLS fails
1841 to set them up, because the GnuTLS error message doesn't include the name
1842 of the failing file when there is a problem reading it.
1844 26. Expand error message when OpenSSL has problems setting up cert/key files.
1847 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1848 (this can affect the format of dates).
1850 28. exim_tidydb, when checking for the continued existence of a message for
1851 which it has found a message-specific retry record, was not finding
1852 messages that were in split spool directories. Consequently, it was
1853 deleting retry records that should have stayed in existence.
1855 29. eximstats updated to version 1.35
1856 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1857 1.35 - bugfix such that pie charts by volume are generated correctly
1859 30. The SPA authentication driver was not abandoning authentication and moving
1860 on to the next authenticator when an expansion was forced to fail,
1861 contradicting the general specification for all authenticators. Instead it
1862 was generating a temporary error. It now behaves as specified.
1864 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1865 (the order specifies the preference for clients). The order is now AES256,
1866 AES128, 3DES, ARCFOUR128.
1868 31. Small patch to Sieve code - explicitly set From: when generating an
1871 32. Exim crashed if a remote delivery caused a very long error message to be
1872 recorded - for instance if somebody sent an entire SpamAssassin report back
1873 as a large number of 550 error lines. This bug was coincidentally fixed by
1874 increasing the size of one of Exim's internal buffers (big_buffer) that
1875 happened as part of the Exiscan merge. However, to be on the safe side, I
1876 have made the code more robust (and fixed the comments that describe what
1879 33. Some experimental protocols are using DNS PTR records for new purposes. The
1880 keys for these records are domain names, not reversed IP addresses. The
1881 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1882 leaves it alone. Component reversal etc. now happens only for IP addresses.
1885 34. The host_aton() function is supposed to be passed a string that is known
1886 to be a valid IP address. However, in the case of IPv6 addresses, it was
1887 not checking this. This is a hostage to fortune. Exim now panics and dies
1888 if the condition is not met. A case was found where this could be provoked
1889 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1890 components; fortuitously, this particular loophole had already been fixed
1891 by change 4.50/55 or 4.44/33 above.
1893 If there are any other similar loopholes, the new check in host_aton()
1894 itself should stop them being exploited. The report I received stated that
1895 data on the command line could provoke the exploit when Exim was running as
1896 exim, but did not say which command line option was involved. All I could
1897 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1898 running as the user.
1901 35. There was a buffer overflow vulnerability in the SPA authentication code
1902 (which came originally from the Samba project). I have added a test to the
1903 spa_base64_to_bits() function which I hope fixes it.
1906 36. The daemon start-up calls getloadavg() while still root for those OS that
1907 need the first call to be done as root, but it missed one case: when
1908 deliver_queue_load_max is set with deliver_drop_privilege. This is
1909 necessary for the benefit of the queue runner, because there is no re-exec
1910 when deliver_drop_privilege is set.
1912 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1913 was in use, was not putting the data itself into the right store pool;
1914 consequently, it could be overwritten for a subsequent message in the same
1915 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1918 38. Sometimes the final signoff response after QUIT could fail to get
1919 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1920 < 0 before doing a fflush(). This bug looks as though it goes back to the
1921 introduction of TLS in release 3.20, but "sometimes" must have been rare
1922 because the tests only now provoked it.
1928 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1929 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1930 Now it returns an integer. A number of calls were still expecting a T/F
1931 return. Fortuitously, in all cases, the tests worked in OK situations,
1932 which is the norm. However, things would have gone wrong on any write
1933 failures on the smtp file descriptor. This function is used when sending
1934 messages over SMTP and also when doing verify callouts.
1936 2. When Exim is called to do synchronous delivery of a locally submitted
1937 message (the -odf or -odi options), it no longer closes stderr before doing
1940 3. Implemented the mua_wrapper option.
1942 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1944 5. Implemented the functions header_remove(), header_testname(),
1945 header_add_at_position(), and receive_remove_recipient(), and exported them
1948 6. If an ACL "warn" statement specified the addition of headers, Exim already
1949 inserted X-ACL-Warn: at the start if there was no header name. However, it
1950 was not making this test for the second and subsequent header lines if
1951 there were newlines in the string. This meant that an invalid header could
1952 be inserted if Exim was badly configured.
1954 7. Allow an ACL "warn" statement to add header lines at the start or after all
1955 the Received: headers, as well as at the end.
1957 8. Added the rcpt_4xx retry error code.
1959 9. Added postmaster_mailfrom=xxx to callout verification option.
1961 10. Added mailfrom=xxxx to the callout verification option, for verify=
1964 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1965 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1967 12. Inserted some casts to stop certain compilers warning when using pointer
1968 differences as field lengths or precisions in printf-type calls (mostly
1969 affecting debugging statements).
1971 13. Added optional readline() support for -be (dynamically loaded).
1973 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1974 same clock tick as a message's arrival, so that its received time was the
1975 same as the "first fail" time on the retry record, and that message
1976 remained on the queue past the ultimate address timeout, every queue runner
1977 would try a delivery (because it was past the ultimate address timeout) but
1978 after another failure, the ultimate address timeout, which should have then
1979 bounced the address, did not kick in. This was a "< instead of <=" error;
1980 in most cases the first failure would have been in the next clock tick
1981 after the received time, and all would be well.
1983 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1984 being recognized when the domain list was tested by the match_domain
1985 condition in an expansion string.
1987 16. Added the ${str2b64: operator.
1989 17. Exim was always calling setrlimit() to set a large limit for the number of
1990 processes, without checking whether the existing limit was already
1991 adequate. (It did check for the limit on file descriptors.) Furthermore,
1992 errors from getrlimit() and setrlimit() were being ignored. Now they are
1993 logged to the main and panic logs, but Exim does carry on, to try to do its
1994 job under whatever limits there are.
1996 18. Imported PCRE 5.0.
1998 19. Trivial typo in log message " temporarily refused connection" (the leading
2001 20. If the log selector return_path_on_delivery was set and an address was
2002 redirected to /dev/null, the delivery process crashed because it assumed
2003 that a return path would always be set for a "successful" delivery. In this
2004 case, the whole delivery is bypassed as an optimization, and therefore no
2007 21. Internal re-arrangement: the function for sending a challenge and reading
2008 a response while authentication was assuming a zero-terminated challenge
2009 string. It's now changed to take a pointer and a length, to allow for
2010 binary data in such strings.
2012 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2014 23. Exim was not respecting finduser_retries when seeking the login of the
2015 uid under which it was called; it was always trying 10 times. (The default
2016 setting of finduser_retries is zero.) Also, it was sleeping after the final
2017 failure, which is pointless.
2019 24. Implemented tls_on_connect_ports.
2021 25. Implemented acl_smtp_predata.
2023 26. If the domain in control=submission is set empty, Exim assumes that the
2024 authenticated id is a complete email address when it generates From: or
2025 Sender: header lines.
2027 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2028 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2029 chown and chgrp in /bin and hostname in /usr/bin.
2031 28. Exim was keeping the "process log" file open after each use, just as it
2032 does for the main log. This opens the possibility of it remaining open for
2033 long periods when the USR1 signal hits a daemon. Occasional processlog
2034 errors were reported, that could have been caused by this. Anyway, it seems
2035 much more sensible not to leave this file open at all, so that is what now
2038 29. The long-running daemon process does not normally write to the log once it
2039 has entered its main loop, and it closes the log before doing so. This is
2040 so that log files can straightforwardly be renamed and moved. However,
2041 there are a couple of unusual error situations where the daemon does write
2042 log entries, and I had neglected to close the log afterwards.
2044 30. The text of an SMTP error response that was received during a remote
2045 delivery was being truncated at 512 bytes. This is too short for some of
2046 the long messages that one sometimes sees. I've increased the limit to
2049 31. It is now possible to make retry rules that apply only when a message has a
2050 specific sender, in particular, an empty sender.
2052 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2053 it possible to be selective about when SMTP synchronization is enforced.
2055 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2057 32. Implemented hosts_connection_nolog.
2059 33. Added an ACL for QUIT.
2061 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2064 35. Added mailbox_size and mailbox_filecount to appendfile.
2066 36. Added control = no_multiline_responses to ACLs.
2068 37. There was a bug in the logic of the code that waits for the clock to tick
2069 in the case where the clock went backwards by a substantial amount such
2070 that the microsecond fraction of "now" was more than the microsecond
2071 fraction of "then" (but the whole seconds number was less).
2073 38. Added support for the libradius Radius client library this is found on
2074 FreeBSD (previously only the radiusclient library was supported).
2080 1. When certain lookups returned multiple values in the form name=value, the
2081 quoting of the values was not always being done properly. Specifically:
2082 (a) If the value started with a double quote, but contained no whitespace,
2084 (b) If the value contained whitespace other than a space character (i.e.
2085 tabs or newlines or carriage returns) it was not quoted.
2086 This fix has been applied to the mysql and pgsql lookups by writing a
2087 separate quoting function and calling it from the lookup code. The fix
2088 should probably also be applied to nisplus, ibase and oracle lookups, but
2089 since I cannot test any of those, I have not disturbed their existing code.
2091 2. A hit in the callout cache for a specific address caused a log line with no
2092 reason for rejecting RCPT. Now it says "Previous (cached) callout
2093 verification failure".
2095 3. There was an off-by-one bug in the queryprogram router. An over-long
2096 return line was truncated at 256 instead of 255 characters, thereby
2097 overflowing its buffer with the terminating zero. As well as fixing this, I
2098 have increased the buffer size to 1024 (and made a note to document this).
2100 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2101 when Exim is waiting for an SMTP response from a remote server, Exim
2102 restarts its select() call on the socket, thereby resetting its timeout.
2103 This is not a problem when such interrupts are rare. Somebody set up a cron
2104 job to run exiwhat every 2 minutes, which is less than the normal select()
2105 timeout (5 or 10 minutes). This meant that the select() timeout never
2106 kicked in because it was always reset. I have fixed this by comparing the
2107 time when an interrupt arrives with the time at the start of the first call
2108 to select(). If more time than the timeout has elapsed, the interrupt is
2109 treated as a timeout.
2111 5. Some internal re-factoring in preparation for the addition of Sieve
2112 extensions (by MH). In particular, the "personal" test is moved to a
2113 separate function, and given an option for scanning Cc: and Bcc: (which is
2114 not set for Exim filters).
2116 6. When Exim created an email address using the login of the caller as the
2117 local part (e.g. when creating a From: or Sender: header line), it was not
2118 quoting the local part when it contained special characters such as @.
2120 7. Installed new OpenBSD configuration files.
2122 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2123 try to make them clearer.
2125 9. Callout options, other than the timeout value, were being ignored when
2126 verifying sender addresses in header lines. For example, when using
2128 verify = header_sender/callout=no_cache
2130 the cache was (incorrectly) being used.
2132 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2133 only the Cygwin environment.
2135 11. When return_path_on_delivery was set as a log selector, if different remote
2136 addresses in the same message used different return paths and parallel
2137 remote delivery occurred, the wrong values would sometimes be logged.
2138 (Whenever a remote delivery process finished, the return path value from
2139 the most recently started remote delivery process was logged.)
2141 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2142 lines when AUTH and/or TLS are in use. This is the "received protocol"
2143 field. Exim used to use "asmtp" for authenticated SMTP, without any
2144 indication (in the protocol name) for TLS use. Now it follows the RFC and
2145 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2146 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2147 names appear in log lines as well as in Received: header lines.
2149 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2150 extensions, and comparison tests, and to fix some bugs.
2152 14. Changes to the "personal" filter test:
2154 (1) The test was buggy in that it was just doing the equivalent of
2155 "contains" tests on header lines. For example, if a user's address was
2156 anne@some.where, the "personal" test would incorrectly be true for
2158 To: susanne@some.where
2160 This test is now done by extracting each address from the header in turn,
2161 and checking the entire address. Other tests that are part of "personal"
2162 are now done using regular expressions (for example, to check local parts
2163 of addresses in From: header lines).
2165 (2) The list of non-personal local parts in From: addresses has been
2166 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2167 taken from the Sieve specification recommendations.
2169 (3) If the message contains any header line starting with "List-" it is
2170 treated as non-personal.
2172 (4) The test for "circular" in the Subject: header line has been removed
2173 because it now seems ill-conceived.
2175 15. Minor typos in src/EDITME comments corrected.
2177 16. Installed latest exipick from John Jetmore.
2179 17. If headers_add on a router specified a text string that was too long for
2180 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2181 of string_sprintf() is now avoided.
2183 18. $message_body_size was not set (it was always zero) when running the DATA
2184 ACL and the local_scan() function.
2186 19. For the "mail" command in an Exim filter, no default was being set for
2187 the once_repeat time, causing a random time value to be used if "once" was
2188 specified. (If the value happened to be <= 0, no repeat happened.) The
2189 default is now 0s, meaning "never repeat". The "vacation" command was OK
2190 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2191 (I found it when inspecting the code).
2193 20. There is now an overall timeout for performing a callout verification. It
2194 defaults to 4 times the callout timeout, which applies to individual SMTP
2195 commands during the callout. The overall timeout applies when there is more
2196 than one host that can be tried. The timeout is checked before trying the
2197 next host. This prevents very long delays if there are a large number of
2198 hosts and all are timing out (e.g. when the network connections are timing
2199 out). The value of the overall timeout can be changed by specifying an
2200 additional sub-option for "callout", called "maxwait". For example:
2202 verify = sender/callout=5s,maxwait=20s
2204 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2205 to the end before writing, but this should make it even safer).
2207 22. Exim was forgetting that it had advertised PIPELINING for the second and
2208 subsequent messages on an SMTP connection. It was also not resetting its
2209 memory on STARTTLS and an internal HELO.
2211 23. When Exim logs an SMTP synchronization error within a session, it now
2212 records whether PIPELINING has been advertised or not.
2214 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2215 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2216 rather than long int.
2218 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2220 26. Added the never_mail option to autoreply.
2226 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2227 crash if the getsockname() call failed; this can happen if a connection is
2228 closed very soon after it is established. The problem was simply in the
2229 order in which certain operations were done, causing Exim to try to write
2230 to the SMTP stream before it had set up the file descriptor. The bug has
2231 been fixed by making things happen in the correct order.
2237 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2238 before the connection was closed, thus losing the rejection response.
2240 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2241 some early Solaris releases, but causes trouble in current releases where
2242 socklen_t is defined.
2244 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2247 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2250 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2251 that releasing the top of it at the end releases what was used for sub-
2252 expansions (unless the block got too big). However, discard this block if
2253 the first thing is a variable or header, so that we can use its block when
2254 it is dynamic (useful for very large $message_headers, for example).
2256 6. Lookups now cache *every* query, not just the most recent. A new, separate
2257 store pool is used for this. It can be recovered when all lookup caches are
2258 flushed. Lookups now release memory at the end of their result strings.
2259 This has involved some general refactoring of the lookup sources.
2261 7. Some code has been added to the store_xxx() functions to reduce the amount
2262 of flapping under certain conditions.
2264 8. log_incoming_interface used to affect only the <= reception log lines. Now
2265 it causes the local interface and port to be added to several more SMTP log
2266 lines, for example "SMTP connection from", and rejection lines.
2268 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2270 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2272 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2273 could be overwritten at the end of the current message (or the start of a
2274 new message if it was set in a HELO ACL). The value is now preserved for
2275 the duration of the SMTP connection.
2277 12. If a transport had a headers_rewrite setting, and a matching header line
2278 contained an unqualified address, that address was qualified, even if it
2279 did not match any rewriting rules. The underlying bug was that the values
2280 of the flags that permit the existence of unqualified sender and recipient
2281 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2282 non-local messages, and by -bnq for local messages) were not being
2283 preserved with the message after it was received.
2285 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2286 "next input=" as part of the text comprising the host identity instead of
2287 the correct text. The code was using the same buffer for two different
2288 strings. However, depending on which order the printing function evaluated
2289 its arguments, the bug did not always show up. Under Linux, for example, my
2290 test suite worked just fine.
2292 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2293 doesn't work with some older versions of Perl. It has been changed to "my",
2294 which in any case is probably the better facility to use.
2296 15. A really picky compiler found some instances of statements for creating
2297 error messages that either had too many or two few arguments for the format
2300 16. The size of the buffer for calls to the DNS resolver has been increased
2301 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2302 for addresses that have a lot of PTR records. This alleviates a problem; it
2303 does not fully solve it.
2305 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2306 buffer now truncates the list and logs the incident, which is the same
2307 action as happens when Exim is looking up a host name and its aliases.
2308 Previously in this situation something unpredictable would happen;
2309 sometimes it was "internal error: store_reset failed".
2311 18. If a server dropped the connection unexpectedly when an Exim client was
2312 using GnuTLS and trying to read a response, the client delivery process
2313 crashed while trying to generate an error log message.
2315 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2316 string, for example:
2318 warn message = H1: something\nH2: something
2320 the text was added as a single header line from Exim's point of view
2321 though it ended up OK in the delivered message. However, searching for the
2322 second and subsequent header lines using $h_h2: did not work. This has been
2323 fixed. Similarly, if a system filter added multiple headers in this way,
2324 the routers could not see them.
2326 20. Expanded the error message when iplsearch is called with an invalid key to
2327 suggest using net-iplsearch in a host list.
2329 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2330 ACLs are no longer actually imposed (and a message to that effect is
2333 22. If a "gecos" field in a passwd entry contained escaped characters, in
2334 particular, if it contained a \" sequence, Exim got it wrong when building
2335 a From: or a Sender: header from that name. A second bug also caused
2336 incorrect handling when an unquoted " was present following a character
2337 that needed quoting.
2339 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2340 was not being matched caselessly.
2342 24. Arranged for all hyphens in the exim.8 source to be escaped with
2345 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2346 properly. Recipient callout cache records were still being keyed to include
2347 the sender, even when use_sender was set false. This led to far more
2348 callouts that were necessary. The sender is no longer included in the key
2349 when use_sender is false.
2351 26. Added "control = submission" modifier to ACLs.
2353 27. Added the ${base62d: operator to decode base 62 numbers.
2355 28. dnsdb lookups can now access SRV records.
2357 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2358 the configuration file.
2360 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2361 (-v) mode. This makes the output for a verbose queue run more intelligible.
2363 31. Added a use_postmaster feature to recipient callouts.
2365 32. Added the $body_zerocount variable, containing the number of binary zero
2366 bytes in the message body.
2368 33. The time of last modification of the "new" subdirectory is now used as the
2369 "mailbox time last read" when there is a quota error for a maildir
2372 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2374 35. Added +ignore_unknown as a special item in host lists.
2376 36. Code for decoding IPv6 addresses in host lists is now included, even if
2377 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2378 address was recognized as an IP address, but was then not correctly decoded
2379 into binary, causing unexpected and incorrect effects when compared with
2386 1. Very minor rewording of debugging text in manualroute to say "list of
2387 hosts" instead of "hostlist".
2389 2. If verify=header_syntax was set, and a header line with an unqualified
2390 address (no domain) and a large number of spaces between the end of the
2391 name and the colon was received, the reception process suffered a buffer
2392 overflow, and (when I tested it) crashed. This was caused by some obsolete
2393 code that should have been removed. The fix is to remove it!
2395 3. When running in the test harness, delay a bit after writing a bounce
2396 message to get a bit more predictability in the log output.
2398 4. Added a call to search_tidyup() just before forking a reception process. In
2399 theory, someone could use a lookup in the expansion of smtp_accept_max_
2400 per_host which, without the tidyup, could leave open a database connection.
2402 5. Added the variables $recipient_data and $sender_data which get set from a
2403 lookup success in an ACL "recipients" or "senders" condition, or a router
2404 "senders" option, similar to $domain_data and $local_part_data.
2406 6. Moved the writing of debug_print from before to after the "senders" test
2409 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2410 problems for message scanning, either using a data ACL, or using
2411 local_scan() because the Received: header was not generated till after they
2412 were called (in order to set the time as the time of reception completion).
2413 I have revised the way this works. The header is now generated after the
2414 body is received, but before the ACL or local_scan() are called. After they
2415 are run, the timestamp in the header is updated.
2421 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2422 before starting a queue runner without re-exec. This happened only when
2423 deliver_drop_privilege was set or when the Exim user was set to root. The
2424 effect of the bug was that timeouts during subsequent deliveries caused
2425 crashes instead of being properly handled. The handler is now left at its
2426 default (and expected) setting.
2428 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2429 message, again when deliver_drop_privilege is set or Exim is run as root.
2430 The bug described in (1) was not present in this case, but the tidying up
2431 of the other signals was missing. I have made the two cases consistent.
2433 3. The ignore_target_hosts setting on a manualroute router was being ignored
2434 for hosts that were looked up using the /MX notation.
2436 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2439 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2440 operated on the sender address. After changing the $sender_address to <>
2441 for the sender address verify, Exim was re-instated it as the original
2442 (before rewriting) address, but remembering that it had rewritten it, so it
2443 wasn't rewriting it again. This bug also had the effect of breaking the
2444 sender address verification caching when the sender address was rewritten.
2446 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2447 This has been changed so that if the ip literal address matches
2448 ignore_target_hosts, the router declines.
2450 7. Added expansion conditions match_domain, match_address, and match_local_
2451 part (NOT match_host).
2453 8. The placeholder for the Received: header didn't have a length field set.
2455 9. Added code to Exim itself and to exim_lock to test for a specific race
2456 condition that could lead to file corruption when using MBX delivery. The
2457 issue is with the lockfile that is created in /tmp. If this file is removed
2458 after a process has opened it but before that process has acquired a lock,
2459 there is the potential for a second process to recreate the file and also
2460 acquire a lock. This could lead to two Exim processes writing to the file
2461 at the same time. The added code performs the same test as UW imapd; it
2462 checks after acquiring the lock that its file descriptor still refers to
2463 the same named file.
2465 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2466 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2469 11. Added the smtp_active_hostname option. If used, this will typically be made
2470 to depend on the incoming interface address. Because $interface_address is
2471 not set up until the daemon has forked a reception process, error responses
2472 that can happen earlier (such as "too many connections") no longer contain
2475 12. If an expansion in a condition on a "warn" statement fails because a lookup
2476 defers, the "warn" statement is abandoned, and the next ACL statement is
2477 processed. Previously this caused the whole ACL to be aborted.
2479 13. Added the iplsearch lookup type.
2481 14. Added ident_timeout as a log selector.
2483 15. Added tls_certificate_verified as a log selector.
2485 16. Added a global option tls_require_ciphers (compare the smtp transport
2486 option of the same name). This controls incoming TLS connections.
2488 17. I finally figured out how to make tls_require_ciphers do a similar thing
2489 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2490 before starting the TLS session.
2492 18. Tabs are now shown as \t in -bP output.
2494 19. If the log selector return_path_on_delivery was set, Exim crashed when
2495 bouncing a message because it had too many Received: header lines.
2497 20. If two routers both had headers_remove settings, and the first one included
2498 a superfluous trailing colon, the final name in the first list and the
2499 first name in the second list were incorrectly joined into one item (with a
2500 colon in the middle).
2506 1. Added -C and -D options to the exinext utility, mainly to make it easier
2507 to include in the automated testing, but these could be helpful when
2508 multiple configurations are in use.
2510 2. The exinext utility was not formatting the output nicely when there was
2511 an alternate port involved in the retry record key, nor when there was a
2512 message id as well (for retries that were specific to a specific message
2513 and a specific host). It was also confused by IPv6 addresses, because of
2514 the additional colons they contain. I have fixed the IPv4 problem, and
2515 patched it up to do a reasonable job for IPv6.
2517 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2518 delivery, the log line now contains "pipelined" if PIPELINING was used.
2520 4. An SMTP transport process used to panic and die if the bind() call to set
2521 an explicit outgoing interface failed. This has been changed; it is now
2522 treated in the same way as a connect() failure.
2524 5. A reference to $sender_host_name in the part of a conditional expansion
2525 that was being skipped was still causing a DNS lookup. This no longer
2528 6. The def: expansion condition was not recognizing references to header lines
2529 that used bh_ and bheader_.
2531 7. Added the _cache feature to named lists.
2533 8. The code for checking quota_filecount in the appendfile transport was
2534 allowing one more file than it should have been.
2536 9. For compatibility with Sendmail, the command line option
2544 and sets the incoming protocol and host name (for trusted callers). The
2545 host name and its colon can be omitted when only the protocol is to be set.
2546 Note the Exim already has two private options, -pd and -ps, that refer to
2547 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2548 "s", but I don't think that's a major issue.
2550 10. A number of refactoring changes to the code, none of which should affect
2553 (a) The number of logging options was getting close to filling up the
2554 32-bit word that was used as a bit map. I have split them into two classes:
2555 those that are passed in the argument to log_write(), and those that are
2556 only ever tested independently outside of that function. These are now in
2557 separate 32-bit words, so there is plenty of room for expansion again.
2558 There is no change in the user interface or the logging behaviour.
2560 (b) When building, for example, log lines, the code previously used a
2561 macro that called string_cat() twice, in order to add two strings. This is
2562 not really sufficiently general. Furthermore, there was one instance where
2563 it was actually wrong because one of the argument was used twice, and in
2564 one call a function was used. (As it happened, calling the function twice
2565 did not affect the overall behaviour.) The macro has been replaced by a
2566 function that can join an arbitrary number of extra strings onto a growing
2569 (c) The code for expansion conditions now uses a table and a binary chop
2570 instead of a serial search (which was left over from when there were very
2571 few conditions). Also, it now recognizes conditions like "pam" even when
2572 the relevant support is not compiled in: a suitably worded error message is
2573 given if an attempt is made to use such a condition.
2575 11. Added ${time_interval:xxxxx}.
2577 12. A bug was causing one of the ddress fields not to be passed back correctly
2578 from remote delivery subprocesses. The field in question was not being
2579 subsequently used, so this caused to problems in practice.
2581 13. Added new log selectors queue_time and deliver_time.
2583 14. Might have fixed a bug in maildirsizefile handling that threw up
2584 "unexpected character" debug warnings, and recalculated the data
2585 unnecessarily. In any case, I expanded the warning message to give more
2588 15. Added the message "Restricted characters in address" to the statements in
2589 the default ACL that block characters like @ and % in local parts.
2591 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2592 Three changes have been made:
2594 (a) There was a serious bug; a negative response to MAIL caused the whole
2595 recipient domain to be cached as invalid, thereby blocking all messages
2596 to all local parts at the same domain, from all senders. This bug has
2597 been fixed. The domain is no longer cached after a negative response to
2598 MAIL if the sender used is not empty.
2600 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2603 (c) A new callout option, "use_sender" has been added for people who want
2604 the modified behaviour.
2610 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2613 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2614 indeed breaks things for older releases.
2616 3. Added additional logging to the case where there is a problem reading data
2617 from a filter that is running in a subprocess using a pipe, in order to
2618 try to track down a specific problem.
2620 4. Testing facility fudge: when running in the test harness and attempting
2621 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2622 getting "No route to host". Convert this to a timeout.
2624 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2627 6. Some OS don't have socklen_t but use size_t instead. This affects the
2628 fifth argument of getsockopt() amongst other things. This is now
2629 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2630 can be set for individual OS. I have set it for SunOS5, OSF1, and
2631 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2632 some earlier ones do not.
2634 7. Change 4.30/15 was not doing the test caselessly.
2636 8. The standard form for an IPv6 address literal was being rejected by address
2637 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2638 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2639 this, as well as the form without the "IPv6" on the front (but only when
2640 address literals are enabled, of course).
2642 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2644 10. Exim crashed if a message with an empty sender address specified by -f
2645 encountered a router with an errors_to setting. This could be provoked only
2646 by a command such as
2650 where an empty string was supplied; "<>" did not hit this bug.
2652 11. Installed PCRE release 4.5.
2654 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2655 remained set. It is now erased.
2657 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2658 times from message ids (which are base 36 rather than the normal 62).
2660 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2661 were being counted as actual protocol errors, and logged if the log
2662 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2663 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2664 and DATA following a set of rejected RCPTs do not count as protocol errors.
2665 In other words, Exim assumes they were pipelined, though this may not
2666 actually be the case. Of course, in all cases the client gets an
2667 appropriate error code.
2669 15. If a lookup fails in an ACL condition, a message about the failure may
2670 be available; it is used if testing the ACL cannot continue, because most
2671 such messages specify what the cause of the deferral is. However, some
2672 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2673 that caused an old message to be retained and used if a later statement
2674 caused a defer, replacing the real cause of the deferral.
2676 16. If an IP address had so many PTR records that the DNS lookup buffer
2677 was not large enough to hold them, Exim could crash while trying to process
2678 the truncated data. It now detects and logs this case.
2680 17. Further to 4.21/58, another change has been made: if (and only if) the
2681 first line of a message (the first header line) ends with CRLF, a bare LF
2682 in a subsequent header line has a space inserted after it, so as not to
2683 terminate the header.
2685 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2686 unnecessarily, used atoi() instead of strtol(), and didn't check the
2687 termination when getting file sizes from file names by regex.
2689 19. Completely re-implemented the support for maildirsize files, in the light
2690 of a number of problems with the previous contributed implementation
2691 (4.30/29). In particular:
2693 . If the quota is zero, the maildirsize file is maintained, but no quota is
2696 . If the maildir directory does not exist, it is created before any attempt
2697 to write a maildirsize file.
2699 . The quota value in the file is just a cache; if the quota is changed in
2700 the transport, the new value overrides.
2702 . A regular expression is available for excluding directories from the
2705 20. The autoreply transport checks the characters in options that define the
2706 message's headers; it allows continued headers, but it was checking with
2707 isspace() after an embedded newline instead of explicitly looking for a
2710 21. If all the "regular" hosts to which an address was routed had passed their
2711 expiry times, and had not reached their retry times, the address was
2712 bounced, even if fallback hosts were defined. Now Exim should go on to try
2715 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2716 equivalent code in the SMTP transport. Some hosts send humungous responses
2717 to HELO/EHLO, more than 1024 it seems.
2719 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2720 gives compiler warnings in some environments. I've now done it "properly",
2723 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2724 (because of problems with the built-in one) was declared to return uschar *
2725 instead of char *, causing compiler failure.
2727 25. Fixed a file descriptor leak when processing alias/forward files.
2729 26. Fixed a minor format string issue in dbfn.c.
2731 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2733 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2734 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2737 29. When a sender address is verified, it is cached, to save repeating the test
2738 when there is more than one recipient in a message. However, when the
2739 verification involves a callout, it is possible for different callout
2740 options to be set for different recipients. It is too complicated to keep
2741 track of this in the cache, so now Exim always runs a verification when a
2742 callout is required, relying on the callout cache for the optimization.
2743 The overhead is duplication of the address routing, but this should not be
2746 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2747 to be verified with callout=postmaster, and the main callout worked but the
2748 postmaster check failed, the verification correctly failed. However, if a
2749 subsequent RCPT command asked for sender verification *without* the
2750 postmaster check, incorrect caching caused this verification also to fail,
2753 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2754 it was not caching the DNS options (qualify_single, search_parents) that
2755 were used when the lookup failed. A subsequent lookup with different
2756 options therefore always gave the same answer, though there were cases
2757 where it should not have. (Example: a "domains = !$mx_any" option on a
2758 dnslookup router: the "domains" option is always processed without any
2759 widening, but the router might have qualify_single set.) Now Exim uses the
2760 cached value only when the same options are set.
2762 32. Added John Jetmore's "exipick" utility to the distribution.
2764 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2765 than a timeout (e.g. a certificate is required, and is not provided), an
2766 Exim server now closes the connection immediately. Previously it waited for
2767 the client to close - but if the client is SSL, it seems that they each
2768 wait for each other, leading to a delay before one of them times out.
2770 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2771 maintained 0.8.x compatibility because I don't think many are using it, and
2772 it is clearly obsolete.
2774 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2777 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2778 client certificate was expired. A simple patch fixes this, though I don't
2779 understand the full logic of why the verify callback is called multiple
2782 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2783 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2784 which causes problems with some clients (such as the Certicom SSL Plus
2785 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2786 disables the coutermeasure allowing Eudora to connect."
2788 38. Exim was not checking that a write() to a log file succeeded. This could
2789 lead to Bad Things if a log got too big, in particular if it hit a file
2790 size limit. Exim now panics and dies if it cannot write to a log file, just
2791 as it does if it cannot open a log file.
2793 39. Modified OS/Makefile-Linux so that it now contains
2795 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2797 The two -D definitions ensure that Exim is compiled with large file
2798 support, which makes it possible to handle log files that are bigger than
2801 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2802 instance) a domain was checked against a named list that involved a lookup,
2803 causing $domain_data to be set, then another domain was checked against the
2804 same list, then the first domain was re-checked, the value of $domain_data
2805 after the final check could be wrong. In particular, if the second check
2806 failed, it could be set empty. This bug probably also applied to
2809 41. The strip_trailing_dot option was not being applied to the address given
2810 with the -f command-line option.
2812 42. The code for reading a message's header from the spool was incrementing
2813 $received_count, but never initializing it. This meant that the value was
2814 incorrect (doubled) while delivering a message in the same process in which
2815 it was received. In the most common configuration of Exim, this never
2816 happens - a fresh exec is done - but it can happen when
2817 deliver_drop_privilege is set.
2819 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2820 it now includes up to 150 characters of the unexpected data in the log
2823 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2824 and building data strings. The size of both of these buffers was 10 000
2825 bytes - far larger than anybody would *ever* want, thought I. Needless to
2826 say, somebody hit the limit. I have increased the maximum line length to
2827 20 000 and the maximum data length of concatenated lines to 100 000. I have
2828 also fixed two bugs, because there was no checking on these buffers. Tsk,
2829 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2830 buffer is too small.
2832 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2833 lsearch lookups. Now it does.
2835 46. When parsing a route_list item in a manualroute router, a fixed-length
2836 buffer was used for the list of hosts. I made this 1024 bytes long,
2837 thinking that nobody would ever have a list of hosts that long. Wrong.
2838 Somebody had a whole pile of complicated expansion conditions, and the
2839 string was silently truncated, leading to an expansion error. It turns out
2840 that it is easier to change to an unlimited length (owing to other changes
2841 that have happened since this code was originally written) than to build
2842 structure for giving a limitation error. The length of the item that
2843 expands into the list of hosts is now unlimited.
2845 47. The lsearch lookup could not handle data where the length of text line was
2846 more than 4095 characters. Such lines were truncated, leading to shortened
2847 data being returned. It should now handle lines of any length.
2849 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2850 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2853 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2854 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2855 no longer happens in comment lines. A list of replacements is now placed
2856 at the head of all of the source files, except those whose only change is
2857 to replace PERL_COMMAND in the very first #! line.
2859 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2860 messages on the queue, with a bottom-up merge sort, using code contributed
2861 by Michael Haardt. This should make operations like -bp somewhat faster on
2862 large queues. It won't affect queue runners, except when queue_run_in_order
2865 51. Installed eximstats 1.31 in the distribution.
2867 52. Added support for SRV lookups to the dnslookup router.
2869 53. If an ACL referred to $message_body or $message_body_end, the value was not
2870 reset for any messages that followed in the same SMTP session.
2872 54. The store-handling optimization for building very long strings was not
2873 differentiating between the different store pools. I don't think this
2874 actually made any difference in practice, but I've tidied it.
2876 55. While running the routers to verify a sender address, $sender_address
2877 was still set to the sender address. This is wrong, because when routing to
2878 send a bounce to the sender, it would be empty. Therefore, I have changed
2879 it so that, while verifying a sender address, $sender_address is set to <>.
2880 (There is no change to what happens when verifying a recipient address.)
2882 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2883 target A or AAAA records (if not already returned) without resetting the
2884 qualify_single or search_parents options of the DNS resolver. These are
2885 inappropriate in this case because the targets of MX and SRV records must
2886 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2887 target that, when qualified, matched something in the local domain. These
2888 two options are now turned off when doing these lookups.
2890 57. It seems that at least some releases of Reiserfs (which does not have the
2891 concept of a fixed number of inodes) returns zero and not -1 for the
2892 number of available inodes. This interacted badly with check_spool_inodes,
2893 which assumed that -1 was the "no such thing" setting. What I have done is
2894 to check that the total number of inodes is greater than zero before doing
2895 the test of how many are available.
2897 58. When a "warn" ACL statement has a log_message modifier, the message is
2898 remembered, and not repeated. This is to avoid a lot of repetition when a
2899 message has many recipients that cause the same warning to be written.
2900 Howewer, Exim was preserving the list of already written lines for an
2901 entire SMTP session, which doesn't seem right. The memory is now reset if a
2902 new message is started.
2904 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2905 debugging output unless log_rewrite was also set.
2907 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2908 of (int)(handle) when we know that handle contains (void *)(-1).
2910 61. The Exim daemon panic-logs an error return when it closes the incoming
2911 connection. However "connection reset by peer" seems to be common, and
2912 isn't really an error worthy of noting specially, so that particular error
2915 62. When Exim is trying to find all the local interfaces, it used to panic and
2916 die if the ioctl to get the interface flags failed. However, it seems that
2917 on at least one OS (Solaris 9) it is possible to have an interface that is
2918 included in the list of interfaces, but for which you get a failure error
2919 for this call. This happens when the interface is not "plumbed" into a
2920 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2921 failure of the "get flags" call assumes that the interface is down.
2923 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2924 makes life easier for people who are doing arithmetic on fields extracted
2925 from dates, where you often get leading zeros that should not be
2926 interpreted as octal.
2928 64. Added qualify_domain to the redirect router, to override the global
2931 65. If a pathologically long header line contained very many addresses (the
2932 report of this problem mentioned 10 000) and each of them was rewritten,
2933 Exim could use up a very large amount of memory. (It kept on making new
2934 copies of the header line as it rewrote, and never released the old ones.)
2935 At the expense of a bit more processing, the header rewriting function has
2936 been changed so that it no longer eats memory in this way.
2938 66. The generation of the Received: header has been moved from the time that a
2939 message starts to be received, to the time that it finishes. The timestamp
2940 in the Received: header should now be very close to that of the <= log
2941 line. There are two side-effects of this change:
2943 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2944 logged header lines no longer include the local Received: line, because
2945 it has not yet been created. The same applies to a copy of the message
2946 that is returned to a non-SMTP sender when a message is rejected.
2948 (b) When a filter file is tested using -bf, no additional Received: header
2949 is added to the test message. After some thought, I decided that this
2952 This change does not affect the value of $received_for. It is still set
2953 after address rewriting, but before local_scan() is called.
2955 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2957 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2958 gave an unhelpful panic error message, and a defer error. I have managed to
2959 change this behaviour so that it now rejects any supplied certificate,
2960 which seems right, as the list of acceptable certificates is empty.
2962 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2963 gave an unhelpful defer error. I have not managed to make this reject any
2964 supplied certificates, but the error message it gives is "no certificate
2965 supplied", which is not helpful.
2967 70. exigrep's output now also includes lines that are not associated with any
2968 message, but which match the given pattern. Implemented by a patch from
2969 Martin Sluka, which also tidied up the Perl a bit.
2971 71. Recipient callout verification, like sender verification, was using <> in
2972 the MAIL FROM command. This isn't really the right thing, since the actual
2973 sender may affect whether the remote host accepts the recipient or not. I
2974 have changed it to use the actual sender in the callout; this means that
2975 the cache record is now keyed on a recipient/sender pair, not just the
2976 recipient address. There doesn't seem to be a real danger of callout loops,
2977 since a callout by the remote host to check the sender would use <>.
2978 [SEE ABOVE: changed after hitting problems.]
2980 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2981 temporary errors. However, in the case of such a code being given after
2982 the end of a data transmission (i.e. after ".") Exim was failing to write
2983 a retry record for the message. (Yes, there was some broken host that was
2984 actually sending 8xx at this point.)
2986 73. An unknown lookup type in a host list could cause Exim to panic-die when
2987 the list was checked. (An example that provoked this was putting <; in the
2988 middle of a list instead of at the start.) If this happened during a DATA
2989 ACL check, a -D file could be left lying around. This kind of configuration
2990 error no longer causes Exim to die; instead it causes a defer errror. The
2991 incident is still logged to the main and panic logs.
2993 74. Buglet left over from Exim 3 conversion. The message "too many messages
2994 in one connection" was written to the rejectlog but not the mainlog, except
2995 when address rewriting (yes!) was being logged.
2997 75. Added write_rejectlog option.
2999 76. When a system filter was run not as root (that is, when system_filter_user
3000 was set), the values of the $n variables were not being returned to the
3001 main process; thus, they were not subsequently available in the $sn
3004 77. Added +return_path_on_delivery log selector.
3006 78. A connection timeout was being treated differently from recipients deferred
3007 when testing hosts_max_try with a message that was older than the host's
3008 retry timeout. (The host should not be counted, thus allowing all hosts to
3009 be tried at least once before bouncing.) This may have been the cause of an
3010 occasionally reported bug whereby a message would remain on the queue
3011 longer than the retry timeout, but would be bounced if a delivery was
3012 forced. I say "may" because I never totally pinned down the problem;
3013 setting up timeout/retry tests is difficult. See also the next item.
3015 79. The ultimate address timeout was not being applied to errors that involved
3016 a combination of host plus message (for example, a timeout on a MAIL
3017 command). When an address resolved to a number of possible hosts, and they
3018 were not all tried for each delivery (e.g. because of hosts_max_try), a
3019 message could remain on the queue longer than the retry timeout.
3021 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3024 81. Fixed an obscure SMTP outgoing bug which required at least the following
3025 conditions: (a) there was another message waiting for the same server;
3026 (b) the server returned 5xx to all RCPT commands in the first message so
3027 that the message was not completed; (c) the server dropped the connection
3028 or gave a negative response to the RSET that Exim sends to abort the
3029 transaction. The observed case was a dropped connection after DATA that had
3030 been sent in pipelining mode. That is, the server had advertised PIPELINING
3031 but was not implementing it correctly. The effect of the bug was incorrect
3032 behaviour, such as trying another host, and this could lead to a crash.
3038 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3039 and daemon.c were passed as pointers to ints; they should have been
3040 pointers to socklen_t variables (which are typically unsigned ints).
3042 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3045 3. Fixed a really odd bug that affected only the testing scheme; patching a
3046 certain fixed string in the binary changed the value of another string that
3047 happened to be identical to the end of the original first string.
3049 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3050 name", it returns that address as the IP address. On some operating
3051 systems (e.g. Solaris), it also passes back the IP address string as the
3052 "host name". However, on others (e.g. Linux), it passes back an empty
3053 string. Exim wasn't checking for this, and was changing the host name to an
3054 empty string, assuming it had been canonicized.
3056 5. Although rare, it is permitted to have more than one PTR record for a given
3057 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3058 all the names associated with an address, because they do in Solaris.
3059 However, it seems that they do not in Linux for data that comes from the
3060 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3061 I found this out when I moved to a new Linux workstation and tried to run
3062 the Exim test suite.
3064 To get round this problem I have changed the code so that it now does its
3065 own call to the DNS to look up PTR records when searching for a host name.
3066 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3067 addresses that are only in /etc/hosts are still found.
3069 This behaviour is, however, controlled by an option called host_lookup_
3070 order, which defaults to "bydns:byaddr". If people want to use the other
3071 order, or indeed, just use one or the other means of lookup, they can
3072 specify it in this variable.
3074 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3075 some operating systems, this comes back from gethostbyaddr() as an empty
3076 string, and this is what Exim used to test for. However, it seems that in
3077 other systems, "." is yielded. Exim now tests for this case too.
3079 7. The values of check_spool_space and check_log_space are now held internally
3080 as a number of kilobytes instead of an absolute number of bytes. If a
3081 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3082 kilobyte. This means that much larger values can be stored.
3084 8. Exim monitor: an attempt to get the action menu when not actually pointing
3085 at a message produces an empty menu entitled "No message selected". This
3086 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3087 no entries in it ("Shell widget menu has zero width and/or height"). So I
3088 have added a single, blank menu entry in this case.
3090 9. Added ${quote_local_part.
3092 10. MIME decoding is now applied to the contents of Subject: header lines when
3095 11. Now that a reference to $sender_host_address automatically causes a reverse
3096 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3097 host lookup before query-style lookups in lists that might use this
3098 variable. This has therefore been abolished, and the "net-" prefix is no
3099 longer necessary for query-style lookups.
3101 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3102 have been a typo for LFLAGS, so it has been changed.
3104 13. The install script calls Exim with "-C /dev/null" in order to find the
3105 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3106 to be output. Howeve, since Exim outputs its version number before the
3107 error, it didn't break the script. It just looked ugly. I fixed this by
3108 always allowing "-C /dev/null" if the caller is root.
3110 14. Ignore overlarge ACL variable number when reading spool file - insurance
3111 against a later release with more variables having written the file.
3113 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3114 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3115 this, as well as the form without the "IPv6" on the front.
3117 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3118 OS/Makefile-Darwin file.
3120 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3121 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3123 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3124 variable. However, some OS use ints for inodes. I've added cast to long int
3125 to get rid of the compiler warning.
3127 19. I had forgotten to lock out "/../" in configuration file names when
3128 ALT_CONFIG_PREFIX was set.
3130 20. Routers used for verification do not need to specify transports. However,
3131 if such a router generated a host list, and callout was configured, Exim
3132 crashed, because it could not find a port number from the (non-existent)
3133 transport. It now assumes port 25 in this circumstance.
3135 21. Added the -t option to exigrep.
3137 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3138 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3139 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3140 them is set, LOOKUP_LSEARCH is forced.
3142 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3144 24. Added sender and host information to the "rejected by local_scan()" log
3145 line; previously there was no indication of these.
3147 25. Added .include_if_exists.
3149 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3150 Linux. It turns out that not all file systems support this. Apparently some
3151 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3152 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3153 not supported on the file descriptor, is now ignored when Exim is trying to
3154 sync a directory. This applies only to Linux.
3156 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3158 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3159 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3162 29. Added support for maildirsize files from supplied patch (modified a bit).
3164 30. The use of :fail: followed by an empty string could lead Exim to respond to
3165 sender verification failures with (e.g.):
3167 550 Verification failed for <xxx>
3168 550 Sender verify failed
3170 where the first response line was missing the '-' that indicates it is not
3171 the final line of the response.
3173 31. The loop for finding the name of the user that called Exim had a hardwired
3174 limit of 10; it now uses the value of finduser_retries, which is used for
3175 all other user lookups.
3177 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3180 33. Exim was neglecting to zero errno before one call of strtol() when
3181 expanding a string and expecting an integer value. On some systems this
3182 resulted in spurious "integer overflow" errors. Also, it was casting the
3183 result into an int without checking.
3185 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3186 did not work. The code looks as if it has *never* worked, though it appears
3187 to have been documented since at least releast 1.62. I have made it work.
3189 35. The "timeout_DNS" error in retry rules, also documented since at least
3190 1.62, also never worked. As it isn't clear exactly what this means, and
3191 clearly it isn't a major issue, I have abolished the feature by treating it
3192 as "timeout", and writing a warning to the main and panic logs.
3194 36. The display of retry rules for -brt wasn't always showing the error code
3197 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3198 timeout_connect_A, timeout_connect_MX.
3200 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3201 to the empty sender.
3203 39. The daemon was not analysing the content of -oX till after it had closed
3204 stderr and disconnected from the controlling terminal. This meant that any
3205 syntax errors were only noted on the panic log, and the return code from
3206 the command was 0. By re-arranging the code a little, I've made the
3207 decoding happen first, so such errors now appear on stderr, and the return
3208 code is 1. However, the actual setting up of the sockets still happens in
3209 the disconnected process, so errors there are still only recorded on the
3212 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3213 connections (as happens on some IP stacks) was logged at start up time as
3214 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3215 it from "IPv6 and IPv4", which means that two separate sockets are being
3218 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3219 says whether AF_INET or AF_INET6 was passed as an argument.
3221 42. Exiwhat output was messed up when time zones were included in log
3224 43. Exiwhat now gives more information about the daemon's listening ports,
3225 and whether -tls-on-connect was used.
3227 44. The "port" option of the smtp transport is now expanded.
3229 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3230 silently ignored. Now an error message is written to the main and panic
3233 46. There's a new ACL modifier called "logwrite" which writes to a log file
3234 as soon as it is encountered.
3236 47. Added $local_user_uid and $local_user_gid at routing time.
3238 48. Exim crashed when trying to verify a sender address that was being
3241 49. Exim was recognizing only a space character after ".include". It now also
3242 recognizes a tab character.
3244 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3245 extracting the relevant information from the specification. The man page no
3246 longer contains scrambled data for the -d option, and I've added a section
3247 at the front about calling Exim under different names.
3249 51. Added "extra_headers" argument to the "mail" command in filter files.
3251 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3254 53. Installed eximstats 1.29.
3256 54. Added transport_filter_timeout as a generic transport option.
3258 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3259 Cc: header lines. This was required by RFC 822, but it not required by RFC
3262 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3263 incoming messages that did not have them. Now it does so only if the
3264 message originates locally, that is, if there is no associated remote host
3265 address. When Resent- header lines are present, this applies to the Resent-
3266 lines rather than the non-Resent- lines.
3268 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3269 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3271 58. Messages for configuration errors now include the name of the main
3272 configuration file - useful now that there may be more than one file in a
3273 list (.included file names were always shown).
3275 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3276 for those rare installations that do not start the daemon as root or run it
3277 setuid root. I've cut out the call to initgroups() if the daemon is not
3280 60. The Exim user and group can now be bound into the binary as text strings
3281 that are looked up at the start of Exim's processing.
3283 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3285 62. Added $mailstore_basename variable.
3287 63. Installed patch to sieve.c from Michael Haardt.
3289 64. When Exim failed to open the panic log after failing to open the main log,
3290 the original message it was trying to log was written to stderr and debug
3291 output, but if they were not available (the usual case in production), it
3292 was lost. Now it is written to syslog before the two lines that record the
3293 failures to open the logs.
3295 65. Users' Exim filters run in subprocesses under the user's uid. It is
3296 possible for a "deliver" command or an alias in a "personal" command to
3297 provoke an address rewrite. If logging of address rewriting is configured,
3298 this fails because the process is not running as root or exim. There may be
3299 a better way of dealing with this, but for the moment (because 4.30 needs
3300 to be released), I have disabled address rewrite logging when running a
3301 filter in a non-root, non-exim process.
3307 1. The buildconfig auxiliary program wasn't quoting the value set for
3308 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3309 not defined. This bug was masked in 4.22 by the effect that was fixed in
3312 2. Some messages that were rejected after a message id was allocated were
3313 shown as "incomplete" by exigrep. It no longer does this for messages that
3314 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3316 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3317 have allow_domain_literals set, the ID did not get logged in the <= line.
3318 Domain literals are now always recognized in Message-ID: header lines.
3320 4. The first argument for a ${extract expansion item is the key name or field
3321 number. Leading and trailing spaces in this item were not being ignored,
3322 causing some misleading effects.
3324 5. When deliver_drop_privilege was set, single queue runner processes started
3325 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3326 same command in the process it spins off) were not dropping privilege.
3328 6. When the daemon running as "exim" started a queue runner, it always
3329 re-executed Exim in the spun-off process. This is a waste of effort when
3330 deliver_drop_privilege is set. The new process now just calls the
3331 queue-runner function directly.
3337 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3340 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3341 ignored. Though the use of -oP was forcing the writing of a pid file, it
3342 was always written to the default place.
3344 3. If the message "no IP address found for host xxxx" is generated during
3345 incoming verification, it is now followed by identification of the incoming
3346 connection (so you can more easily find what provoked it).
3348 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3350 5. Added some features to "harden" Exim a bit more against certain attacks:
3352 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3353 be put in Local/Makefile. This is like the never_users runtime option,
3354 but it cannot be overridden. The default setting is "root".
3356 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3357 prefix string with which any file named in a -C command line option
3360 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3361 is retained for -C and -D only if the caller of Exim is root. Without
3362 it, the exim user may also use -C and -D and retain privilege.
3364 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3365 command line option is disabled.
3367 6. Macro names set by the -D option must start with an upper case letter, just
3368 like macro names defined in the configuration file.
3370 7. Added "dereference=" facility to LDAP.
3372 8. Two instances of the typo "uknown" in the source files are fixed.
3374 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3375 the Configure-Makefile script screwed up while processing it.
3377 10. Incorporated PCRE 4.4.
3379 11. The SMTP synchronization check was not operating right at the start of an
3380 SMTP session. For example, it could not catch a HELO sent before the client
3381 waited for the greeting. There is now a check for outstanding input at the
3382 point when the greeting is written. Because of the duplex, asynchronous
3383 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3384 way, but not yet received, when the check is performed.
3386 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3387 on TCP/IP sockets, because this apparently causes some broken clients to
3390 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3391 unchanged) from the Cygwin maintainer.
3393 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3394 was supported instead of testing for mbx. Effectively a typo.
3396 15. The spa authenticator server code was not checking that the input it
3397 received was valid base64.
3399 16. The debug output line for the "set" modifier in ACLs was not showing the
3400 name of the variable that was being set.
3402 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3404 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3405 lookup on its own. Something else was needed to trigger the lookup. For
3406 example, a match in host_lookup or the need for a host name in a host list.
3407 Now, if $sender_host_name is referenced and the host name has not yet been
3408 looked up, a lookup is performed. If the lookup fails, the variable remains
3409 empty, and $host_lookup_failed is set to "1".
3411 19. Added "eqi" as a case-independent comparison operator.
3413 20. The saslauthd authentication condition could segfault if neither service
3414 nor realm was specified.
3416 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3417 error message that was logged was misleading, and incoming SMTP
3418 connections were dropped. The message is now more accurate, and temporary
3419 errors are given to SMTP connections.
3421 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3422 (except RSET) with a 421 error, until QUIT is received. However, it was
3423 failing to send a response to QUIT.
3425 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3426 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3427 ACL is now run after the helo_try_verify_hosts code.
3429 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3430 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3431 case-independent, so other case variants are also recognized). Apparently
3432 some systems use these upper case variants.
3434 25. If more than two messages were waiting for the same host, and a transport
3435 filter was specified for the transport, Exim sent two messages over the
3436 same TCP/IP connection, and then failed with "socket operation on non-
3437 socket" when it tried to send the third.
3439 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3441 27. The extern definition of crypt16() in expand.c was not being excluded when
3442 the OS had its own crypt16() function.
3444 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3445 as a preferred synonym for return_size_limit, both as an option and as an
3448 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3450 30. Changed the default configuration ACL to relax the local part checking rule
3451 for addresses that are not in any local domains. For these addresses,
3452 slashes and pipe symbols are allowed within local parts, but the sequence
3453 /../ is explicitly forbidden.
3455 31. SPA server authentication was not clearing the challenge buffer before
3458 32. log_message in a "warn" ACL statement was writing to the reject log as
3459 well as to the main log, which contradicts the documentation and doesn't
3460 seem right (because no rejection is happening). So I have stopped it.
3462 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3463 However, I am unable to do any testing of this.
3465 34. Fixed an infelicity in the appendfile transport. When checking directories
3466 for a mailbox, to see if any needed to be created, it was accidentally
3467 using path names with one or more superfluous leading slashes; tracing
3468 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3470 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3471 contents are added to the log line that is written for every discarded
3472 recipient. (Previously a log_message setting was ignored.)
3474 36. The ${quote: operator now quotes the string if it is empty.
3476 37. The install script runs exim in order to find its version number. If for
3477 some reason other than non-existence or emptiness, which it checks, it
3478 could not run './exim', it was installing it with an empty version number,
3479 i.e. as "exim-". This error state is now caught, and the installation is
3482 38. An argument was missing from the function that creates an error message
3483 when Exim fails to connect to the socket for saslauthd authentication.
3484 This could cause Exim to crash, or give a corrupted message.
3486 39. Added isip, isip4, and isip6 to ${if conditions.
3488 40. The ACL variables $acl_xx are now saved with the message, and can be
3489 accessed later in routers, transports, and filters.
3491 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3492 strings in the file are not string-expanded.
3494 42. If a MAIL command specified a SIZE value that was too large to fit into an
3495 int variable, the check against message_size_limit failed. Such values are
3496 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3497 day this will have to be increased, but I don't think I want to be around
3498 when emails are that large.
3505 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3506 iconv() is not standard in FreeBSD.
3508 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3509 IPv6 enabled. The observed symptom was a segmentation fault on return from
3510 the function os_common_find_running_interfaces() in src/os.c.
3512 3. In the check_special_case() function in daemon.c I had used "errno" as an
3513 argument name, which causes warnings on some systems. This was basically a
3514 typo, since it was named "eno" in the comments!
3516 4. The code that waits for the clock to tick (at a resolution of some fraction
3517 of a second) so as to ensure message-id uniqueness was always waiting for
3518 at least one whole tick, when it could have waited for less. [This is
3519 almost certainly not relevant at current processor speeds, where it is
3520 unlikely to ever wait at all. But we try to future-proof.]
3522 5. The function that sleeps for a time interval that includes fractions of a
3523 second contained a race. It did not block SIGALRM between setting the
3524 timer, and suspending (a couple of lines later). If the interval was short
3525 and the sigsuspend() was delayed until after it had expired, the suspension
3526 never ended. On busy systems this could lead to processes getting stuck for
3529 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3530 process, before it forks any delivery processes. The open lookup caching
3531 mechanism meant that the open file or database connection was passed into
3532 the delivery process. The problem was that delivery processes always tidy
3533 up cached lookup data. This could cause a problem for the next delivery
3534 process started by the queue runner, because the external queue runner
3535 process does not know about the closure. So the next delivery process
3536 still has data in the lookup cache. In the case of a file lookup, there was
3537 no problem because closing a file descriptor in a subprocess doesn't affect
3538 the parent. However, if the lookup was caching a connection to a database,
3539 the connection was closed, and the second delivery process was likely to
3540 see errors such as "PGSQL: query failed: server closed the connection
3541 unexpectedly". The problem has been fixed by closing all cached lookups
3542 in a queue runner before running a delivery process.
3544 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3545 seem to have the "const" qualifier which it has on other OS. I've
3548 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3549 *of the same type* (client or server) with the same public name that an
3550 error should be diagnosed.
3552 9. When Exim looked up a host name for an IP address, but failed to find the
3553 original IP address when looking up the host name (a safety check), it
3554 output the message "<ip address> does not match any IP for NULL", which was
3555 confusing, to say the least. The bug was that the host name should have
3556 appeared instead of "NULL".
3558 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3559 user that is built into the binary, and the -C or -D options is used, root
3560 privilege is dropped before the configuration file is read. In addition,
3561 logging is switched to stderr instead of the normal log files. If the
3562 configuration then re-defines the Exim user, the unprivileged environment
3563 is probably not what is expected, so Exim logs a panic warning message (but
3566 However, if deliver_drop_privilege is set, the unprivileged state may well
3567 be exactly what is intended, so the warning has been cut out in that case,
3568 and Exim is allowed to try to write to its normal log files.
3574 1. smtp_return_error_details was not giving details for temporary sender
3575 or receiver verification errors.
3577 2. Diagnose a configuration error if two authenticators have the same public
3580 3. Exim used not to create the message log file for a message until the first
3581 delivery attempt. This could be confusing when incoming messages were held
3582 for policy or load reasons. The message log file is now created at the time
3583 the message is received, and an initial "Received" line is written to it.
3585 4. The automatically generated man page for command line options had a minor
3586 bug that caused no ill effects; however, a more serious problem was that
3587 the procedure for building the man page automatically didn't always
3588 operate. Consequently, release 4.20 contains an out-of-date version. This
3589 shouldn't happen again.
3591 5. When building Exim with embedded Perl support, the script that builds the
3592 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3593 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3595 6. The freeze_tell option was not being used for messages that were frozen on
3596 arrival, either by an ACL or by local_scan().
3598 7. Added the smtp_incomplete_transaction log selector.
3600 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3601 was accepting AUTH without a new EHLO.
3603 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3604 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3607 10. Logging of TCP/IP connections (when configured) now happens in the main
3608 daemon process instead of the child process, so that the TCP/IP connection
3609 count is more accurate (but it can never be perfect).
3611 11. The use of "drop" in a nested ACL was not being handled correctly in the
3612 outer ACL. Now, if condition failure induced by the nested "drop" causes
3613 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3614 or "require"), the connection is dropped.
3616 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3617 that yield "discard" can now be used with an "accept" or a "discard" verb,
3618 but an error is generated for any others (because I can't see a useful way
3619 to define what should happen).
3621 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3622 are now processed in the same way as lines in the Exim configuration file.
3623 In particular, continuation lines are supported.
3625 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3627 15. Added -ti meaning -t -i.
3629 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3630 domains, and warn by logging if others are found.
3632 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3633 returned from GIFCONF when Exim is trying to find the list of interfaces on
3634 a host. The code in os.c has been modified to copy each ifreq to an aligned
3635 structure in all cases.
3637 Also, in some cases, the returned ifreq's were being copied to a 'struct
3638 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3639 means the last couple of bytes of an IPv6 address could be chopped if the
3640 ifreq contained only a normal sockaddr (14 bytes storage).
3642 18. Named domain lists were not supported in the hosts_treat_as_local option.
3643 An entry such as +xxxx was not recognized, and was treated as a literal
3646 19. Ensure that header lines added by a DATA ACL are included in the reject log
3647 if the ACL subsequently rejects the message.
3649 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3650 MD5 (which is deprecated).
3652 21. When testing a filter file using -bf, Exim was writing a message when it
3653 took the sender from a "From " line in the message, but it was not doing so
3654 when it took $return_path from a Return-Path: header line. It now does.
3656 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3657 with a valid header line field name (a series of printing characters
3658 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3660 23. Changed "disc" in the source to "disk" to conform to the documentation and
3661 the book and for uniformity.
3663 24. Ignore Sendmail's -Ooption=value command line item.
3665 25. When execve() failed while trying to run a command in a pipe transport,
3666 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3667 could be confused with a return value of 69 from the command itself. This
3668 has been changed to 127, the value the shell returns if it is asked to run
3669 a non-existent command. The wording for the related log line suggests a
3670 non-existent command as the problem.
3672 26. If received_header_text expands to an empty string, do not add a Received:
3673 header line to the message. (Well, it adds a token one on the spool, but
3674 marks it "old" so that it doesn't get used or transmitted.)
3676 27. Installed eximstats 1.28 (addition of -nt option).
3678 28. There was no check for failure on the call to getsockname() in the daemon
3679 code. This can fail if there is a shortage of resources on the system, with
3680 ENOMEM, for example. A temporary error is now given on failure.
3682 29. Contrary to the C standard, it seems that in some environments, the
3683 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3684 program. Exim now does this explicitly; it affects the formatting of
3685 timestamps using strftime().
3687 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3688 complaints. I've now initialized all the variables, to avoid this.
3690 32. Header lines added by a system filter were not being "seen" during
3691 transport-time rewrites.
3693 33. The info_callback() function passed to OpenSSL is set up with type void
3694 (*)(SSL *, int, int), as described somewhere. However, when calling the
3695 function (actually a macro) that sets it up, the type void(*)() is
3696 expected. I've put in a cast to prevent warnings from picky compilers.
3698 34. If a DNS black list lookup found a CNAME record, but there were no A
3699 records associated with the domain it pointed at, Exim crashed.
3701 35. If a DNS black list lookup returned more than one A record, Exim ignored
3702 all but the first. It now scans all returned addresses if a particular IP
3703 value is being sought. In this situation, the contents of the
3704 $dnslist_value variable are a list of all the addresses, separated by a
3707 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3708 to accept a host name and all its aliases if the forward lookup for any of
3709 them yielded the IP address of the incoming connection. Now it accepts only
3710 those names whose forward lookup yields the correct IP address. Any other
3711 names are discarded. This closes a loophole whereby a rogue DNS
3712 administrator could create reverse DNS records to break through a
3713 wildcarded host restriction in an ACL.
3715 37. If a user filter or a system filter that ran in a subprocess used any of
3716 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3717 the wrong values were passed to the pipe command ($thisaddress had the
3718 value of $0, $0 had the value of $1, etc). This bug was introduced by
3719 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3721 38. Improved the line breaking for long SMTP error messages from ACLs.
3722 Previously, if there was no break point between 40 and 75 characters, Exim
3723 left the rest of the message alone. Two changes have been made: (a) I've
3724 reduced the minimum length to 35 characters; (b) if it can't find a break
3725 point between 35 and 75 characters, it looks ahead and uses the first one
3726 that it finds. This may give the occasional overlong line, but at least the
3727 remaining text gets split now.
3729 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3730 file descriptors might be low, and that setting 1000 would always raise it.
3731 It turns out that in some environments, the limit is already over 1000 and
3732 that lowering it causes trouble. So now Exim takes care not to decrease it.
3734 40. When delivering a message, the value of $return_path is set to $sender_
3735 address at the start of routing (routers may change the value). By an
3736 oversight, this default was not being set up when an address was tested by
3737 -bt or -bv, which affected the outcome if any router or filter referred to
3740 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3741 log only when "message" or "log_message" are set. However, if one of the
3742 conditions was an address verification, or a call to a nested ACL, the
3743 messages generated by the underlying test were being passed through. This
3744 no longer happens. The underlying message is available in $acl_verify_
3745 message for both "message" and "log_message" expansions, so it can be
3746 passed through if needed.
3748 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3749 new expansion $bh_ to give the encoded byte string without charset
3750 translation. Translation happens only if iconv() is available; HAVE_ICONV
3751 indicates this at build time. HEADERS_CHARSET gives the charset to
3752 translate to; headers_charset can change it in the configuration, and
3753 "headers charset" can change it in an individual filter file.
3755 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3756 that creates RFC 2047 encoded "words" labels them as that charset instead
3757 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3758 expansion operator; (ii) when Exim creates a From: line for a local
3759 message; (iii) when a header line is rewritten to include a "phrase" part.
3761 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3762 buggy, causing it to skip the first lines of messages whose message ID
3763 ended in 'D'. This would not have bitten before Exim release 4.14, because
3764 message IDs were unlikely to end in 'D' before then. The effect was to have
3765 incorrect size information for certain domains.
3767 45. #include "config.h" was missing at the start of the crypt16.c module. This
3768 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3771 46. If there was a timeout during a "random" callout check, Exim treated it as
3772 a failure of the random address, and carried on sending RSET and the real
3773 address. If the delay was just some slowness somewhere, the response to the
3774 original RCPT would be taken as a response to RSET and so on, causing
3775 mayhem of various kinds.
3777 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3778 when I implemented it. It didn't allow for the fact that some option values
3779 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3780 the right test for positive values.
3782 48. Domain names in DNS records are case-independent. Exim always looks them up
3783 in lower case. Some resolvers return domain names in exactly the case they
3784 appear in the zone file, that is, they may contain uppercase letters. Not
3785 all resolvers do this - some return always lower case. Exim was treating a
3786 change of case by a resolver as a change of domain, similar to a widening
3787 of a domain abbreviation. This triggered its re-routing code and so it was
3788 trying to route what was effectively the same domain again. This normally
3789 caused routing to fail (because the router wouldn't handle the domain
3790 twice). Now Exim checks for this case specially, and just changes the
3791 casing of the domain that it ultimately uses when it transmits the message
3794 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3797 50. If a filter generated a file delivery with a non-absolute name (possible if
3798 no home directory exists for the router), the forbid_file option was not
3801 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3802 the existing equality matching.
3804 52. Exim was using ints instead of ino_t variables in some places where it was
3805 dealing with inode numbers.
3807 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3808 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3809 called TMPDIR, and if it finds it is different, it changes its value.
3811 54. The smtp_printf() function is now made available to local_scan() so
3812 additional output lines can be written before returning. There is also an
3813 smtp_fflush() function to enable the detection of a dropped connection.
3814 The variables smtp_input and smtp_batched_input are exported to
3817 55. Changed the default runtime configuration: the message "Unknown user"
3818 has been removed from the ACL, and instead placed on the localuser router,
3819 using the cannot_route_message feature. This means that any verification
3820 failures that generate their own messages won't get overridden. Similarly,
3821 the "Unrouteable address" message that was in the ACL for unverifiable
3822 relay addresses has also been removed.
3824 56. Added hosts_avoid_esmtp to the smtp transport.
3826 57. The exicyclog script was not checking for the esoteric option
3827 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3828 will work only if exicyclog is run under the appropriate euid.
3830 58. Following a discussion on the list, the rules by which Exim recognises line
3831 endings on incoming messages have been changed. The -dropcr and drop_cr
3832 options are now no-ops, retained only for backwards compatibility. The
3833 following line terminators are recognized: LF CRLF CR. However, special
3834 processing applies to CR:
3836 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3837 nor a local message in the state where . is a terminator.
3839 (ii) If a bare CR is encountered in a header line, an extra space is added
3840 after the line terminator so as not to end the header. The reasoning
3841 behind this is that bare CRs in header lines are most likely either
3842 to be mistakes, or people trying to play silly games.
3844 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3845 was being incorrectly given as 18 bytes larger than it should have been.
3846 This is a VOB (very old bug).
3848 60. This may never have affected anything current, but just in case it has:
3849 When the local host is found other than at the start of a list of hosts,
3850 the local host, those with the same MX, and any that follow, are discarded.
3851 When the list in question was part of a longer list of hosts, the following
3852 hosts (not currently being processed) were also being discarded. This no
3853 longer happens. I'm not sure if this situation could ever has previously
3856 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3859 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3860 header. This is something that is recommended in a new Internet Draft, and
3861 is something that is documented as being done by Sendmail. There are two
3862 possible values. For messages generated by the autoreply transport, Exim
3865 Auto-Submitted: auto-replied
3867 whereas for all other generated messages (e.g. bounces) it adds
3869 Auto-Submitted: auto-generated
3871 63. The "personal" condition in filters now includes a test for the
3872 Auto-Submitted: header. If it contains the string "auto-" the message it
3873 not considered personal.
3875 64. Added rcpt_include_affixes as a generic transport option.
3877 65. Added queue_only_override (default true).
3879 66. Added the syslog_duplication option.
3881 67. If what should have been the first header line of a message consisted of
3882 a space followed by a colon, Exim was mis-interpreting it as a header line.
3883 It isn't of course - it is syntactically invalid and should therefore be
3884 treated as the start of the message body. The misbehaviour could have
3885 caused a number of strange effects, including loss of data in subsequent
3886 header lines, and spool format errors.
3888 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3889 client host had authenticated. This control can now be exercised by an ACL
3890 for more flexibility.
3892 69. By default, callouts do not happen when testing with -bh. There is now a
3893 variant, -bhc, which does actually run the callout code, including
3894 consulting and updating the callout cache.
3896 70. Added support for saslauthd authentication, courtesy of Alexander
3899 71. If statvfs() failed on the spool or log directories while checking their
3900 size for availability, Exim confusingly gave the error "space shortage".
3901 Furthermore, in debugging mode it crashed with a floating point exception.
3902 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3903 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3904 serious problem, Exim now writes to the main and panic logs when this
3905 happens, with details of the failure. It then refuses to accept the
3906 incoming message, giving the message "spool directory problem" or "log
3907 directory problem" with a 421 code for SMTP messages.
3909 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3910 0, 1, and 2 exist, because some OS complain for execs without them (see
3911 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3912 descriptors. However, the code omitted to check that the open succeeded,
3913 causing mysterious errors if for some reason the permissions on /dev/null
3914 got screwed. Now Exim writes a message to the main and panic logs, and
3915 bombs out if it can't open /dev/null.
3917 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3918 interact so that it is all more flexible. It is supposed to remain
3919 backwards compatible. Also added extra_local_interfaces.
3921 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3922 to bomb out with an assertion failure - to the client this appears as a
3923 connection drop. This problem occurs in the part of the code that was taken
3924 from the Samba project. Fortunately, the assertion is in a very simple
3925 function, so I have fixed this by reproducing the function inline in the
3926 one place where it is called, and arranging for authentication to fail
3927 instead of killing the process with assert().
3929 75. The SPA client code was not working when the server requested OEM rather
3930 than Unicode encoding.
3932 76. Added code to make require_files with a specific uid setting more usable in
3933 the case where statting the file as root fails - usually a non-root-mounted
3934 NFS file system. When this happens and the failure is EACCES, Exim now
3935 forks a subprocess and does the per-uid checking as the relevant uid.
3937 77. Added process_log_path.
3939 78. If log_file_path was not explicitly set, a setting of check_log_space or
3940 check_log_inodes was ignored.
3942 79. If a space check for the spool or log partitions fails, the incident is now
3943 logged. Of course, in the latter case the data may get lost...
3945 80. Added the %p formatting code to string_format() so that it can be used to
3946 print addresses in debug_print(). Adjusted all the address printing in the
3947 debugging in store.c to use %p rather than %d.
3949 81. There was a concern that a line of code in smtp_in.c could overflow a
3950 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3951 initially expressed, the concern was not well-founded, because trailing
3952 spaces are removed early. However, if the trailing spaces were followed by
3953 a NULL, they did not get removed, so the overflow was possible. Two fixes
3956 (a) I re-wrote the offending code in a cleaner fashion.
3957 (b) If an incoming SMTP command contains a NULL character, it is rejected
3960 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3961 runs initgroups(), so that if the Exim user is in any additional groups,
3962 they will be used during message reception.
3968 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob