1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.347 2006/04/25 10:44:57 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
10 other effects) broke the use of negated acl sub-conditions.
12 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
15 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
16 "Deny" causes Exim to reject the incoming connection with a 554 error.
17 Unfortunately, if there is a major crisis, such as a disk failure,
18 tcp-wrappers gives "deny", whereas what one would like would be some
19 kind of temporary error. A kludge has been added to help with this.
20 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
21 554 error is used if errno is still zero or contains ENOENT (which occurs
22 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
25 PH/03 Add -lutil to the default FreeBSD LIBS setting.
27 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
28 errors. Otherwise a message that provokes a temporary error (when other
29 messages do not) can cause a whole host to time out.
31 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
32 addresses were routed directly to files or pipes from a redirect router.
33 File deliveries just didn't batch; pipe deliveries might have suffered
36 PH/06 A failure to get a lock for a hints database would erroneously always say
37 "Failed to get write lock", even when it was really a read lock.
39 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
40 of 0600. This has been changed to use the value of the lockfile_mode
41 option (which defaults to 0600).
43 PH/08 Applied small patch from the Sieve maintainer.
49 PH/01 The code for finding all the local interface addresses on a FreeBSD
50 system running IPv6 was broken. This may well have applied to all BSD
51 systems, as well as to others that have similar system calls. The broken
52 code found IPv4 interfaces correctly, but gave incorrect values for the
53 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
54 that it would not match correctly against @[] and not recognize the IPv6
57 PH/02 The ipliteral router was not recognizing addresses of the form user@
58 [ipv6:....] because it didn't know about the "ipv6:" prefix.
60 PH/03 Added disable_ipv6.
62 PH/04 Changed $reply_address to use the raw form of the headers instead of the
63 decoded form, because it is most often used to construct To: headers
64 lines in autoreplies, and the decoded form may well be syntactically
65 invalid. However, $reply_address has leading white space removed, and all
66 newlines turned into spaces so that the autoreply transport does not
69 PH/05 If group was specified without a user on a router, and no group or user
70 was specified on a transport, the group from the router was ignored.
72 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
73 for visible compile-time settings that can be used to change these
74 numbers, for those that want even more. Backwards compatibility with old
75 spool files has been maintained. However, going back to a previous Exim
76 release will lost any variables that are in spool files.
78 PH/07 Two small changes when running in the test harness: increase delay when
79 passing a TCP/IP connection to a new process, in case the original
80 process has to generate a bounce, and remove special handling of
81 127.0.0.2 (sic), which is no longer necessary.
83 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
84 be the same on different OS.
86 PH/09 Moved a debug statement in filter processing to avoid a race problem when
89 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
90 whether --show-vars was specified or not
92 JJ/02 exipick: Added support for new ACL variable spool format introduced
95 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
96 syntactically invalid From: or Reply-to: line, and a filter used this to
97 generate an autoreply, and therefore failed to obtain an address for the
98 autoreply, Exim could try to deliver to a non-existent relative file
99 name, causing unrelated and misleading errors. What now happens is that
100 it logs this as a hard delivery error, but does not attempt to create a
103 PH/11 The exinext utility has a -C option for testing purposes, but although
104 the given file was scanned by exinext itself; it wasn't being passed on
105 when Exim was called.
107 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
108 an end-of-file indication when reading a command response.
110 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
111 compiled. In many other places in Exim, IPv6 addresses are always
112 recognized, so I have changed this. It also means that IPv4 domain
113 literals of the form [IPV4:n.n.n.n] are now always recognized.
115 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
116 used if the router is not running as root, for example, when verifying at
117 ACL time, or when using -bh. The debugging output from this situation was
118 non-existent - all you got was a failure to exec. I have made two
121 (a) Failures to set uid/gid, the current directory, or a process leader
122 in a subprocess such as that created by queryprogram now generate
123 suitable debugging ouput when -d is set.
125 (b) The queryprogram router detects when it is not running as root,
126 outputs suitable debugging information if -d is set, and then runs
127 the subprocess without attempting to change uid/gid.
129 PH/15 Minor change to Makefile for building test_host (undocumented testing
132 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
133 additional section of a DNS packet that returns MX or SRV records.
134 Instead, it always explicitly searches for A/AAAA records. This avoids
135 major problems that occur when a DNS server includes only records of one
136 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
137 fixed another bug: if SRV records were looked up and the corresponding
138 address records were *not* found in the additional section, the port
139 values from the SRV records were lost.
141 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
142 using the correct key (the original address) when searching the retry
143 rules in order to find which one to use for generating the retry hint.
145 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
146 adding the default one. Similarly, if it contains a Reply-To: header, the
147 errors_reply_to option, if set, is not used.
149 PH/19 When calculating a retry time, Exim used to measure the "time since
150 failure" by looking at the "first failed" field in the retry record. Now
151 it does not use this if it is later than than the arrival time of the
152 message. Instead it uses the arrival time. This makes for better
153 behaviour in cases where some deliveries succeed, thus re-setting the
154 "first failed" field. An example is a quota failure for a huge message
155 when small messages continue to be delivered. Without this change, the
156 "time since failure" will always be short, possible causing more frequent
157 delivery attempts for the huge message than are intended.
158 [Note: This change was subsequently modified - see PH/04 for 4.62.]
160 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
161 $1, $2, $3) because the numerical variables can be reset during some
162 expansion items (e.g. "match"), thereby losing the authentication data.
164 PH/21 Make -bV show the size of off_t variables so that the test suite can
165 decide whether to run tests for quotas > 2G.
167 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
168 mailbox_size, and mailbox_filecount in the appendfile transport. If a
169 filecount value is greater than 2G or if a quota value is greater than 2G
170 on a system where the size of off_t is not greater than 4, a panic error
173 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
174 never match. The debug and -bh output now contains an explicit error
175 message indicating a malformed IPv4 address or mask.
177 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
178 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
181 PH/25 Do not write to syslog when running in the test harness. The only
182 occasion when this arises is a failure to open the main or panic logs
183 (for which there is an explicit test).
185 PH/26 Added the /no_tell option to "control=freeze".
187 PH/27 If a host name lookup failed very early in a connection, for example, if
188 the IP address matched host_lookup and the reverse lookup yielded a name
189 that did not have a forward lookup, an error message of the form "no IP
190 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
191 could be logged. Now it outputs the IP address instead of "NULL".
193 PH/28 An enabling patch from MH: add new function child_open_exim2() which
194 allows the sender and the authenticated sender to be set when
195 submitting a message from within Exim. Since child_open_exim() is
196 documented for local_scan(), the new function should be too.
198 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
199 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
200 results in an empty string is now treated as unset.
202 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
204 PH/31 Added sender_verify_fail logging option.
206 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
207 needed by RFC 822 but not by RFC 2822 was commented out. I have now
208 tidied the source and removed it altogether.
210 PH/33 When a queue run was abandoned because the load average was too high, a
211 log line was always written; now it is written only if the queue_run log
212 selector is set. In addition, the log line for abandonment now contains
213 information about the queue run such as the pid. This is always present
214 in "start" and "stop" lines but was omitted from the "abandon" line.
216 PH/34 Omit spaces between a header name and the colon in the error message that
217 is given when verify = headers_syntax fails (if there are lots of them,
218 the message gets confusing).
220 PH/35 Change the default for dns_check_names_pattern to allow slashes within
221 names, as there are now some PTR records that contain slashes. This check
222 is only to protect against broken name servers that fall over on strange
223 characters, so the fact that it applies to all lookups doesn't matter.
225 PH/36 Now that the new test suite is complete, we can remove some of the
226 special code in Exim that was needed for the old test suite. For example,
227 sorting DNS records because real resolvers return them in an arbitrary
228 order. The new test suite's fake resolver always returns records in the
231 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
232 bounces) except when queue_only is set, to avoid logging races between
233 the different processes.
235 PH/38 Panic-die if .include specifies a non-absolute path.
237 PH/39 A tweak to the "H" retry rule from its user.
239 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
240 a label. They prevented compilation on older perls.
242 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
243 a warning to be raised on newish perls.
245 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
246 on queue. Changes to match documented behaviour of showing count of
247 messages matching specified criteria.
249 PH/40 Changed the default ident timeout from 30s to 5s.
251 PH/41 Added support for the use of login_cap features, on those BSD systems
252 that have them, for controlling the resources used by pipe deliveries.
254 PH/42 The content-scanning code uses fopen() to create files in which to put
255 message data. Previously it was not paying any attention to the mode of
256 the files. Exim runs with umask(0) because the rest of the code creates
257 files with open(), and sets the required mode explicitly. Thus, these
258 files were ending up world-writeable. This was not a big issue, because,
259 being within the spool directory, they were not world-accessible. I have
260 created a function called modefopen, which takes an additional mode
261 argument. It sets umask(777), creates the file, chmods it to the required
262 mode, then resets the umask. All the relevant calls to fopen() in the
263 content scanning code have been changed to use this function.
265 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
266 to 24 hours. This avoids potential overflow problems when processing G
267 and H retry rules. I suspect nobody ever tinkers with this value.
269 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
271 PH/45 When the plaintext authenticator is running as a client, the server's
272 challenges are checked to ensure they are valid base64 strings. By
273 default, the authentication attempt is cancelled if an invalid string is
274 received. Setting client_ignore_invalid_base64 true ignores these errors.
275 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
276 they are received. Thus, the responses can be made to depend on the
277 challenges. If an invalid string is ignored, an empty string is placed in
280 PH/46 Messages that are created by the autoreply transport now contains a
281 References: header, in accordance with RFCs 2822 and 3834.
283 PH/47 Added authenticated_sender_force to the smtp transport.
285 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
287 PH/49 Installed latest patch from the Sieve maintainer.
289 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
290 was also set, the appendfile transport was unnecessarily scanning a
291 directory of message files (e.g. for maildir delivery) to find the count
292 of files (along with the size), even though it did not need this
293 information. It now does the scan only if it needs to find either the
294 size of the count of files.
296 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
298 PH/52 Two bugs concerned with error handling when the smtp transport is
301 (i) Exim was not creating retry information for temporary errors given
302 for individual recipients after the DATA command when the smtp transport
303 was used in LMTP mode. This meant that they could be retried too
304 frequently, and not timed out correctly.
306 (ii) Exim was setting the flag that allows error details to be returned
307 for LMTP errors on RCPT commands, but not for LMTP errors for individual
308 recipients that were returned after the DATA command.
310 PH/53 This is related to PH/52, but is more general: for any failing address,
311 when detailed error information was permitted to be returned to the
312 sender, but the error was temporary, then after the final timeout, only
313 "retry timeout exceeded" was returned. Now it returns the full error as
314 well as "retry timeout exceeded".
316 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
317 do this, and (what is worse) MTAs that accept it.
319 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
320 will now be deprecated.
322 PH/56 New os.c-cygwin from the Cygwin maintainer.
324 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
325 formats (previously only available in exim formats via -bpr, -bpru,
326 and -bpra. Now also available in native and exiqgrep formats)
328 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
329 with very large, slow to parse queues
331 JJ/08 exipick: added ! as generic prefix to negate any criteria format
333 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
335 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
336 responses to authentication challenges, though it was showing the
337 challenges; (ii) I've removed the CR characters from the debug output for
340 PH/58 Allow for the insertion of a newline as well as a space when a string
341 is turned into more than one encoded-word during RFC 2047 encoding. The
342 Sieve code now uses this.
344 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
345 data_4xx, lost_connection, tls_required.
347 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
348 message was being sent as an SMTP response.
350 PH/61 Add -l and -k options to exicyclog.
352 PH/62 When verifying, if an address was redirected to one new address, so that
353 verification continued, and the new address failed or deferred after
354 having set something in $address_data, the value of $address_data was not
355 passed back to the ACL. This was different to the case when no
356 redirection occurred. The value is now passed back in both cases.
358 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
359 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
360 use login_cap.h, so on its own it isn't the distinguishing feature. The
361 new name refers directly to the setclassresources() function.
363 PH/65 Added configuration files for NetBSD3.
365 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
367 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
368 is preferred over IPv4.
370 PH/68 The bounce_return_message and bounce_return_body options were not being
371 honoured for bounces generated during the reception of non-SMTP messages.
372 In particular, this applied to messages rejected by the ACL. This bug has
373 been fixed. However, if bounce_return_message is true and bounce_return_
374 body is false, the headers that are returned for a non-SMTP message
375 include only those that have been read before the error was detected.
376 (In the case of an ACL rejection, they have all been read.)
378 PH/69 The HTML version of the specification is now built in a directory called
379 spec_html instead of spec.html, because the latter looks like a path with
380 a MIME-type, and this confuses some software.
382 PH/70 Catch two compiler warnings in sieve.c.
384 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
385 function verify_get_ident() calls ip_connect() to connect a socket, but
386 if the "connect()" function timed out, ip_connect() used to close the
387 socket. However, verify_get_ident() also closes the socket later, and in
388 between Exim writes to the log, which may get opened at this point. When
389 the socket was closed in ip_connect(), the log could get the same file
390 descriptor number as the socket. This naturally causes chaos. The fix is
391 not to close the socket in ip_connect(); the socket should be closed by
392 the function that creates it. There was only one place in the code where
393 this was missing, in the iplookup router, which I don't think anybody now
394 uses, but I've fixed it anyway.
396 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
397 well as to direct DNS lookups. Otherwise the handling of names in host
398 lists is inconsistent and therefore confusing.
404 PH/01 Two changes to the default runtime configuration:
406 (1) Move the checks for relay_from_hosts and authenticated clients from
407 after to before the (commented out) DNS black list checks.
409 (2) Add control=submission to the relay_from_hosts and authenticated
410 clients checks, on the grounds that messages accepted by these
411 statements are most likely to be submissions.
413 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
415 (1) Generate an error if the third argument for the ${prvs expansion is
418 (2) Treat a missing third argument of ${prvscheck as if it were an empty
421 (3) Reset the variables that are obtained from the first argument of
422 ${prvscheck and used in the second argument before leaving the code,
423 because their memory is reclaimed, so using them afterwards may do
426 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
427 one (it's much easier than Tom thought :-).
429 (5) Because of (4), we can now allow for the use of $prvscheck_result
430 inside the third argument.
432 PH/03 For some reason, the default setting of PATH when running a command from
433 a pipe transport was just "/usr/bin". I have changed it to
436 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
437 anything to be listed in the output from -bV.
439 PH/05 When a filter generated an autoreply, the entire To: header line was
440 quoted in the delivery log line, like this:
442 => >A.N.Other <ano@some.domain> <original@ddress> ...
444 This has been changed so that it extracts the operative address. There
445 may be more than one such address. If so, they are comma-separated, like
448 => >ano@some.domain,ona@other.domain <original@ddress> ...
450 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
451 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
452 not being looked up in the rDNS to get a host name, Exim was showing the
453 IP address twice in Received: lines, even though the IP addresses were
454 identical. For example:
456 Received: from [1.2.3.4] (helo=[1.2.3.4])
458 However, if the real host name was known, it was omitting the HELO data
459 if it matched the actual IP address. This has been tidied up so that it
460 doesn't show the same IP address twice.
462 PH/07 When both +timestamp and +memory debugging was on, the value given by
463 $tod_xxx expansions could be wrong, because the tod_stamp() function was
464 called by the debug printing, thereby overwriting the timestamp buffer.
465 Debugging no longer uses the tod_stamp() function when +timestamp is set.
467 PH/08 When the original message was included in an autoreply transport, it
468 always said "this is a copy of the message, including all the headers",
469 even if body_only or headers_only was set. It now gives an appropriate
472 PH/09 Applied a patch from the Sieve maintainer which:
474 o fixes some comments
475 o adds the (disabled) notify extension core
476 o adds some debug output for the result of if/elsif tests
477 o points to the current vacation draft in the documentation
478 and documents the missing references header update
482 o fixes a bug in processing the envelope test (when testing
483 multiple envelope elements, the last element determinted the
486 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
487 Electronic Mail") by including:
489 Auto-submitted: auto-generated
491 in the messages that it generates (bounce messages and others, such as
492 warnings). In the case of bounce messages for non-SMTP mesages, there was
493 also a typo: it was using "Auto_submitted" (underscore instead of
494 hyphen). Since every message generated by Exim is necessarily in response
495 to another message, thes have all been changed to:
497 Auto-Submitted: auto-replied
499 in accordance with these statements in the RFC:
501 The auto-replied keyword:
503 - SHOULD be used on messages sent in direct response to another
504 message by an automatic process,
506 - MUST NOT be used on manually-generated messages,
508 - MAY be used on Delivery Status Notifications (DSNs) and Message
509 Disposition Notifications (MDNs),
511 - MUST NOT be used on messages generated by automatic or periodic
512 processes, except for messages which are automatic responses to
515 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
516 to the default Received: header definition.
518 PH/12 Added log selector acl_warn_skipped (default on).
520 PH/13 After a successful wildlsearch lookup, discard the values of numeric
521 variables because (a) they are in the wrong storage pool and (b) even if
522 they were copied, it wouldn't work properly because of the caching.
524 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
525 checking when decoding. Apparently there are clients that generate
526 overlong encoded strings. Why am I not surprised?
528 PH/15 If the first argument of "${if match_address" was not empty, but did not
529 contain an "@" character, Exim crashed. Now it writes a panic log message
530 and treats the condition as false.
532 PH/16 In autoreply, treat an empty string for "once" the same as unset.
534 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
535 extension "envelope-auth". The code is finished and in agreement with
536 other implementations, but there is no documentation so far and in fact,
537 nobody wrote the draft yet. This extension is currently #undef'ed, thus
538 not changing the active code.
540 Print executed "if" and "elsif" statements when debugging is used. This
541 helps a great deal to understand what a filter does.
543 Document more things not specified clearly in RFC3028. I had all this
544 sorted out, when out of a sudden new issues came to my mind. Oops."
546 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
549 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
550 canonical form (as documented). However, after a host name lookup from
551 the IP address, check_host() was doing a simple string comparison with
552 addresses acquired from the DNS when checking that the found name did
553 have the original IP as one of its addresses. Since any found IPv6
554 addresses are likely to be in abbreviated form, the comparison could
555 fail. Luckily, there already exists a function for doing the comparison
556 by converting both addresses to binary, so now that is used instead of
559 PH/20 There was another similar case to PH/19, when a complete host name was
560 given in a host list; looking up its IP address could give an abbreviated
561 form, whereas the current host's name might or might not be abbreviated.
562 The same fix has been applied.
568 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
569 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
572 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
573 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
575 PH/03 Typo: missing ".o" in src/pcre/Makefile.
577 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
578 header line, restrict the check to what is listed in RFCs 2369 and 2929.
579 Also, for "Auto-Submitted", treat anything other than "no" as
580 non-personal, in accordance with RFC 3834. (Previously it treated
581 anything starting "auto-" as non-personal.)
583 TF/01 The control=submission/name=... option had a problem with syntax
584 errors if the name included a slash character. The /name= option
585 now slurps the rest of the string, so it can include any characters
586 but it must come last in the list of options (after /sender_retain
589 PH/05 Some modifications to the interface to the fake nameserver for the new
597 TK/01 Added the "success_on_redirect" address verification option. See
598 NewStuff for rationale and an example.
600 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
602 PH/02 Patch to exigrep to allow it to work on syslog lines.
604 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
605 fread() to skip over the body file's header line, because in Cygwin the
606 header line is locked and is inaccessible.
608 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
609 co-exist for some time) to make it clear that it is the Exim ID that is
610 referenced, not the Message-ID: header line.
612 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
613 string_format() function, because snprintf() does not exist on all
616 PH/06 The use of forbid_filter_existstest now also locks out the use of the
617 ${stat: expansion item.
619 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
620 protocol synchronization error", to keep the pedants happy.
622 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
623 well as for IRIX systems, when gcc is being used. See the host.c source
626 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
628 PH/10 Named domain lists were not working if used in a queue_smtp_domains
631 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
632 transport and to the smtp transport in LMTP mode.
634 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
636 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
637 run a filter in a subprocess. This could lead to confusion in subsequent
638 lookups in the parent process. There should also be a search_tidyup() at
639 the end of the subprocess.
641 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
642 only if the host matched helo_try_verify_hosts, which caused the
643 verification to occur when the EHLO/HELO command was issued. The ACL just
644 tested the remembered result. Now, if a previous verification attempt has
645 not happened, "verify = helo" does it there and then.
647 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
649 TK/03 Fix log output including CR from clamd.
651 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
652 exist provoked a memory error which could cause a segfault.
654 PH/15 Installed PCRE 6.2
656 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
658 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
659 of the problem. Specifically, suggested +O2 rather than +O1 for the
662 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
664 PH/20 If a delivery was routed to a non-standard port by means of an SRV
665 record, the port was not correctly logged when the outgoing_port log
666 selector was set (it logged the transort's default port).
668 PH/21 Added support for host-specific ports to manualroute, queryprogram,
669 fallback_hosts, and "hosts" in the smtp transport.
671 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
672 host errors such as "Connection refused".
674 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
675 authentication with radiusclient 0.4.9:
677 - Error returned from rc_read_config was caught wrongly
678 - Username/password not passed on to radius server due to wrong length.
680 The presumption is that some radiusclient API changes for 4.51/PH/17
681 were not taken care of correctly. The code is still untested by me (my
682 Linux distribution still has 0.3.2 of radiusclient), but it was
683 contributed by a Radius user.
685 PH/24 When doing a callout, the value of $domain wasn't set correctly when
686 expanding the "port" option of the smtp transport.
688 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
689 while reading a MIME header. Thanks to Tom Hughes for a patch.
691 PH/24 Include config.h inside local_scan.h so that configuration settings are
694 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
695 that in an ACL for RCPT (for example), you can examine exactly what was
698 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
699 commands, but it was not correctly comparing the address with the actual
700 client host address. Thus, it would show the EHLO address in Received:
701 header lines when this was not necessary.
703 PH/27 Added the % operator to ${eval:}.
705 PH/28 Exim tries to create and chdir to its spool directory when it starts;
706 it should be ignoring failures (because with -C, for example, it has lost
707 privilege). It wasn't ignoring creation failures other than "already
710 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
713 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
714 because an input line was too long, either on its own, or by virtue of
715 too many continuations, the temporary file was not being removed, and the
716 return code was incorrect.
718 PH/31 Missing "BOOL" in function definition in filtertest.c.
720 PH/32 Applied Sieve patches from the maintainer.
722 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
724 PH/33 Added "verify = not_blind".
726 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
727 Local/Makefile (with some defaults set). These are used in built scripts
728 such as exicyclog, but they have never been used in the exim_install
729 script (though there are many overriding facilities there). I have
730 arranged that the exim_install script now takes note of these two
733 PH/35 Installed configuration files for Dragonfly.
735 PH/36 When a locally submitted message by a trusted user did not contain a
736 From: header, and the sender address was obtained from -f or from an SMTP
737 MAIL command, and the trusted user did not use -F to supply a sender
738 name, $originator_name was incorrectly used when constructing a From:
739 header. Furthermore, $originator_name was used for submission mode
740 messages from external hosts without From: headers in a similar way,
741 which is clearly wrong.
743 PH/37 Added control=suppress_local_fixups.
745 PH/38 When log_selector = +received_sender was set, and the addition of the
746 sender made the log line's construction buffer exactly full, or one byte
747 less than full, an overflow happened when the terminating "\n" was
750 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
751 when the result of a list match is failure because a DNS lookup failed.
753 PH/40 RM_COMMAND is now used in the building process.
755 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
756 the "build-* directories that it finds.
758 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
759 address in a domain literal was a prefix of an interface address.
761 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
762 when verifying a sender address, unless rewrite_headers is false.
764 PH/44 Wrote a long comment about why errors_to addresses are verified as
765 recipients, not senders.
767 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
768 the ratelimit ACL was added.
770 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
772 PH/46 Added extra information about PostgreSQL errors to the error string.
774 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
775 avoiding the need to install special zones in a real server. This is
776 backwards compatible; if it can't find the fake resolver, it drops back.
777 Thus, both old and new test suites can be run.
779 TF/02 Added util/ratelimit.pl
781 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
782 clock is set back in time.
784 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
785 Candler <B.Candler@pobox.com>.
787 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
788 OK for addresses that are the result of redirections.
790 PH/48 A number of further additions for the benefit of the new test suite,
791 including a fake gethostbyname() that interfaces to the fake DNS resolver
794 TF/06 The fix for widen_domains has also been applied to qualify_single and
795 search_parents which are the other dnslookup options that can cause
798 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
801 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
803 TF/07 Exim produced the error message "an SRV record indicated no SMTP
804 service" if it encountered an MX record with an empty target hostname.
805 The message is now "an MX or SRV record indicated no SMTP service".
807 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
808 if the DNS of the sending site is misconfigured. This is quite a
809 common situation. This change restores the behaviour of treating a
810 helo verification defer as a failure.
812 PH/51 If self=fail was set on a router, the bounce message did not include the
813 actual error message.
819 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
821 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
822 command itself ended in error, the underlying message about the transport
823 filter timeout was being overwritten with the pipe command error. Now the
824 underlying error message should be appended to the second error message.
826 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
828 PH/02 Reduce the amount of output that "make" produces by default. Full output
829 can still be requested.
831 PH/03 The warning log line about a condition test deferring for a "warn" verb
832 was being output only once per connection, rather than after each
833 occurrence (because it was using the same function as for successful
834 "warn" verbs). This seems wrong, so I have changed it.
836 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
837 it should not have, which might have caused a crash in the right
838 circumstances, but probably never did.
840 PH/04 Installed a modified version of Tony Finch's patch to make submission
841 mode fix the return path as well as the Sender: header line, and to
842 add a /name= option so that you can make the user's friendly name appear
845 TF/03 Added the control = fakedefer ACL modifier.
847 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
848 Mark Lowes for thorough testing.
850 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
852 TK/03 Merged latest SRS patch from Miles Wilton.
854 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
855 with the definition in sysexits.h (which is #included earlier).
856 Fortunately, Exim does not actually use EX_OK. The code used to try to
857 preserve the sysexits.h value, by assumimg that macro definitions were
858 scanned for macro replacements. I have been disabused of this notion,
859 so now the code just undefines EX_OK before #including unistd.h.
861 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
862 in the smtp transport. When a block could not be written in a single
863 write() function, the timeout was being re-applied to each part-write.
864 This seems wrong - if the receiver was accepting one byte at a time it
865 would take for ever. The timeout is now adjusted when this happens. It
866 doesn't have to be particularly precise.
868 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
869 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
871 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
872 without a domain if the check to <postmaster@domain> fails.
874 SC/01 Eximstats: added -xls and the ability to specify output files
875 (patch written by Frank Heydlauf).
877 SC/02 Eximstats: use FileHandles for outputing results.
879 SC/03 Eximstats: allow any combination of xls, txt, and html output.
881 SC/04 Eximstats: fixed display of large numbers with -nvr option
883 SC/05 Eximstats: fixed merging of reports with empty tables.
885 SC/06 Eximstats: added the -include_original_destination flag
887 SC/07 Eximstats: removed tabs and trailing whitespace.
889 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
891 TK/06 MBOX spool code: Add real "From " MBOX separator line
892 so the .eml file is really in mbox format (even though
893 most programs do not really care). Patch from Alex Miller.
895 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
896 The latter is generated from $received_to and is only set if the
897 message has one envelope recipient. SA can use these headers,
898 obviously out-of-the-box. Patch from Alex Miller.
900 PH/08 The ${def test on a variable was returning false if the variable's
901 value was "0", contrary to what the specification has always said!
902 The result should be true unless the variable is empty.
904 PH/09 The syntax error of a character other than { following "${if
905 def:variable_name" (after optional whitespace) was not being diagnosed.
906 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
907 accidental colon was present, for example, could give incorrect results.
909 PH/10 Tidied the code in a number of places where the st_size field of a stat()
910 result is used (not including appendfile, where other changes are about
913 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
914 This involved changing a lot of size variables from int to off_t. It
915 should work with maildirs and everything.
917 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
918 spamd dying while we are connected to it.
920 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
923 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
924 function for BDB 4.3.
926 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
928 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
929 the sources for PCRE have been re-arranged and more files are now
932 PH/15 The code I had for printing potentially long long variables in PH/11
933 above was not the best (it lost precision). The length of off_t variables
934 is now inspected at build time, and an appropriate printing format (%ld
935 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
936 to be "long long int" or "long int". This is needed for the internal
937 formatting function string_vformat().
939 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
940 the configuration file to be ":syslog", then the script "guesses" where
941 the logs files are, rather than using the compiled in default. In our
942 case the guess is not the same as the compiled default, so the script
943 suddenly stopped working when I started to use syslog. The patch checks
944 to see if log_file_path is "". If so, it attempts to read it from exim
945 with no configuration file to get the compiled in version, before it
946 falls back to the previous guessing code."
948 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
949 implementing BATV in an Exim configuration. See NewStuff for the gory
952 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
953 Makefile that are specific to HP-UX.
955 PH/18 If the "use_postmaster" option was set for a recipient callout together
956 with the "random" option, the postmaster address was used as the MAIL
957 FROM address for the random test, but not for the subsequent recipient
958 test. It is now used for both.
960 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
961 patch removes a few documentation additions to RFC 3028, because the
962 latest draft now contains them. It adds the new en;ascii-case comparator
963 and a new error check for 8bit text in MIME parts. Comparator and
964 require names are now matched exactly. I enabled the subaddress
965 extension, but it is not well tested yet (read: it works for me)."
967 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
968 rework some of the code of TK/09 above to avoid the hardwired use of
969 "%lld" and "long long". Replaced the call to snprintf() with a call to
972 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
973 records point to non-existent hosts", "retry timeout exceeded", and
974 "retry time not reached for any host after a long failure period".
976 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
977 experimental DomainKeys support:
979 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
980 (2) On an error such as an illegally used "control", the wrong name for
981 the control was given.
983 These problems did NOT occur unless DomainKeys support was compiled.
985 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
987 PH/24 Added ${if match_ip condition.
989 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
990 databases so that it will be absolutely obvious if a crash occurs in the
991 DB library. This is a regular occurrence (often caused by mis-matched
994 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
995 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
996 picked up on a user's system that detects such things. There doesn't seem
997 to be a gcc warning option for this - only an attribute that has to be
998 put on the function's prototype. It seems that in Fedora Core 4 they have
999 set this on a number of new functions. No doubt there will be more in due
1002 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1003 specify a transport. However, if an address that was verified by such a
1004 router was the subject of a callout, Exim crashed because it tried to
1005 read the rcpt_include_affixes from the non-existent transport. Now it
1006 just assumes that the setting of that option is false. This bug was
1007 introduced by 4.51/PH/31.
1009 PH/28 Changed -d+all to exclude +memory, because that information is very
1010 rarely of interest, but it makes the output a lot bigger. People tend to
1011 do -d+all out of habit.
1013 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1014 code in os-type was giving problems when libc.so lives in lib64, like on
1017 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1018 aren't the modern standard, and it seems that some systems' include files
1019 don't always have them. Exim was already checking for some of the newer
1020 ones like T_AAAA, and defining it itself. I've added checks for all the
1021 record types that Exim uses.
1023 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1024 not automatically generating a new one, as it is supposed to. This
1025 prevented TLS from working. If the file did exist, but contained invalid
1026 data, a new version was generated, as expected. It was only the case of a
1027 non-existent file that was broken.
1029 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1030 with a change in libdomainkeys > 0.64.
1032 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1033 from DNS. If the selector record carries the flag, it now has
1034 precedence over the domain-wide flag.
1036 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1038 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1039 the use of a port name that isn't defined in /etc/services) occurred, the
1040 message was deferred as in a normal delivery, and thus remained on the
1041 spool, instead of being failed because of the mua_wrapper setting. This
1042 is now fixed, and I tidied up some of the mua_wrapper messages at the
1045 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1046 the messages in a hash of arrays rather than using individual hashes.
1047 This is a bit cleaner and results in dramatic memory savings, albeit
1048 at a slight CPU cost.
1050 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1051 as requested by Marc Sherman.
1053 SC/10 Eximstats: added histograms for user specified patterns as requested
1056 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1058 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1059 fopen() in the content-scanning modules that did not already have it.
1065 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1066 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1068 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1070 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1072 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1073 to test Sieve filters that use "vacation".
1075 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1076 that changes the way the GnuTLS parameters are stored in the cache file.
1077 The new format can be generated externally. For backward compatibility,
1078 if the data in the cache doesn't make sense, Exim assumes it has read an
1079 old-format file, and it generates new data and writes a new file. This
1080 means that you can't go back to an older release without removing the
1083 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1084 work if there are any delivery delays because "one_time" forces the
1085 parent to be marked "delivered", so its unseen clone is never tried
1086 again. For this reason, Exim now forbids the simultaneous setting of
1089 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1090 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1091 entry if you want to know the details. The fix, however, neglected to
1092 consider the case when local delivery batching is involved. The test for
1093 "previously delivered" was not happening when checking to see if an
1094 address could be batched with a previous (undelivered) one; under
1095 certain circumstances this could lead to multiple deliveries to the same
1098 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1099 in its include files, and this causes problems building Exim.
1101 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1102 header_syntax) but Exim was just ignoring anything given after a slash.
1103 In particular, this caused confusion with an attempt to use "verify =
1104 reverse_host_lookup/defer_ok". An error is now given when options are
1105 supplied for verify items that do not have them. (Maybe reverse_host_
1106 lookup should have a defer_ok option, but that's a different point.)
1108 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1109 defined by RFC 821) to 2048, because there were problems with some AUTH
1110 commands, and RFC 1869 says the size should be increased for extended
1111 SMTP commands that take arguments.
1113 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1116 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1117 "unknown" error; now it says that the functionality isn't in the binary.
1119 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1120 an address' error message when a string expansion fails (syntax or
1121 whatever). Otherwise the password may appear in the log. Following change
1122 PH/42 below, there is no longer a chance of it appearing in a bounce
1125 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1127 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1128 of its addresses was ever tried. (Bugzilla bug #2.)
1130 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1131 the result incorrectly in the debug output. (It correctly added a newline
1132 to what was transported.)
1134 TF/01 Added $received_time.
1136 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1137 commented out examples of how to interface to a virus scanner and to
1138 SpamAssassin. Also added commented examples of av_scanner and
1139 spamd_address settings.
1141 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1142 and controls are allowed in which ACLs. There were a couple of minor
1143 errors. Some of the entries in the conditions table (which is a table of
1144 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1145 negation of where the condition IS allowed.
1147 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1149 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1150 header file does not have a version number, so I've had to invent a new
1151 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1152 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1153 radiusclient), but it was contributed by a Radius user.
1155 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1156 files or directories, for OpenSSL.
1158 PH/19 When an Exim process that is running as root has to create an Exim log
1159 file, it does so in a subprocess that runs as exim:exim so as to get the
1160 ownership right at creation (otherwise, other Exim processes might see
1161 the file with the wrong ownership). There was no test for failure of this
1162 fork() call, which would lead to the process getting stuck as it waited
1163 for a non-existent subprocess. Forks do occasionally fail when resources
1164 run out. I reviewed all the other calls to fork(); they all seem to check
1167 PH/20 When checking for unexpected SMTP input at connect time (before writing
1168 the banner), Exim was not dealing correctly with a non-positive return
1169 from the read() function. If the client had disconnected by this time,
1170 the result was a log entry for a synchronization error with an empty
1171 string after "input=" when read() returned zero. If read() returned -1
1172 (an event I could not check), uninitialized data bytes were printed.
1173 There were reports of junk text (parts of files, etc) appearing after
1176 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1178 PH/22 Added support for macro redefinition, and (re)definition in between
1179 driver and ACL definitions.
1181 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1182 forgetting to use the resulting value; it was using the unexpanded value.
1184 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1185 hadn't been configured. The fix is from Juergen Kreileder, who
1186 understands it better than I do:
1188 "Here's what I see happening with three configured cyrus_sasl
1189 authenticators configured (plain, login, cram-md5):
1191 On startup auth_cyrus_sasl_init() gets called for each of these.
1192 This means three calls to sasl_listmech() without a specified mech_list.
1193 => SASL tests which mechs of all available mechs actually work
1194 => three warnings about OTP not working
1195 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1197 With the patch, sasl_listmech() also gets called three times. But now
1198 SASL's mech_list option is set to the server_mech specified in the the
1199 authenticator. Or in other words, the answer from sasl_listmech()
1200 gets limited to just the mech you're testing for (which is different
1202 => the return list contains just 'plain' or 'login', 'cram-md5' or
1203 nothing depending on the value of ob->server_mech.
1205 I've just tested the patch: Authentication still works fine,
1206 unavailable mechs specified in the exim configuration are still
1207 caught, and the auth.log warnings about OTP are gone."
1209 PH/25 When debugging is enabled, the contents of the command line are added
1210 to the debugging output, even when log_selector=+arguments is not
1213 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1214 answer is "GNU", and only if the return is "GNU/something" is the answer
1217 PH/27 $acl_verify_message is now set immediately after the failure of a
1218 verification in an ACL, and so is available in subsequent modifiers. In
1219 particular, the message can be preserved by coding like this:
1221 warn !verify = sender
1222 set acl_m0 = $acl_verify_message
1224 Previously, $acl_verify_message was set only while expanding "message"
1225 and "log_message" when a very denied access.
1227 PH/28 Modified OS/os.c-Linux with
1229 -#ifndef OS_LOAD_AVERAGE
1230 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1232 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1233 nomenclature these days.)
1235 PH/29 Installed patch from the Sieve maintainer that adds the options
1236 sieve_useraddress and sieve_subaddress to the redirect router.
1238 PH/30 In these circumstances:
1239 . Two addresses routed to the same list of hosts;
1240 . First host does not offer TLS;
1241 . First host accepts first address;
1242 . First host gives temporary error to second address;
1243 . Second host offers TLS and a TLS session is established;
1244 . Second host accepts second address.
1245 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1246 and peerdn, if requested) that were in fact used only for the second
1249 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1250 attention to any local part prefix or suffix that was matched by the
1251 router that accepted the address. It now behaves in the same way as it
1252 does for delivery: the affixes are removed from the local part unless
1253 rcpt_include_affixes is set on the transport.
1255 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1256 timeout during the DATA phase of an incoming message.
1258 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1259 applied a patch sanctioned by the Sieve maintainer.
1261 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1262 the uid or gid is negative. A case of a negative gid caused this to be
1263 noticed. The fix allows for either to be negative.
1265 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1266 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1269 PH/36 The change PH/12 above was broken. Fixed it.
1271 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1272 the grounds that routing the same address twice would always produce the
1273 same answer. This might have been true once, but it is certainly no
1274 longer true now. Routing a child address may depend on the previous
1275 routing that produced that child. Some complicated redirection strategies
1276 went wrong when messages had multiple recipients, and made Exim's
1277 behaviour dependent on the order in which the addresses were given.
1279 I have moved the duplicate checking until after the routing is complete.
1280 Exim scans the addresses that are assigned to local and remote
1281 transports, and removes any duplicates. This means that more work will be
1282 done, as duplicates will always all be routed, but duplicates are
1283 presumably rare, so I don't expect this is of any significance.
1285 For deliveries to pipes, files, and autoreplies, the duplicate checking
1286 still happens during the routing process, since they are not going to be
1289 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1290 It corrects a timeout issue with spamd. This is Ian's comment: "The
1291 background is that sometimes spamd either never reads data from a
1292 connection it has accepted, or it never writes response data. The exiscan
1293 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1294 blindly assumes that writes won't block so it may never time out."
1296 PH/39 Allow G after quota size as well as K and M.
1298 PH/40 The value set for $authenticated_id in an authenticator may not contain
1299 binary zeroes or newlines because the value is written to log lines and
1300 to spool files. There was no check on this. Now the value is run through
1301 the string_printing() function so that such characters are converted to
1302 printable escape sequences.
1304 PH/41 $message_linecount is a new variable that contains the total number of
1305 lines in the message. Compare $body_linecount, which is the count for the
1308 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1309 bounce and delay warning messages, except in certain special cases, which
1312 (a) An SMTP error message from a remote host;
1313 (b) A message specified in a :fail: redirection;
1314 (c) A message specified in a "fail" command in a system filter;
1315 (d) A message specified in a FAIL return from the queryprogram router;
1316 (e) A message specified by the cannot_route_message router option.
1318 In these cases only, Exim does include the error details in bounce and
1319 warning messages. There are also a few cases where bland messages such
1320 as "unrouteable address" or "local delivery error" are given.
1322 PH/43 $value is now also set for the "else" part of a ${run expansion.
1324 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1325 being worked on, but at least Exim now implements the latest version to
1328 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1329 process to complete was treated as a delivery failure, a timeout while
1330 writing the message to the pipe was logged, but erroneously treated as a
1331 successful delivery. Such timeouts include transport filter timeouts. For
1332 consistency with the overall process timeout, these timeouts are now
1333 treated as errors, giving rise to delivery failures by default. However,
1334 there is now a new Boolean option for the pipe transport called
1335 timeout_defer, which, if set TRUE, converts the failures into defers for
1336 both kinds of timeout. A transport filter timeout is now identified in
1339 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1340 systems where "make" and "gmake" are different, calling "gmake" at top
1341 level broke things. I've arranged for the value of $(MAKE) to be passed
1342 from the Makefile to this script so that it can call the same version of
1346 A note about Exim versions 4.44 and 4.50
1347 ----------------------------------------
1349 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1350 changes of various kinds. As a consequence, a big documentation update was
1351 needed. This delayed the release for rather longer than seemed good, especially
1352 in the light of a couple of (minor) security issues. Therefore, the changes
1353 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1354 release. So 4.44 and 4.50 are in effect two different branches that both start
1357 I have left the 4.50 change log unchanged; it contains all the changes since
1358 4.43. The change log for 4.44 is below; many of its items are identical to
1359 those for 4.50. This seems to be the most sensible way to preserve the
1360 historical information.
1366 1. Minor wording change to the doc/README.SIEVE file.
1368 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1369 computation of the current number of files was incorrect.
1371 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1372 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1375 4. Give more explanation in the error message when the command for a transport
1376 filter fails to execute.
1378 5. There are several places where Exim runs a non-Exim command in a
1379 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1380 was being done only for the command run by the queryprogram router. It is
1381 now done for all such subprocesses. The other cases are: ${run, transport
1382 filters, and the commands run by the lmtp and pipe transports.
1384 6. Added CONFIGURE_GROUP build-time option.
1386 7. Some older OS have a limit of 256 on the maximum number of file
1387 descriptors. Exim was using setrlimit() to set 1000 as a large value
1388 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1389 systems. I've change it so that if it can't get 1000, it tries for 256.
1391 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1392 was an oversight, and furthermore, ever since the addition of extra
1393 controls (e.g. 4.43/32), the checks on when to allow different forms of
1394 "control" were broken. There should now be diagnostics for all cases when a
1395 control that does not make sense is encountered.
1397 9. Added the /retain_sender option to "control=submission".
1399 10. $recipients is now available in the predata ACL (oversight).
1401 11. Tidy the search cache before the fork to do a delivery from a message
1402 received from the command line. Otherwise the child will trigger a lookup
1403 failure and thereby defer the delivery if it tries to use (for example) a
1404 cached ldap connection that the parent has called unbind on.
1406 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1407 of $address_data from the recipient verification was clobbered by the
1408 sender verification.
1410 13. The value of address_data from a sender verification is now available in
1411 $sender_address_data in subsequent conditions in the ACL statement.
1413 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1415 15. Added a new option "connect=<time>" to callout options, to set a different
1418 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1419 was its contents. (It was OK if the option was not defined at all.)
1421 17. A "Completed" log line is now written for messages that are removed from
1422 the spool by the -Mrm option.
1424 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1425 information about exactly what failed.
1427 19. Added -dd to debug only the daemon process.
1429 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1430 handles timeouts, both on the server side and network timeouts. Renamed the
1431 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1433 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1434 It is now set to "smtps".
1436 22. $host_address is now set to the target address during the checking of
1437 ignore_target_hosts.
1439 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1440 being passed; this would have caused $sender_host_name to have been used if
1441 matching the list had actually called for a host name (not very likely,
1442 since this list is usually IP addresses). A host name is now passed as
1445 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1446 code that specifies a non-restarting handler (typically sigaction() in
1447 modern systems) in an attempt to fix a rare and obscure crash bug.
1449 25. Narrowed the window for a race in the daemon that could cause it to ignore
1450 SIGCHLD signals. This is not a major problem, because they are used only to
1451 wake it up if nothing else does.
1453 26. A malformed maildirsize file could cause Exim to calculate negative values
1454 for the mailbox size or file count. Odd effects could occur as a result.
1455 The maildirsize information is now recalculated if the size or filecount
1458 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1459 support for a long time. Removed HAVE_SYS_VFS_H.
1461 28. Installed the latest version of exipick from John Jetmore.
1463 29. In an address list, if the pattern was not a regular expression, an empty
1464 subject address (from a bounce message) matched only if the pattern was an
1465 empty string. Non-empty patterns were not even tested. This was the wrong
1466 because it is perfectly reasonable to use an empty address as part of a
1467 database query. An empty address is now tested by patterns that are
1468 lookups. However, all the other forms of pattern expect the subject to
1469 contain a local part and a domain, and therefore, for them, an empty
1470 address still always fails if the pattern is not itself empty.
1472 30. Exim went into a mad DNS loop when attempting to do a callout where the
1473 host was specified on an smtp transport, and looking it up yielded more
1474 than one IP address.
1476 31. Re-factored the code for checking spool and log partition space into a
1477 function that finds that data and another that does the check. The former
1478 is then used to implement four new variables: $spool_space, $log_space,
1479 $spool_inodes, and $log_inodes.
1481 32. The RFC2047 encoding function was originally intended for short strings
1482 such as real names; it was not keeping to the 75-character limit for
1483 encoded words that the RFC imposes. It now respects the limit, and
1484 generates multiple encoded words if necessary. To be on the safe side, I
1485 have increased the buffer size for the ${rfc2047: expansion operator from
1488 33. It is now permitted to omit both strings after an "if" condition; if the
1489 condition is true, the result is "true". As before, when the second string
1490 is omitted, a false condition yields an empty string. This makes it less
1491 cumbersome to write custom ACL and router conditions.
1493 34. Failure to deliver a bounce message always caused it to be frozen, even if
1494 there was an errors_to setting on the router. The errors_to setting is now
1497 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1498 canonical form (fully expanded) before being placed in
1499 $sender_host_address.
1501 36. The table in the code that translates DNS record types into text (T_A to
1502 "A" for instance) was missing entries for NS and CNAME. It is just possible
1503 that this could have caused confusion if both these types were looked up
1504 for the same domain, because the text type is used as part of Exim's
1505 per-process caching. But the chance of anyone hitting this buglet seems
1508 37. The dnsdb lookup has been extended in a number of ways.
1510 (1) There is a new type, "zns", which walks up the domain tree until it
1511 finds some nameserver records. It should be used with care.
1513 (2) There is a new type, "mxh", which is like "mx" except that it returns
1514 just the host names, not the priorities.
1516 (3) It is now possible to give a list of domains (or IP addresses) to be
1517 looked up. The behaviour when one of the lookups defers can be
1518 controlled by a keyword.
1520 (4) It is now possible to specify the separator character for use when
1521 multiple records are returned.
1523 38. The dnslists ACL condition has been extended: it is now possible to supply
1524 a list of IP addresses and/or domains to be looked up in a particular DNS
1527 39. Added log_selector=+queue_time_overall.
1529 40. When running the queue in the test harness, wait just a tad after forking a
1530 delivery process, to get repeatability of debugging output.
1532 41. Include certificate and key file names in error message when GnuTLS fails
1533 to set them up, because the GnuTLS error message doesn't include the name
1534 of the failing file when there is a problem reading it.
1536 42. Allow both -bf and -bF in the same test run.
1538 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1540 44. The "Exiscan patch" is now merged into the mainline Exim source.
1542 45. Sometimes the final signoff response after QUIT could fail to get
1543 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1544 < 0 before doing a fflush(). This bug looks as though it goes back to the
1545 introduction of TLS in release 3.20, but "sometimes" must have been rare
1546 because the tests only now provoked it.
1548 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1549 (this can affect the format of dates).
1551 47. exim_tidydb, when checking for the continued existence of a message for
1552 which it has found a message-specific retry record, was not finding
1553 messages that were in split spool directories. Consequently, it was
1554 deleting retry records that should have stayed in existence.
1556 48. Steve fixed some bugs in eximstats.
1558 49. The SPA authentication driver was not abandoning authentication and moving
1559 on to the next authenticator when an expansion was forced to fail,
1560 contradicting the general specification for all authenticators. Instead it
1561 was generating a temporary error. It now behaves as specified.
1563 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1564 (the order specifies the preference for clients). The order is now AES256,
1565 AES128, 3DES, ARCFOUR128.
1567 51. Small patch to Sieve code - explicitly set From: when generating an
1570 52. Exim crashed if a remote delivery caused a very long error message to be
1571 recorded - for instance if somebody sent an entire SpamAssassin report back
1572 as a large number of 550 error lines. This bug was coincidentally fixed by
1573 increasing the size of one of Exim's internal buffers (big_buffer) that
1574 happened as part of the Exiscan merge. However, to be on the safe side, I
1575 have made the code more robust (and fixed the comments that describe what
1578 53. Now that there can be additional text after "Completed" in log lines (if
1579 the queue_time_overall log selector is set), a one-byte patch to exigrep
1580 was needed to allow it to recognize "Completed" as not the last thing in
1583 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1584 patch that reportedly fixes this has been added. I am not expert enough to
1585 create a test for it. This is what the patch creator wrote:
1587 "I found a little strange behaviour of ldap code when working with
1588 Windows 2003 AD Domain, where users was placed in more than one
1589 Organization Units. When I tried to give exim partial DN, the exit code
1590 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1591 But simultaneously result of request was absolutely normal ldap result,
1592 so I produce this patch..."
1594 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1595 REFERENCE, so I have modified the code to exclude the patch when that macro
1598 55. Some experimental protocols are using DNS PTR records for new purposes. The
1599 keys for these records are domain names, not reversed IP addresses. The
1600 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1601 leaves it alone. Component reversal etc. now happens only for IP addresses.
1604 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1606 57. Double the size of the debug message buffer (to 2048) so that more of very
1607 long debug lines gets shown.
1609 58. The exicyclog utility now does better if the number of log files to keep
1610 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1612 59. Two changes related to the smtp_active_hostname option:
1614 (1) $smtp_active_hostname is now available as a variable.
1615 (2) The default for smtp_banner uses $smtp_active_hostname instead
1616 of $primary_hostname.
1618 60. The host_aton() function is supposed to be passed a string that is known
1619 to be a valid IP address. However, in the case of IPv6 addresses, it was
1620 not checking this. This is a hostage to fortune. Exim now panics and dies
1621 if the condition is not met. A case was found where this could be provoked
1622 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1623 components; fortuitously, this particular loophole had already been fixed
1624 by change 4.50/55 above.
1626 If there are any other similar loopholes, the new check in host_aton()
1627 itself should stop them being exploited. The report I received stated that
1628 data on the command line could provoke the exploit when Exim was running as
1629 exim, but did not say which command line option was involved. All I could
1630 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1631 running as the user.
1634 61. There was a buffer overflow vulnerability in the SPA authentication code
1635 (which came originally from the Samba project). I have added a test to the
1636 spa_base64_to_bits() function which I hope fixes it.
1639 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1640 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1642 63. The daemon start-up calls getloadavg() while still root for those OS that
1643 need the first call to be done as root, but it missed one case: when
1644 deliver_queue_load_max is set with deliver_drop_privilege. This is
1645 necessary for the benefit of the queue runner, because there is no re-exec
1646 when deliver_drop_privilege is set.
1648 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1649 This has been fixed.
1651 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1652 was in use, was not putting the data itself into the right store pool;
1653 consequently, it could be overwritten for a subsequent message in the same
1654 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1657 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1659 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1660 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1661 the function were treating the return as a boolean value, which happened to
1662 work because 0=false and not-0=true, but is not correct code.
1664 68. The host_aton() function was not handling scoped IPv6 addresses (those
1665 with, for example, "%eth0" on the end) correctly.
1667 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1668 negated items (that is, ~something) in unsigned ints. Some compilers
1669 apparently mutter when there is no cast.
1671 70. If an address verification called from an ACL failed, and did not produce a
1672 user-specific message (i.e. there was only a "system" message), nothing was
1673 put in $acl_verify_message. In this situation, it now puts the system
1676 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1677 session; change 4.31/43 added the unwanted input to the log line - except
1678 that it did not do this in the start of session case. It now does.
1680 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1681 This is wrong and can cause the other end to generate a synchronization
1682 error if it is another Exim or anything else that does the synchronization
1683 check. A QUIT command is no longer sent after a timeout.
1685 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1686 during host lookups.
1688 74. The defer_ok option of callout verification was not working if it was used
1689 when verifying addresses in header lines, that is, for this case:
1691 verify = header_sender/callout=defer_ok
1693 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1694 those file descriptors could be used for SMTP connections. If anything
1695 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1696 could be sent to the SMTP client, causing chaos. The daemon now opens
1697 stdin, stdout, and stderr to /dev/null when it puts itself into the
1700 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1701 log by default. The user can override this with suitable Perl magic.
1703 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1704 the log message when discard triggers, was not working for the DATA ACL or
1705 for the non-SMTP ACL.
1707 78. Error message wording change in sieve.c.
1709 79. If smtp_accept_max_per_host was set, the number of connections could be
1710 restricted to fewer than expected, because the daemon was trying to set up
1711 a new connection before checking whether the processes handling previous
1712 connections had finished. The check for completed processes is now done
1713 earlier. On busy systems, this bug wouldn't be noticed because something
1714 else would have woken the daemon, and it would have reaped the completed
1717 80. If a message was submitted locally by a user whose login name contained one
1718 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1719 It caused a spool format error. I have fixed the spool reading code. A
1720 related problem was that the "from" clause in the Received: line became
1721 illegal because of the space(s). It is now covered by ${quote_local_part.
1723 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1726 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1728 83. Patch from Sieve maintainer to fix unterminated string problem in
1729 "vacation" handling.
1731 84. Some minor changes to the Linux configuration files to help with other
1732 OS variants using glibc.
1734 85. One more patch for Sieve to update vacation handling to latest spec.
1737 ----------------------------------------------------
1738 See the note above about the 4.44 and 4.50 releases.
1739 ----------------------------------------------------
1745 1. Change 4.43/35 introduced a bug that caused file counts to be
1746 incorrectly computed when quota_filecount was set in an appendfile
1749 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1750 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1753 3. Give more explanation in the error message when the command for a transport
1754 filter fails to execute.
1756 4. There are several places where Exim runs a non-Exim command in a
1757 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1758 was being done only for the command run by the queryprogram router. It is
1759 now done for all such subprocesses. The other cases are: ${run, transport
1760 filters, and the commands run by the lmtp and pipe transports.
1762 5. Some older OS have a limit of 256 on the maximum number of file
1763 descriptors. Exim was using setrlimit() to set 1000 as a large value
1764 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1765 systems. I've change it so that if it can't get 1000, it tries for 256.
1767 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1768 was an oversight, and furthermore, ever since the addition of extra
1769 controls (e.g. 4.43/32), the checks on when to allow different forms of
1770 "control" were broken. There should now be diagnostics for all cases when a
1771 control that does not make sense is encountered.
1773 7. $recipients is now available in the predata ACL (oversight).
1775 8. Tidy the search cache before the fork to do a delivery from a message
1776 received from the command line. Otherwise the child will trigger a lookup
1777 failure and thereby defer the delivery if it tries to use (for example) a
1778 cached ldap connection that the parent has called unbind on.
1780 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1781 of $address_data from the recipient verification was clobbered by the
1782 sender verification.
1784 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1785 was its contents. (It was OK if the option was not defined at all.)
1787 11. A "Completed" log line is now written for messages that are removed from
1788 the spool by the -Mrm option.
1790 12. $host_address is now set to the target address during the checking of
1791 ignore_target_hosts.
1793 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1794 being passed; this would have caused $sender_host_name to have been used if
1795 matching the list had actually called for a host name (not very likely,
1796 since this list is usually IP addresses). A host name is now passed as
1799 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1800 code that specifies a non-restarting handler (typically sigaction() in
1801 modern systems) in an attempt to fix a rare and obscure crash bug.
1803 15. Narrowed the window for a race in the daemon that could cause it to ignore
1804 SIGCHLD signals. This is not a major problem, because they are used only to
1805 wake it up if nothing else does.
1807 16. A malformed maildirsize file could cause Exim to calculate negative values
1808 for the mailbox size or file count. Odd effects could occur as a result.
1809 The maildirsize information is now recalculated if the size or filecount
1812 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1813 support for a long time. Removed HAVE_SYS_VFS_H.
1815 18. Updated exipick to current release from John Jetmore.
1817 19. Allow an empty sender to be matched against a lookup in an address list.
1818 Previously the only cases considered were a regular expression, or an
1821 20. Exim went into a mad DNS lookup loop when doing a callout where the
1822 host was specified on the transport, if the DNS lookup yielded more than
1825 21. The RFC2047 encoding function was originally intended for short strings
1826 such as real names; it was not keeping to the 75-character limit for
1827 encoded words that the RFC imposes. It now respects the limit, and
1828 generates multiple encoded words if necessary. To be on the safe side, I
1829 have increased the buffer size for the ${rfc2047: expansion operator from
1832 22. Failure to deliver a bounce message always caused it to be frozen, even if
1833 there was an errors_to setting on the router. The errors_to setting is now
1836 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1837 canonical form (fully expanded) before being placed in
1838 $sender_host_address.
1840 24. Updated eximstats to version 1.33
1842 25. Include certificate and key file names in error message when GnuTLS fails
1843 to set them up, because the GnuTLS error message doesn't include the name
1844 of the failing file when there is a problem reading it.
1846 26. Expand error message when OpenSSL has problems setting up cert/key files.
1849 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1850 (this can affect the format of dates).
1852 28. exim_tidydb, when checking for the continued existence of a message for
1853 which it has found a message-specific retry record, was not finding
1854 messages that were in split spool directories. Consequently, it was
1855 deleting retry records that should have stayed in existence.
1857 29. eximstats updated to version 1.35
1858 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1859 1.35 - bugfix such that pie charts by volume are generated correctly
1861 30. The SPA authentication driver was not abandoning authentication and moving
1862 on to the next authenticator when an expansion was forced to fail,
1863 contradicting the general specification for all authenticators. Instead it
1864 was generating a temporary error. It now behaves as specified.
1866 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1867 (the order specifies the preference for clients). The order is now AES256,
1868 AES128, 3DES, ARCFOUR128.
1870 31. Small patch to Sieve code - explicitly set From: when generating an
1873 32. Exim crashed if a remote delivery caused a very long error message to be
1874 recorded - for instance if somebody sent an entire SpamAssassin report back
1875 as a large number of 550 error lines. This bug was coincidentally fixed by
1876 increasing the size of one of Exim's internal buffers (big_buffer) that
1877 happened as part of the Exiscan merge. However, to be on the safe side, I
1878 have made the code more robust (and fixed the comments that describe what
1881 33. Some experimental protocols are using DNS PTR records for new purposes. The
1882 keys for these records are domain names, not reversed IP addresses. The
1883 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1884 leaves it alone. Component reversal etc. now happens only for IP addresses.
1887 34. The host_aton() function is supposed to be passed a string that is known
1888 to be a valid IP address. However, in the case of IPv6 addresses, it was
1889 not checking this. This is a hostage to fortune. Exim now panics and dies
1890 if the condition is not met. A case was found where this could be provoked
1891 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1892 components; fortuitously, this particular loophole had already been fixed
1893 by change 4.50/55 or 4.44/33 above.
1895 If there are any other similar loopholes, the new check in host_aton()
1896 itself should stop them being exploited. The report I received stated that
1897 data on the command line could provoke the exploit when Exim was running as
1898 exim, but did not say which command line option was involved. All I could
1899 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1900 running as the user.
1903 35. There was a buffer overflow vulnerability in the SPA authentication code
1904 (which came originally from the Samba project). I have added a test to the
1905 spa_base64_to_bits() function which I hope fixes it.
1908 36. The daemon start-up calls getloadavg() while still root for those OS that
1909 need the first call to be done as root, but it missed one case: when
1910 deliver_queue_load_max is set with deliver_drop_privilege. This is
1911 necessary for the benefit of the queue runner, because there is no re-exec
1912 when deliver_drop_privilege is set.
1914 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1915 was in use, was not putting the data itself into the right store pool;
1916 consequently, it could be overwritten for a subsequent message in the same
1917 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1920 38. Sometimes the final signoff response after QUIT could fail to get
1921 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1922 < 0 before doing a fflush(). This bug looks as though it goes back to the
1923 introduction of TLS in release 3.20, but "sometimes" must have been rare
1924 because the tests only now provoked it.
1930 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1931 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1932 Now it returns an integer. A number of calls were still expecting a T/F
1933 return. Fortuitously, in all cases, the tests worked in OK situations,
1934 which is the norm. However, things would have gone wrong on any write
1935 failures on the smtp file descriptor. This function is used when sending
1936 messages over SMTP and also when doing verify callouts.
1938 2. When Exim is called to do synchronous delivery of a locally submitted
1939 message (the -odf or -odi options), it no longer closes stderr before doing
1942 3. Implemented the mua_wrapper option.
1944 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1946 5. Implemented the functions header_remove(), header_testname(),
1947 header_add_at_position(), and receive_remove_recipient(), and exported them
1950 6. If an ACL "warn" statement specified the addition of headers, Exim already
1951 inserted X-ACL-Warn: at the start if there was no header name. However, it
1952 was not making this test for the second and subsequent header lines if
1953 there were newlines in the string. This meant that an invalid header could
1954 be inserted if Exim was badly configured.
1956 7. Allow an ACL "warn" statement to add header lines at the start or after all
1957 the Received: headers, as well as at the end.
1959 8. Added the rcpt_4xx retry error code.
1961 9. Added postmaster_mailfrom=xxx to callout verification option.
1963 10. Added mailfrom=xxxx to the callout verification option, for verify=
1966 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1967 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1969 12. Inserted some casts to stop certain compilers warning when using pointer
1970 differences as field lengths or precisions in printf-type calls (mostly
1971 affecting debugging statements).
1973 13. Added optional readline() support for -be (dynamically loaded).
1975 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1976 same clock tick as a message's arrival, so that its received time was the
1977 same as the "first fail" time on the retry record, and that message
1978 remained on the queue past the ultimate address timeout, every queue runner
1979 would try a delivery (because it was past the ultimate address timeout) but
1980 after another failure, the ultimate address timeout, which should have then
1981 bounced the address, did not kick in. This was a "< instead of <=" error;
1982 in most cases the first failure would have been in the next clock tick
1983 after the received time, and all would be well.
1985 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1986 being recognized when the domain list was tested by the match_domain
1987 condition in an expansion string.
1989 16. Added the ${str2b64: operator.
1991 17. Exim was always calling setrlimit() to set a large limit for the number of
1992 processes, without checking whether the existing limit was already
1993 adequate. (It did check for the limit on file descriptors.) Furthermore,
1994 errors from getrlimit() and setrlimit() were being ignored. Now they are
1995 logged to the main and panic logs, but Exim does carry on, to try to do its
1996 job under whatever limits there are.
1998 18. Imported PCRE 5.0.
2000 19. Trivial typo in log message " temporarily refused connection" (the leading
2003 20. If the log selector return_path_on_delivery was set and an address was
2004 redirected to /dev/null, the delivery process crashed because it assumed
2005 that a return path would always be set for a "successful" delivery. In this
2006 case, the whole delivery is bypassed as an optimization, and therefore no
2009 21. Internal re-arrangement: the function for sending a challenge and reading
2010 a response while authentication was assuming a zero-terminated challenge
2011 string. It's now changed to take a pointer and a length, to allow for
2012 binary data in such strings.
2014 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2016 23. Exim was not respecting finduser_retries when seeking the login of the
2017 uid under which it was called; it was always trying 10 times. (The default
2018 setting of finduser_retries is zero.) Also, it was sleeping after the final
2019 failure, which is pointless.
2021 24. Implemented tls_on_connect_ports.
2023 25. Implemented acl_smtp_predata.
2025 26. If the domain in control=submission is set empty, Exim assumes that the
2026 authenticated id is a complete email address when it generates From: or
2027 Sender: header lines.
2029 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2030 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2031 chown and chgrp in /bin and hostname in /usr/bin.
2033 28. Exim was keeping the "process log" file open after each use, just as it
2034 does for the main log. This opens the possibility of it remaining open for
2035 long periods when the USR1 signal hits a daemon. Occasional processlog
2036 errors were reported, that could have been caused by this. Anyway, it seems
2037 much more sensible not to leave this file open at all, so that is what now
2040 29. The long-running daemon process does not normally write to the log once it
2041 has entered its main loop, and it closes the log before doing so. This is
2042 so that log files can straightforwardly be renamed and moved. However,
2043 there are a couple of unusual error situations where the daemon does write
2044 log entries, and I had neglected to close the log afterwards.
2046 30. The text of an SMTP error response that was received during a remote
2047 delivery was being truncated at 512 bytes. This is too short for some of
2048 the long messages that one sometimes sees. I've increased the limit to
2051 31. It is now possible to make retry rules that apply only when a message has a
2052 specific sender, in particular, an empty sender.
2054 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2055 it possible to be selective about when SMTP synchronization is enforced.
2057 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2059 32. Implemented hosts_connection_nolog.
2061 33. Added an ACL for QUIT.
2063 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2066 35. Added mailbox_size and mailbox_filecount to appendfile.
2068 36. Added control = no_multiline_responses to ACLs.
2070 37. There was a bug in the logic of the code that waits for the clock to tick
2071 in the case where the clock went backwards by a substantial amount such
2072 that the microsecond fraction of "now" was more than the microsecond
2073 fraction of "then" (but the whole seconds number was less).
2075 38. Added support for the libradius Radius client library this is found on
2076 FreeBSD (previously only the radiusclient library was supported).
2082 1. When certain lookups returned multiple values in the form name=value, the
2083 quoting of the values was not always being done properly. Specifically:
2084 (a) If the value started with a double quote, but contained no whitespace,
2086 (b) If the value contained whitespace other than a space character (i.e.
2087 tabs or newlines or carriage returns) it was not quoted.
2088 This fix has been applied to the mysql and pgsql lookups by writing a
2089 separate quoting function and calling it from the lookup code. The fix
2090 should probably also be applied to nisplus, ibase and oracle lookups, but
2091 since I cannot test any of those, I have not disturbed their existing code.
2093 2. A hit in the callout cache for a specific address caused a log line with no
2094 reason for rejecting RCPT. Now it says "Previous (cached) callout
2095 verification failure".
2097 3. There was an off-by-one bug in the queryprogram router. An over-long
2098 return line was truncated at 256 instead of 255 characters, thereby
2099 overflowing its buffer with the terminating zero. As well as fixing this, I
2100 have increased the buffer size to 1024 (and made a note to document this).
2102 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2103 when Exim is waiting for an SMTP response from a remote server, Exim
2104 restarts its select() call on the socket, thereby resetting its timeout.
2105 This is not a problem when such interrupts are rare. Somebody set up a cron
2106 job to run exiwhat every 2 minutes, which is less than the normal select()
2107 timeout (5 or 10 minutes). This meant that the select() timeout never
2108 kicked in because it was always reset. I have fixed this by comparing the
2109 time when an interrupt arrives with the time at the start of the first call
2110 to select(). If more time than the timeout has elapsed, the interrupt is
2111 treated as a timeout.
2113 5. Some internal re-factoring in preparation for the addition of Sieve
2114 extensions (by MH). In particular, the "personal" test is moved to a
2115 separate function, and given an option for scanning Cc: and Bcc: (which is
2116 not set for Exim filters).
2118 6. When Exim created an email address using the login of the caller as the
2119 local part (e.g. when creating a From: or Sender: header line), it was not
2120 quoting the local part when it contained special characters such as @.
2122 7. Installed new OpenBSD configuration files.
2124 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2125 try to make them clearer.
2127 9. Callout options, other than the timeout value, were being ignored when
2128 verifying sender addresses in header lines. For example, when using
2130 verify = header_sender/callout=no_cache
2132 the cache was (incorrectly) being used.
2134 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2135 only the Cygwin environment.
2137 11. When return_path_on_delivery was set as a log selector, if different remote
2138 addresses in the same message used different return paths and parallel
2139 remote delivery occurred, the wrong values would sometimes be logged.
2140 (Whenever a remote delivery process finished, the return path value from
2141 the most recently started remote delivery process was logged.)
2143 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2144 lines when AUTH and/or TLS are in use. This is the "received protocol"
2145 field. Exim used to use "asmtp" for authenticated SMTP, without any
2146 indication (in the protocol name) for TLS use. Now it follows the RFC and
2147 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2148 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2149 names appear in log lines as well as in Received: header lines.
2151 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2152 extensions, and comparison tests, and to fix some bugs.
2154 14. Changes to the "personal" filter test:
2156 (1) The test was buggy in that it was just doing the equivalent of
2157 "contains" tests on header lines. For example, if a user's address was
2158 anne@some.where, the "personal" test would incorrectly be true for
2160 To: susanne@some.where
2162 This test is now done by extracting each address from the header in turn,
2163 and checking the entire address. Other tests that are part of "personal"
2164 are now done using regular expressions (for example, to check local parts
2165 of addresses in From: header lines).
2167 (2) The list of non-personal local parts in From: addresses has been
2168 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2169 taken from the Sieve specification recommendations.
2171 (3) If the message contains any header line starting with "List-" it is
2172 treated as non-personal.
2174 (4) The test for "circular" in the Subject: header line has been removed
2175 because it now seems ill-conceived.
2177 15. Minor typos in src/EDITME comments corrected.
2179 16. Installed latest exipick from John Jetmore.
2181 17. If headers_add on a router specified a text string that was too long for
2182 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2183 of string_sprintf() is now avoided.
2185 18. $message_body_size was not set (it was always zero) when running the DATA
2186 ACL and the local_scan() function.
2188 19. For the "mail" command in an Exim filter, no default was being set for
2189 the once_repeat time, causing a random time value to be used if "once" was
2190 specified. (If the value happened to be <= 0, no repeat happened.) The
2191 default is now 0s, meaning "never repeat". The "vacation" command was OK
2192 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2193 (I found it when inspecting the code).
2195 20. There is now an overall timeout for performing a callout verification. It
2196 defaults to 4 times the callout timeout, which applies to individual SMTP
2197 commands during the callout. The overall timeout applies when there is more
2198 than one host that can be tried. The timeout is checked before trying the
2199 next host. This prevents very long delays if there are a large number of
2200 hosts and all are timing out (e.g. when the network connections are timing
2201 out). The value of the overall timeout can be changed by specifying an
2202 additional sub-option for "callout", called "maxwait". For example:
2204 verify = sender/callout=5s,maxwait=20s
2206 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2207 to the end before writing, but this should make it even safer).
2209 22. Exim was forgetting that it had advertised PIPELINING for the second and
2210 subsequent messages on an SMTP connection. It was also not resetting its
2211 memory on STARTTLS and an internal HELO.
2213 23. When Exim logs an SMTP synchronization error within a session, it now
2214 records whether PIPELINING has been advertised or not.
2216 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2217 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2218 rather than long int.
2220 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2222 26. Added the never_mail option to autoreply.
2228 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2229 crash if the getsockname() call failed; this can happen if a connection is
2230 closed very soon after it is established. The problem was simply in the
2231 order in which certain operations were done, causing Exim to try to write
2232 to the SMTP stream before it had set up the file descriptor. The bug has
2233 been fixed by making things happen in the correct order.
2239 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2240 before the connection was closed, thus losing the rejection response.
2242 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2243 some early Solaris releases, but causes trouble in current releases where
2244 socklen_t is defined.
2246 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2249 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2252 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2253 that releasing the top of it at the end releases what was used for sub-
2254 expansions (unless the block got too big). However, discard this block if
2255 the first thing is a variable or header, so that we can use its block when
2256 it is dynamic (useful for very large $message_headers, for example).
2258 6. Lookups now cache *every* query, not just the most recent. A new, separate
2259 store pool is used for this. It can be recovered when all lookup caches are
2260 flushed. Lookups now release memory at the end of their result strings.
2261 This has involved some general refactoring of the lookup sources.
2263 7. Some code has been added to the store_xxx() functions to reduce the amount
2264 of flapping under certain conditions.
2266 8. log_incoming_interface used to affect only the <= reception log lines. Now
2267 it causes the local interface and port to be added to several more SMTP log
2268 lines, for example "SMTP connection from", and rejection lines.
2270 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2272 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2274 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2275 could be overwritten at the end of the current message (or the start of a
2276 new message if it was set in a HELO ACL). The value is now preserved for
2277 the duration of the SMTP connection.
2279 12. If a transport had a headers_rewrite setting, and a matching header line
2280 contained an unqualified address, that address was qualified, even if it
2281 did not match any rewriting rules. The underlying bug was that the values
2282 of the flags that permit the existence of unqualified sender and recipient
2283 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2284 non-local messages, and by -bnq for local messages) were not being
2285 preserved with the message after it was received.
2287 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2288 "next input=" as part of the text comprising the host identity instead of
2289 the correct text. The code was using the same buffer for two different
2290 strings. However, depending on which order the printing function evaluated
2291 its arguments, the bug did not always show up. Under Linux, for example, my
2292 test suite worked just fine.
2294 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2295 doesn't work with some older versions of Perl. It has been changed to "my",
2296 which in any case is probably the better facility to use.
2298 15. A really picky compiler found some instances of statements for creating
2299 error messages that either had too many or two few arguments for the format
2302 16. The size of the buffer for calls to the DNS resolver has been increased
2303 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2304 for addresses that have a lot of PTR records. This alleviates a problem; it
2305 does not fully solve it.
2307 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2308 buffer now truncates the list and logs the incident, which is the same
2309 action as happens when Exim is looking up a host name and its aliases.
2310 Previously in this situation something unpredictable would happen;
2311 sometimes it was "internal error: store_reset failed".
2313 18. If a server dropped the connection unexpectedly when an Exim client was
2314 using GnuTLS and trying to read a response, the client delivery process
2315 crashed while trying to generate an error log message.
2317 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2318 string, for example:
2320 warn message = H1: something\nH2: something
2322 the text was added as a single header line from Exim's point of view
2323 though it ended up OK in the delivered message. However, searching for the
2324 second and subsequent header lines using $h_h2: did not work. This has been
2325 fixed. Similarly, if a system filter added multiple headers in this way,
2326 the routers could not see them.
2328 20. Expanded the error message when iplsearch is called with an invalid key to
2329 suggest using net-iplsearch in a host list.
2331 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2332 ACLs are no longer actually imposed (and a message to that effect is
2335 22. If a "gecos" field in a passwd entry contained escaped characters, in
2336 particular, if it contained a \" sequence, Exim got it wrong when building
2337 a From: or a Sender: header from that name. A second bug also caused
2338 incorrect handling when an unquoted " was present following a character
2339 that needed quoting.
2341 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2342 was not being matched caselessly.
2344 24. Arranged for all hyphens in the exim.8 source to be escaped with
2347 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2348 properly. Recipient callout cache records were still being keyed to include
2349 the sender, even when use_sender was set false. This led to far more
2350 callouts that were necessary. The sender is no longer included in the key
2351 when use_sender is false.
2353 26. Added "control = submission" modifier to ACLs.
2355 27. Added the ${base62d: operator to decode base 62 numbers.
2357 28. dnsdb lookups can now access SRV records.
2359 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2360 the configuration file.
2362 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2363 (-v) mode. This makes the output for a verbose queue run more intelligible.
2365 31. Added a use_postmaster feature to recipient callouts.
2367 32. Added the $body_zerocount variable, containing the number of binary zero
2368 bytes in the message body.
2370 33. The time of last modification of the "new" subdirectory is now used as the
2371 "mailbox time last read" when there is a quota error for a maildir
2374 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2376 35. Added +ignore_unknown as a special item in host lists.
2378 36. Code for decoding IPv6 addresses in host lists is now included, even if
2379 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2380 address was recognized as an IP address, but was then not correctly decoded
2381 into binary, causing unexpected and incorrect effects when compared with
2388 1. Very minor rewording of debugging text in manualroute to say "list of
2389 hosts" instead of "hostlist".
2391 2. If verify=header_syntax was set, and a header line with an unqualified
2392 address (no domain) and a large number of spaces between the end of the
2393 name and the colon was received, the reception process suffered a buffer
2394 overflow, and (when I tested it) crashed. This was caused by some obsolete
2395 code that should have been removed. The fix is to remove it!
2397 3. When running in the test harness, delay a bit after writing a bounce
2398 message to get a bit more predictability in the log output.
2400 4. Added a call to search_tidyup() just before forking a reception process. In
2401 theory, someone could use a lookup in the expansion of smtp_accept_max_
2402 per_host which, without the tidyup, could leave open a database connection.
2404 5. Added the variables $recipient_data and $sender_data which get set from a
2405 lookup success in an ACL "recipients" or "senders" condition, or a router
2406 "senders" option, similar to $domain_data and $local_part_data.
2408 6. Moved the writing of debug_print from before to after the "senders" test
2411 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2412 problems for message scanning, either using a data ACL, or using
2413 local_scan() because the Received: header was not generated till after they
2414 were called (in order to set the time as the time of reception completion).
2415 I have revised the way this works. The header is now generated after the
2416 body is received, but before the ACL or local_scan() are called. After they
2417 are run, the timestamp in the header is updated.
2423 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2424 before starting a queue runner without re-exec. This happened only when
2425 deliver_drop_privilege was set or when the Exim user was set to root. The
2426 effect of the bug was that timeouts during subsequent deliveries caused
2427 crashes instead of being properly handled. The handler is now left at its
2428 default (and expected) setting.
2430 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2431 message, again when deliver_drop_privilege is set or Exim is run as root.
2432 The bug described in (1) was not present in this case, but the tidying up
2433 of the other signals was missing. I have made the two cases consistent.
2435 3. The ignore_target_hosts setting on a manualroute router was being ignored
2436 for hosts that were looked up using the /MX notation.
2438 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2441 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2442 operated on the sender address. After changing the $sender_address to <>
2443 for the sender address verify, Exim was re-instated it as the original
2444 (before rewriting) address, but remembering that it had rewritten it, so it
2445 wasn't rewriting it again. This bug also had the effect of breaking the
2446 sender address verification caching when the sender address was rewritten.
2448 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2449 This has been changed so that if the ip literal address matches
2450 ignore_target_hosts, the router declines.
2452 7. Added expansion conditions match_domain, match_address, and match_local_
2453 part (NOT match_host).
2455 8. The placeholder for the Received: header didn't have a length field set.
2457 9. Added code to Exim itself and to exim_lock to test for a specific race
2458 condition that could lead to file corruption when using MBX delivery. The
2459 issue is with the lockfile that is created in /tmp. If this file is removed
2460 after a process has opened it but before that process has acquired a lock,
2461 there is the potential for a second process to recreate the file and also
2462 acquire a lock. This could lead to two Exim processes writing to the file
2463 at the same time. The added code performs the same test as UW imapd; it
2464 checks after acquiring the lock that its file descriptor still refers to
2465 the same named file.
2467 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2468 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2471 11. Added the smtp_active_hostname option. If used, this will typically be made
2472 to depend on the incoming interface address. Because $interface_address is
2473 not set up until the daemon has forked a reception process, error responses
2474 that can happen earlier (such as "too many connections") no longer contain
2477 12. If an expansion in a condition on a "warn" statement fails because a lookup
2478 defers, the "warn" statement is abandoned, and the next ACL statement is
2479 processed. Previously this caused the whole ACL to be aborted.
2481 13. Added the iplsearch lookup type.
2483 14. Added ident_timeout as a log selector.
2485 15. Added tls_certificate_verified as a log selector.
2487 16. Added a global option tls_require_ciphers (compare the smtp transport
2488 option of the same name). This controls incoming TLS connections.
2490 17. I finally figured out how to make tls_require_ciphers do a similar thing
2491 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2492 before starting the TLS session.
2494 18. Tabs are now shown as \t in -bP output.
2496 19. If the log selector return_path_on_delivery was set, Exim crashed when
2497 bouncing a message because it had too many Received: header lines.
2499 20. If two routers both had headers_remove settings, and the first one included
2500 a superfluous trailing colon, the final name in the first list and the
2501 first name in the second list were incorrectly joined into one item (with a
2502 colon in the middle).
2508 1. Added -C and -D options to the exinext utility, mainly to make it easier
2509 to include in the automated testing, but these could be helpful when
2510 multiple configurations are in use.
2512 2. The exinext utility was not formatting the output nicely when there was
2513 an alternate port involved in the retry record key, nor when there was a
2514 message id as well (for retries that were specific to a specific message
2515 and a specific host). It was also confused by IPv6 addresses, because of
2516 the additional colons they contain. I have fixed the IPv4 problem, and
2517 patched it up to do a reasonable job for IPv6.
2519 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2520 delivery, the log line now contains "pipelined" if PIPELINING was used.
2522 4. An SMTP transport process used to panic and die if the bind() call to set
2523 an explicit outgoing interface failed. This has been changed; it is now
2524 treated in the same way as a connect() failure.
2526 5. A reference to $sender_host_name in the part of a conditional expansion
2527 that was being skipped was still causing a DNS lookup. This no longer
2530 6. The def: expansion condition was not recognizing references to header lines
2531 that used bh_ and bheader_.
2533 7. Added the _cache feature to named lists.
2535 8. The code for checking quota_filecount in the appendfile transport was
2536 allowing one more file than it should have been.
2538 9. For compatibility with Sendmail, the command line option
2546 and sets the incoming protocol and host name (for trusted callers). The
2547 host name and its colon can be omitted when only the protocol is to be set.
2548 Note the Exim already has two private options, -pd and -ps, that refer to
2549 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2550 "s", but I don't think that's a major issue.
2552 10. A number of refactoring changes to the code, none of which should affect
2555 (a) The number of logging options was getting close to filling up the
2556 32-bit word that was used as a bit map. I have split them into two classes:
2557 those that are passed in the argument to log_write(), and those that are
2558 only ever tested independently outside of that function. These are now in
2559 separate 32-bit words, so there is plenty of room for expansion again.
2560 There is no change in the user interface or the logging behaviour.
2562 (b) When building, for example, log lines, the code previously used a
2563 macro that called string_cat() twice, in order to add two strings. This is
2564 not really sufficiently general. Furthermore, there was one instance where
2565 it was actually wrong because one of the argument was used twice, and in
2566 one call a function was used. (As it happened, calling the function twice
2567 did not affect the overall behaviour.) The macro has been replaced by a
2568 function that can join an arbitrary number of extra strings onto a growing
2571 (c) The code for expansion conditions now uses a table and a binary chop
2572 instead of a serial search (which was left over from when there were very
2573 few conditions). Also, it now recognizes conditions like "pam" even when
2574 the relevant support is not compiled in: a suitably worded error message is
2575 given if an attempt is made to use such a condition.
2577 11. Added ${time_interval:xxxxx}.
2579 12. A bug was causing one of the ddress fields not to be passed back correctly
2580 from remote delivery subprocesses. The field in question was not being
2581 subsequently used, so this caused to problems in practice.
2583 13. Added new log selectors queue_time and deliver_time.
2585 14. Might have fixed a bug in maildirsizefile handling that threw up
2586 "unexpected character" debug warnings, and recalculated the data
2587 unnecessarily. In any case, I expanded the warning message to give more
2590 15. Added the message "Restricted characters in address" to the statements in
2591 the default ACL that block characters like @ and % in local parts.
2593 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2594 Three changes have been made:
2596 (a) There was a serious bug; a negative response to MAIL caused the whole
2597 recipient domain to be cached as invalid, thereby blocking all messages
2598 to all local parts at the same domain, from all senders. This bug has
2599 been fixed. The domain is no longer cached after a negative response to
2600 MAIL if the sender used is not empty.
2602 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2605 (c) A new callout option, "use_sender" has been added for people who want
2606 the modified behaviour.
2612 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2615 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2616 indeed breaks things for older releases.
2618 3. Added additional logging to the case where there is a problem reading data
2619 from a filter that is running in a subprocess using a pipe, in order to
2620 try to track down a specific problem.
2622 4. Testing facility fudge: when running in the test harness and attempting
2623 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2624 getting "No route to host". Convert this to a timeout.
2626 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2629 6. Some OS don't have socklen_t but use size_t instead. This affects the
2630 fifth argument of getsockopt() amongst other things. This is now
2631 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2632 can be set for individual OS. I have set it for SunOS5, OSF1, and
2633 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2634 some earlier ones do not.
2636 7. Change 4.30/15 was not doing the test caselessly.
2638 8. The standard form for an IPv6 address literal was being rejected by address
2639 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2640 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2641 this, as well as the form without the "IPv6" on the front (but only when
2642 address literals are enabled, of course).
2644 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2646 10. Exim crashed if a message with an empty sender address specified by -f
2647 encountered a router with an errors_to setting. This could be provoked only
2648 by a command such as
2652 where an empty string was supplied; "<>" did not hit this bug.
2654 11. Installed PCRE release 4.5.
2656 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2657 remained set. It is now erased.
2659 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2660 times from message ids (which are base 36 rather than the normal 62).
2662 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2663 were being counted as actual protocol errors, and logged if the log
2664 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2665 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2666 and DATA following a set of rejected RCPTs do not count as protocol errors.
2667 In other words, Exim assumes they were pipelined, though this may not
2668 actually be the case. Of course, in all cases the client gets an
2669 appropriate error code.
2671 15. If a lookup fails in an ACL condition, a message about the failure may
2672 be available; it is used if testing the ACL cannot continue, because most
2673 such messages specify what the cause of the deferral is. However, some
2674 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2675 that caused an old message to be retained and used if a later statement
2676 caused a defer, replacing the real cause of the deferral.
2678 16. If an IP address had so many PTR records that the DNS lookup buffer
2679 was not large enough to hold them, Exim could crash while trying to process
2680 the truncated data. It now detects and logs this case.
2682 17. Further to 4.21/58, another change has been made: if (and only if) the
2683 first line of a message (the first header line) ends with CRLF, a bare LF
2684 in a subsequent header line has a space inserted after it, so as not to
2685 terminate the header.
2687 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2688 unnecessarily, used atoi() instead of strtol(), and didn't check the
2689 termination when getting file sizes from file names by regex.
2691 19. Completely re-implemented the support for maildirsize files, in the light
2692 of a number of problems with the previous contributed implementation
2693 (4.30/29). In particular:
2695 . If the quota is zero, the maildirsize file is maintained, but no quota is
2698 . If the maildir directory does not exist, it is created before any attempt
2699 to write a maildirsize file.
2701 . The quota value in the file is just a cache; if the quota is changed in
2702 the transport, the new value overrides.
2704 . A regular expression is available for excluding directories from the
2707 20. The autoreply transport checks the characters in options that define the
2708 message's headers; it allows continued headers, but it was checking with
2709 isspace() after an embedded newline instead of explicitly looking for a
2712 21. If all the "regular" hosts to which an address was routed had passed their
2713 expiry times, and had not reached their retry times, the address was
2714 bounced, even if fallback hosts were defined. Now Exim should go on to try
2717 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2718 equivalent code in the SMTP transport. Some hosts send humungous responses
2719 to HELO/EHLO, more than 1024 it seems.
2721 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2722 gives compiler warnings in some environments. I've now done it "properly",
2725 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2726 (because of problems with the built-in one) was declared to return uschar *
2727 instead of char *, causing compiler failure.
2729 25. Fixed a file descriptor leak when processing alias/forward files.
2731 26. Fixed a minor format string issue in dbfn.c.
2733 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2735 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2736 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2739 29. When a sender address is verified, it is cached, to save repeating the test
2740 when there is more than one recipient in a message. However, when the
2741 verification involves a callout, it is possible for different callout
2742 options to be set for different recipients. It is too complicated to keep
2743 track of this in the cache, so now Exim always runs a verification when a
2744 callout is required, relying on the callout cache for the optimization.
2745 The overhead is duplication of the address routing, but this should not be
2748 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2749 to be verified with callout=postmaster, and the main callout worked but the
2750 postmaster check failed, the verification correctly failed. However, if a
2751 subsequent RCPT command asked for sender verification *without* the
2752 postmaster check, incorrect caching caused this verification also to fail,
2755 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2756 it was not caching the DNS options (qualify_single, search_parents) that
2757 were used when the lookup failed. A subsequent lookup with different
2758 options therefore always gave the same answer, though there were cases
2759 where it should not have. (Example: a "domains = !$mx_any" option on a
2760 dnslookup router: the "domains" option is always processed without any
2761 widening, but the router might have qualify_single set.) Now Exim uses the
2762 cached value only when the same options are set.
2764 32. Added John Jetmore's "exipick" utility to the distribution.
2766 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2767 than a timeout (e.g. a certificate is required, and is not provided), an
2768 Exim server now closes the connection immediately. Previously it waited for
2769 the client to close - but if the client is SSL, it seems that they each
2770 wait for each other, leading to a delay before one of them times out.
2772 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2773 maintained 0.8.x compatibility because I don't think many are using it, and
2774 it is clearly obsolete.
2776 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2779 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2780 client certificate was expired. A simple patch fixes this, though I don't
2781 understand the full logic of why the verify callback is called multiple
2784 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2785 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2786 which causes problems with some clients (such as the Certicom SSL Plus
2787 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2788 disables the coutermeasure allowing Eudora to connect."
2790 38. Exim was not checking that a write() to a log file succeeded. This could
2791 lead to Bad Things if a log got too big, in particular if it hit a file
2792 size limit. Exim now panics and dies if it cannot write to a log file, just
2793 as it does if it cannot open a log file.
2795 39. Modified OS/Makefile-Linux so that it now contains
2797 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2799 The two -D definitions ensure that Exim is compiled with large file
2800 support, which makes it possible to handle log files that are bigger than
2803 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2804 instance) a domain was checked against a named list that involved a lookup,
2805 causing $domain_data to be set, then another domain was checked against the
2806 same list, then the first domain was re-checked, the value of $domain_data
2807 after the final check could be wrong. In particular, if the second check
2808 failed, it could be set empty. This bug probably also applied to
2811 41. The strip_trailing_dot option was not being applied to the address given
2812 with the -f command-line option.
2814 42. The code for reading a message's header from the spool was incrementing
2815 $received_count, but never initializing it. This meant that the value was
2816 incorrect (doubled) while delivering a message in the same process in which
2817 it was received. In the most common configuration of Exim, this never
2818 happens - a fresh exec is done - but it can happen when
2819 deliver_drop_privilege is set.
2821 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2822 it now includes up to 150 characters of the unexpected data in the log
2825 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2826 and building data strings. The size of both of these buffers was 10 000
2827 bytes - far larger than anybody would *ever* want, thought I. Needless to
2828 say, somebody hit the limit. I have increased the maximum line length to
2829 20 000 and the maximum data length of concatenated lines to 100 000. I have
2830 also fixed two bugs, because there was no checking on these buffers. Tsk,
2831 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2832 buffer is too small.
2834 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2835 lsearch lookups. Now it does.
2837 46. When parsing a route_list item in a manualroute router, a fixed-length
2838 buffer was used for the list of hosts. I made this 1024 bytes long,
2839 thinking that nobody would ever have a list of hosts that long. Wrong.
2840 Somebody had a whole pile of complicated expansion conditions, and the
2841 string was silently truncated, leading to an expansion error. It turns out
2842 that it is easier to change to an unlimited length (owing to other changes
2843 that have happened since this code was originally written) than to build
2844 structure for giving a limitation error. The length of the item that
2845 expands into the list of hosts is now unlimited.
2847 47. The lsearch lookup could not handle data where the length of text line was
2848 more than 4095 characters. Such lines were truncated, leading to shortened
2849 data being returned. It should now handle lines of any length.
2851 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2852 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2855 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2856 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2857 no longer happens in comment lines. A list of replacements is now placed
2858 at the head of all of the source files, except those whose only change is
2859 to replace PERL_COMMAND in the very first #! line.
2861 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2862 messages on the queue, with a bottom-up merge sort, using code contributed
2863 by Michael Haardt. This should make operations like -bp somewhat faster on
2864 large queues. It won't affect queue runners, except when queue_run_in_order
2867 51. Installed eximstats 1.31 in the distribution.
2869 52. Added support for SRV lookups to the dnslookup router.
2871 53. If an ACL referred to $message_body or $message_body_end, the value was not
2872 reset for any messages that followed in the same SMTP session.
2874 54. The store-handling optimization for building very long strings was not
2875 differentiating between the different store pools. I don't think this
2876 actually made any difference in practice, but I've tidied it.
2878 55. While running the routers to verify a sender address, $sender_address
2879 was still set to the sender address. This is wrong, because when routing to
2880 send a bounce to the sender, it would be empty. Therefore, I have changed
2881 it so that, while verifying a sender address, $sender_address is set to <>.
2882 (There is no change to what happens when verifying a recipient address.)
2884 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2885 target A or AAAA records (if not already returned) without resetting the
2886 qualify_single or search_parents options of the DNS resolver. These are
2887 inappropriate in this case because the targets of MX and SRV records must
2888 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2889 target that, when qualified, matched something in the local domain. These
2890 two options are now turned off when doing these lookups.
2892 57. It seems that at least some releases of Reiserfs (which does not have the
2893 concept of a fixed number of inodes) returns zero and not -1 for the
2894 number of available inodes. This interacted badly with check_spool_inodes,
2895 which assumed that -1 was the "no such thing" setting. What I have done is
2896 to check that the total number of inodes is greater than zero before doing
2897 the test of how many are available.
2899 58. When a "warn" ACL statement has a log_message modifier, the message is
2900 remembered, and not repeated. This is to avoid a lot of repetition when a
2901 message has many recipients that cause the same warning to be written.
2902 Howewer, Exim was preserving the list of already written lines for an
2903 entire SMTP session, which doesn't seem right. The memory is now reset if a
2904 new message is started.
2906 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2907 debugging output unless log_rewrite was also set.
2909 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2910 of (int)(handle) when we know that handle contains (void *)(-1).
2912 61. The Exim daemon panic-logs an error return when it closes the incoming
2913 connection. However "connection reset by peer" seems to be common, and
2914 isn't really an error worthy of noting specially, so that particular error
2917 62. When Exim is trying to find all the local interfaces, it used to panic and
2918 die if the ioctl to get the interface flags failed. However, it seems that
2919 on at least one OS (Solaris 9) it is possible to have an interface that is
2920 included in the list of interfaces, but for which you get a failure error
2921 for this call. This happens when the interface is not "plumbed" into a
2922 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2923 failure of the "get flags" call assumes that the interface is down.
2925 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2926 makes life easier for people who are doing arithmetic on fields extracted
2927 from dates, where you often get leading zeros that should not be
2928 interpreted as octal.
2930 64. Added qualify_domain to the redirect router, to override the global
2933 65. If a pathologically long header line contained very many addresses (the
2934 report of this problem mentioned 10 000) and each of them was rewritten,
2935 Exim could use up a very large amount of memory. (It kept on making new
2936 copies of the header line as it rewrote, and never released the old ones.)
2937 At the expense of a bit more processing, the header rewriting function has
2938 been changed so that it no longer eats memory in this way.
2940 66. The generation of the Received: header has been moved from the time that a
2941 message starts to be received, to the time that it finishes. The timestamp
2942 in the Received: header should now be very close to that of the <= log
2943 line. There are two side-effects of this change:
2945 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2946 logged header lines no longer include the local Received: line, because
2947 it has not yet been created. The same applies to a copy of the message
2948 that is returned to a non-SMTP sender when a message is rejected.
2950 (b) When a filter file is tested using -bf, no additional Received: header
2951 is added to the test message. After some thought, I decided that this
2954 This change does not affect the value of $received_for. It is still set
2955 after address rewriting, but before local_scan() is called.
2957 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2959 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2960 gave an unhelpful panic error message, and a defer error. I have managed to
2961 change this behaviour so that it now rejects any supplied certificate,
2962 which seems right, as the list of acceptable certificates is empty.
2964 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2965 gave an unhelpful defer error. I have not managed to make this reject any
2966 supplied certificates, but the error message it gives is "no certificate
2967 supplied", which is not helpful.
2969 70. exigrep's output now also includes lines that are not associated with any
2970 message, but which match the given pattern. Implemented by a patch from
2971 Martin Sluka, which also tidied up the Perl a bit.
2973 71. Recipient callout verification, like sender verification, was using <> in
2974 the MAIL FROM command. This isn't really the right thing, since the actual
2975 sender may affect whether the remote host accepts the recipient or not. I
2976 have changed it to use the actual sender in the callout; this means that
2977 the cache record is now keyed on a recipient/sender pair, not just the
2978 recipient address. There doesn't seem to be a real danger of callout loops,
2979 since a callout by the remote host to check the sender would use <>.
2980 [SEE ABOVE: changed after hitting problems.]
2982 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2983 temporary errors. However, in the case of such a code being given after
2984 the end of a data transmission (i.e. after ".") Exim was failing to write
2985 a retry record for the message. (Yes, there was some broken host that was
2986 actually sending 8xx at this point.)
2988 73. An unknown lookup type in a host list could cause Exim to panic-die when
2989 the list was checked. (An example that provoked this was putting <; in the
2990 middle of a list instead of at the start.) If this happened during a DATA
2991 ACL check, a -D file could be left lying around. This kind of configuration
2992 error no longer causes Exim to die; instead it causes a defer errror. The
2993 incident is still logged to the main and panic logs.
2995 74. Buglet left over from Exim 3 conversion. The message "too many messages
2996 in one connection" was written to the rejectlog but not the mainlog, except
2997 when address rewriting (yes!) was being logged.
2999 75. Added write_rejectlog option.
3001 76. When a system filter was run not as root (that is, when system_filter_user
3002 was set), the values of the $n variables were not being returned to the
3003 main process; thus, they were not subsequently available in the $sn
3006 77. Added +return_path_on_delivery log selector.
3008 78. A connection timeout was being treated differently from recipients deferred
3009 when testing hosts_max_try with a message that was older than the host's
3010 retry timeout. (The host should not be counted, thus allowing all hosts to
3011 be tried at least once before bouncing.) This may have been the cause of an
3012 occasionally reported bug whereby a message would remain on the queue
3013 longer than the retry timeout, but would be bounced if a delivery was
3014 forced. I say "may" because I never totally pinned down the problem;
3015 setting up timeout/retry tests is difficult. See also the next item.
3017 79. The ultimate address timeout was not being applied to errors that involved
3018 a combination of host plus message (for example, a timeout on a MAIL
3019 command). When an address resolved to a number of possible hosts, and they
3020 were not all tried for each delivery (e.g. because of hosts_max_try), a
3021 message could remain on the queue longer than the retry timeout.
3023 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3026 81. Fixed an obscure SMTP outgoing bug which required at least the following
3027 conditions: (a) there was another message waiting for the same server;
3028 (b) the server returned 5xx to all RCPT commands in the first message so
3029 that the message was not completed; (c) the server dropped the connection
3030 or gave a negative response to the RSET that Exim sends to abort the
3031 transaction. The observed case was a dropped connection after DATA that had
3032 been sent in pipelining mode. That is, the server had advertised PIPELINING
3033 but was not implementing it correctly. The effect of the bug was incorrect
3034 behaviour, such as trying another host, and this could lead to a crash.
3040 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3041 and daemon.c were passed as pointers to ints; they should have been
3042 pointers to socklen_t variables (which are typically unsigned ints).
3044 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3047 3. Fixed a really odd bug that affected only the testing scheme; patching a
3048 certain fixed string in the binary changed the value of another string that
3049 happened to be identical to the end of the original first string.
3051 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3052 name", it returns that address as the IP address. On some operating
3053 systems (e.g. Solaris), it also passes back the IP address string as the
3054 "host name". However, on others (e.g. Linux), it passes back an empty
3055 string. Exim wasn't checking for this, and was changing the host name to an
3056 empty string, assuming it had been canonicized.
3058 5. Although rare, it is permitted to have more than one PTR record for a given
3059 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3060 all the names associated with an address, because they do in Solaris.
3061 However, it seems that they do not in Linux for data that comes from the
3062 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3063 I found this out when I moved to a new Linux workstation and tried to run
3064 the Exim test suite.
3066 To get round this problem I have changed the code so that it now does its
3067 own call to the DNS to look up PTR records when searching for a host name.
3068 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3069 addresses that are only in /etc/hosts are still found.
3071 This behaviour is, however, controlled by an option called host_lookup_
3072 order, which defaults to "bydns:byaddr". If people want to use the other
3073 order, or indeed, just use one or the other means of lookup, they can
3074 specify it in this variable.
3076 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3077 some operating systems, this comes back from gethostbyaddr() as an empty
3078 string, and this is what Exim used to test for. However, it seems that in
3079 other systems, "." is yielded. Exim now tests for this case too.
3081 7. The values of check_spool_space and check_log_space are now held internally
3082 as a number of kilobytes instead of an absolute number of bytes. If a
3083 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3084 kilobyte. This means that much larger values can be stored.
3086 8. Exim monitor: an attempt to get the action menu when not actually pointing
3087 at a message produces an empty menu entitled "No message selected". This
3088 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3089 no entries in it ("Shell widget menu has zero width and/or height"). So I
3090 have added a single, blank menu entry in this case.
3092 9. Added ${quote_local_part.
3094 10. MIME decoding is now applied to the contents of Subject: header lines when
3097 11. Now that a reference to $sender_host_address automatically causes a reverse
3098 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3099 host lookup before query-style lookups in lists that might use this
3100 variable. This has therefore been abolished, and the "net-" prefix is no
3101 longer necessary for query-style lookups.
3103 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3104 have been a typo for LFLAGS, so it has been changed.
3106 13. The install script calls Exim with "-C /dev/null" in order to find the
3107 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3108 to be output. Howeve, since Exim outputs its version number before the
3109 error, it didn't break the script. It just looked ugly. I fixed this by
3110 always allowing "-C /dev/null" if the caller is root.
3112 14. Ignore overlarge ACL variable number when reading spool file - insurance
3113 against a later release with more variables having written the file.
3115 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3116 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3117 this, as well as the form without the "IPv6" on the front.
3119 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3120 OS/Makefile-Darwin file.
3122 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3123 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3125 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3126 variable. However, some OS use ints for inodes. I've added cast to long int
3127 to get rid of the compiler warning.
3129 19. I had forgotten to lock out "/../" in configuration file names when
3130 ALT_CONFIG_PREFIX was set.
3132 20. Routers used for verification do not need to specify transports. However,
3133 if such a router generated a host list, and callout was configured, Exim
3134 crashed, because it could not find a port number from the (non-existent)
3135 transport. It now assumes port 25 in this circumstance.
3137 21. Added the -t option to exigrep.
3139 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3140 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3141 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3142 them is set, LOOKUP_LSEARCH is forced.
3144 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3146 24. Added sender and host information to the "rejected by local_scan()" log
3147 line; previously there was no indication of these.
3149 25. Added .include_if_exists.
3151 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3152 Linux. It turns out that not all file systems support this. Apparently some
3153 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3154 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3155 not supported on the file descriptor, is now ignored when Exim is trying to
3156 sync a directory. This applies only to Linux.
3158 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3160 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3161 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3164 29. Added support for maildirsize files from supplied patch (modified a bit).
3166 30. The use of :fail: followed by an empty string could lead Exim to respond to
3167 sender verification failures with (e.g.):
3169 550 Verification failed for <xxx>
3170 550 Sender verify failed
3172 where the first response line was missing the '-' that indicates it is not
3173 the final line of the response.
3175 31. The loop for finding the name of the user that called Exim had a hardwired
3176 limit of 10; it now uses the value of finduser_retries, which is used for
3177 all other user lookups.
3179 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3182 33. Exim was neglecting to zero errno before one call of strtol() when
3183 expanding a string and expecting an integer value. On some systems this
3184 resulted in spurious "integer overflow" errors. Also, it was casting the
3185 result into an int without checking.
3187 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3188 did not work. The code looks as if it has *never* worked, though it appears
3189 to have been documented since at least releast 1.62. I have made it work.
3191 35. The "timeout_DNS" error in retry rules, also documented since at least
3192 1.62, also never worked. As it isn't clear exactly what this means, and
3193 clearly it isn't a major issue, I have abolished the feature by treating it
3194 as "timeout", and writing a warning to the main and panic logs.
3196 36. The display of retry rules for -brt wasn't always showing the error code
3199 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3200 timeout_connect_A, timeout_connect_MX.
3202 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3203 to the empty sender.
3205 39. The daemon was not analysing the content of -oX till after it had closed
3206 stderr and disconnected from the controlling terminal. This meant that any
3207 syntax errors were only noted on the panic log, and the return code from
3208 the command was 0. By re-arranging the code a little, I've made the
3209 decoding happen first, so such errors now appear on stderr, and the return
3210 code is 1. However, the actual setting up of the sockets still happens in
3211 the disconnected process, so errors there are still only recorded on the
3214 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3215 connections (as happens on some IP stacks) was logged at start up time as
3216 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3217 it from "IPv6 and IPv4", which means that two separate sockets are being
3220 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3221 says whether AF_INET or AF_INET6 was passed as an argument.
3223 42. Exiwhat output was messed up when time zones were included in log
3226 43. Exiwhat now gives more information about the daemon's listening ports,
3227 and whether -tls-on-connect was used.
3229 44. The "port" option of the smtp transport is now expanded.
3231 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3232 silently ignored. Now an error message is written to the main and panic
3235 46. There's a new ACL modifier called "logwrite" which writes to a log file
3236 as soon as it is encountered.
3238 47. Added $local_user_uid and $local_user_gid at routing time.
3240 48. Exim crashed when trying to verify a sender address that was being
3243 49. Exim was recognizing only a space character after ".include". It now also
3244 recognizes a tab character.
3246 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3247 extracting the relevant information from the specification. The man page no
3248 longer contains scrambled data for the -d option, and I've added a section
3249 at the front about calling Exim under different names.
3251 51. Added "extra_headers" argument to the "mail" command in filter files.
3253 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3256 53. Installed eximstats 1.29.
3258 54. Added transport_filter_timeout as a generic transport option.
3260 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3261 Cc: header lines. This was required by RFC 822, but it not required by RFC
3264 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3265 incoming messages that did not have them. Now it does so only if the
3266 message originates locally, that is, if there is no associated remote host
3267 address. When Resent- header lines are present, this applies to the Resent-
3268 lines rather than the non-Resent- lines.
3270 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3271 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3273 58. Messages for configuration errors now include the name of the main
3274 configuration file - useful now that there may be more than one file in a
3275 list (.included file names were always shown).
3277 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3278 for those rare installations that do not start the daemon as root or run it
3279 setuid root. I've cut out the call to initgroups() if the daemon is not
3282 60. The Exim user and group can now be bound into the binary as text strings
3283 that are looked up at the start of Exim's processing.
3285 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3287 62. Added $mailstore_basename variable.
3289 63. Installed patch to sieve.c from Michael Haardt.
3291 64. When Exim failed to open the panic log after failing to open the main log,
3292 the original message it was trying to log was written to stderr and debug
3293 output, but if they were not available (the usual case in production), it
3294 was lost. Now it is written to syslog before the two lines that record the
3295 failures to open the logs.
3297 65. Users' Exim filters run in subprocesses under the user's uid. It is
3298 possible for a "deliver" command or an alias in a "personal" command to
3299 provoke an address rewrite. If logging of address rewriting is configured,
3300 this fails because the process is not running as root or exim. There may be
3301 a better way of dealing with this, but for the moment (because 4.30 needs
3302 to be released), I have disabled address rewrite logging when running a
3303 filter in a non-root, non-exim process.
3309 1. The buildconfig auxiliary program wasn't quoting the value set for
3310 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3311 not defined. This bug was masked in 4.22 by the effect that was fixed in
3314 2. Some messages that were rejected after a message id was allocated were
3315 shown as "incomplete" by exigrep. It no longer does this for messages that
3316 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3318 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3319 have allow_domain_literals set, the ID did not get logged in the <= line.
3320 Domain literals are now always recognized in Message-ID: header lines.
3322 4. The first argument for a ${extract expansion item is the key name or field
3323 number. Leading and trailing spaces in this item were not being ignored,
3324 causing some misleading effects.
3326 5. When deliver_drop_privilege was set, single queue runner processes started
3327 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3328 same command in the process it spins off) were not dropping privilege.
3330 6. When the daemon running as "exim" started a queue runner, it always
3331 re-executed Exim in the spun-off process. This is a waste of effort when
3332 deliver_drop_privilege is set. The new process now just calls the
3333 queue-runner function directly.
3339 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3342 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3343 ignored. Though the use of -oP was forcing the writing of a pid file, it
3344 was always written to the default place.
3346 3. If the message "no IP address found for host xxxx" is generated during
3347 incoming verification, it is now followed by identification of the incoming
3348 connection (so you can more easily find what provoked it).
3350 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3352 5. Added some features to "harden" Exim a bit more against certain attacks:
3354 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3355 be put in Local/Makefile. This is like the never_users runtime option,
3356 but it cannot be overridden. The default setting is "root".
3358 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3359 prefix string with which any file named in a -C command line option
3362 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3363 is retained for -C and -D only if the caller of Exim is root. Without
3364 it, the exim user may also use -C and -D and retain privilege.
3366 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3367 command line option is disabled.
3369 6. Macro names set by the -D option must start with an upper case letter, just
3370 like macro names defined in the configuration file.
3372 7. Added "dereference=" facility to LDAP.
3374 8. Two instances of the typo "uknown" in the source files are fixed.
3376 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3377 the Configure-Makefile script screwed up while processing it.
3379 10. Incorporated PCRE 4.4.
3381 11. The SMTP synchronization check was not operating right at the start of an
3382 SMTP session. For example, it could not catch a HELO sent before the client
3383 waited for the greeting. There is now a check for outstanding input at the
3384 point when the greeting is written. Because of the duplex, asynchronous
3385 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3386 way, but not yet received, when the check is performed.
3388 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3389 on TCP/IP sockets, because this apparently causes some broken clients to
3392 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3393 unchanged) from the Cygwin maintainer.
3395 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3396 was supported instead of testing for mbx. Effectively a typo.
3398 15. The spa authenticator server code was not checking that the input it
3399 received was valid base64.
3401 16. The debug output line for the "set" modifier in ACLs was not showing the
3402 name of the variable that was being set.
3404 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3406 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3407 lookup on its own. Something else was needed to trigger the lookup. For
3408 example, a match in host_lookup or the need for a host name in a host list.
3409 Now, if $sender_host_name is referenced and the host name has not yet been
3410 looked up, a lookup is performed. If the lookup fails, the variable remains
3411 empty, and $host_lookup_failed is set to "1".
3413 19. Added "eqi" as a case-independent comparison operator.
3415 20. The saslauthd authentication condition could segfault if neither service
3416 nor realm was specified.
3418 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3419 error message that was logged was misleading, and incoming SMTP
3420 connections were dropped. The message is now more accurate, and temporary
3421 errors are given to SMTP connections.
3423 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3424 (except RSET) with a 421 error, until QUIT is received. However, it was
3425 failing to send a response to QUIT.
3427 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3428 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3429 ACL is now run after the helo_try_verify_hosts code.
3431 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3432 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3433 case-independent, so other case variants are also recognized). Apparently
3434 some systems use these upper case variants.
3436 25. If more than two messages were waiting for the same host, and a transport
3437 filter was specified for the transport, Exim sent two messages over the
3438 same TCP/IP connection, and then failed with "socket operation on non-
3439 socket" when it tried to send the third.
3441 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3443 27. The extern definition of crypt16() in expand.c was not being excluded when
3444 the OS had its own crypt16() function.
3446 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3447 as a preferred synonym for return_size_limit, both as an option and as an
3450 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3452 30. Changed the default configuration ACL to relax the local part checking rule
3453 for addresses that are not in any local domains. For these addresses,
3454 slashes and pipe symbols are allowed within local parts, but the sequence
3455 /../ is explicitly forbidden.
3457 31. SPA server authentication was not clearing the challenge buffer before
3460 32. log_message in a "warn" ACL statement was writing to the reject log as
3461 well as to the main log, which contradicts the documentation and doesn't
3462 seem right (because no rejection is happening). So I have stopped it.
3464 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3465 However, I am unable to do any testing of this.
3467 34. Fixed an infelicity in the appendfile transport. When checking directories
3468 for a mailbox, to see if any needed to be created, it was accidentally
3469 using path names with one or more superfluous leading slashes; tracing
3470 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3472 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3473 contents are added to the log line that is written for every discarded
3474 recipient. (Previously a log_message setting was ignored.)
3476 36. The ${quote: operator now quotes the string if it is empty.
3478 37. The install script runs exim in order to find its version number. If for
3479 some reason other than non-existence or emptiness, which it checks, it
3480 could not run './exim', it was installing it with an empty version number,
3481 i.e. as "exim-". This error state is now caught, and the installation is
3484 38. An argument was missing from the function that creates an error message
3485 when Exim fails to connect to the socket for saslauthd authentication.
3486 This could cause Exim to crash, or give a corrupted message.
3488 39. Added isip, isip4, and isip6 to ${if conditions.
3490 40. The ACL variables $acl_xx are now saved with the message, and can be
3491 accessed later in routers, transports, and filters.
3493 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3494 strings in the file are not string-expanded.
3496 42. If a MAIL command specified a SIZE value that was too large to fit into an
3497 int variable, the check against message_size_limit failed. Such values are
3498 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3499 day this will have to be increased, but I don't think I want to be around
3500 when emails are that large.
3507 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3508 iconv() is not standard in FreeBSD.
3510 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3511 IPv6 enabled. The observed symptom was a segmentation fault on return from
3512 the function os_common_find_running_interfaces() in src/os.c.
3514 3. In the check_special_case() function in daemon.c I had used "errno" as an
3515 argument name, which causes warnings on some systems. This was basically a
3516 typo, since it was named "eno" in the comments!
3518 4. The code that waits for the clock to tick (at a resolution of some fraction
3519 of a second) so as to ensure message-id uniqueness was always waiting for
3520 at least one whole tick, when it could have waited for less. [This is
3521 almost certainly not relevant at current processor speeds, where it is
3522 unlikely to ever wait at all. But we try to future-proof.]
3524 5. The function that sleeps for a time interval that includes fractions of a
3525 second contained a race. It did not block SIGALRM between setting the
3526 timer, and suspending (a couple of lines later). If the interval was short
3527 and the sigsuspend() was delayed until after it had expired, the suspension
3528 never ended. On busy systems this could lead to processes getting stuck for
3531 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3532 process, before it forks any delivery processes. The open lookup caching
3533 mechanism meant that the open file or database connection was passed into
3534 the delivery process. The problem was that delivery processes always tidy
3535 up cached lookup data. This could cause a problem for the next delivery
3536 process started by the queue runner, because the external queue runner
3537 process does not know about the closure. So the next delivery process
3538 still has data in the lookup cache. In the case of a file lookup, there was
3539 no problem because closing a file descriptor in a subprocess doesn't affect
3540 the parent. However, if the lookup was caching a connection to a database,
3541 the connection was closed, and the second delivery process was likely to
3542 see errors such as "PGSQL: query failed: server closed the connection
3543 unexpectedly". The problem has been fixed by closing all cached lookups
3544 in a queue runner before running a delivery process.
3546 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3547 seem to have the "const" qualifier which it has on other OS. I've
3550 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3551 *of the same type* (client or server) with the same public name that an
3552 error should be diagnosed.
3554 9. When Exim looked up a host name for an IP address, but failed to find the
3555 original IP address when looking up the host name (a safety check), it
3556 output the message "<ip address> does not match any IP for NULL", which was
3557 confusing, to say the least. The bug was that the host name should have
3558 appeared instead of "NULL".
3560 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3561 user that is built into the binary, and the -C or -D options is used, root
3562 privilege is dropped before the configuration file is read. In addition,
3563 logging is switched to stderr instead of the normal log files. If the
3564 configuration then re-defines the Exim user, the unprivileged environment
3565 is probably not what is expected, so Exim logs a panic warning message (but
3568 However, if deliver_drop_privilege is set, the unprivileged state may well
3569 be exactly what is intended, so the warning has been cut out in that case,
3570 and Exim is allowed to try to write to its normal log files.
3576 1. smtp_return_error_details was not giving details for temporary sender
3577 or receiver verification errors.
3579 2. Diagnose a configuration error if two authenticators have the same public
3582 3. Exim used not to create the message log file for a message until the first
3583 delivery attempt. This could be confusing when incoming messages were held
3584 for policy or load reasons. The message log file is now created at the time
3585 the message is received, and an initial "Received" line is written to it.
3587 4. The automatically generated man page for command line options had a minor
3588 bug that caused no ill effects; however, a more serious problem was that
3589 the procedure for building the man page automatically didn't always
3590 operate. Consequently, release 4.20 contains an out-of-date version. This
3591 shouldn't happen again.
3593 5. When building Exim with embedded Perl support, the script that builds the
3594 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3595 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3597 6. The freeze_tell option was not being used for messages that were frozen on
3598 arrival, either by an ACL or by local_scan().
3600 7. Added the smtp_incomplete_transaction log selector.
3602 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3603 was accepting AUTH without a new EHLO.
3605 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3606 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3609 10. Logging of TCP/IP connections (when configured) now happens in the main
3610 daemon process instead of the child process, so that the TCP/IP connection
3611 count is more accurate (but it can never be perfect).
3613 11. The use of "drop" in a nested ACL was not being handled correctly in the
3614 outer ACL. Now, if condition failure induced by the nested "drop" causes
3615 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3616 or "require"), the connection is dropped.
3618 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3619 that yield "discard" can now be used with an "accept" or a "discard" verb,
3620 but an error is generated for any others (because I can't see a useful way
3621 to define what should happen).
3623 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3624 are now processed in the same way as lines in the Exim configuration file.
3625 In particular, continuation lines are supported.
3627 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3629 15. Added -ti meaning -t -i.
3631 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3632 domains, and warn by logging if others are found.
3634 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3635 returned from GIFCONF when Exim is trying to find the list of interfaces on
3636 a host. The code in os.c has been modified to copy each ifreq to an aligned
3637 structure in all cases.
3639 Also, in some cases, the returned ifreq's were being copied to a 'struct
3640 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3641 means the last couple of bytes of an IPv6 address could be chopped if the
3642 ifreq contained only a normal sockaddr (14 bytes storage).
3644 18. Named domain lists were not supported in the hosts_treat_as_local option.
3645 An entry such as +xxxx was not recognized, and was treated as a literal
3648 19. Ensure that header lines added by a DATA ACL are included in the reject log
3649 if the ACL subsequently rejects the message.
3651 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3652 MD5 (which is deprecated).
3654 21. When testing a filter file using -bf, Exim was writing a message when it
3655 took the sender from a "From " line in the message, but it was not doing so
3656 when it took $return_path from a Return-Path: header line. It now does.
3658 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3659 with a valid header line field name (a series of printing characters
3660 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3662 23. Changed "disc" in the source to "disk" to conform to the documentation and
3663 the book and for uniformity.
3665 24. Ignore Sendmail's -Ooption=value command line item.
3667 25. When execve() failed while trying to run a command in a pipe transport,
3668 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3669 could be confused with a return value of 69 from the command itself. This
3670 has been changed to 127, the value the shell returns if it is asked to run
3671 a non-existent command. The wording for the related log line suggests a
3672 non-existent command as the problem.
3674 26. If received_header_text expands to an empty string, do not add a Received:
3675 header line to the message. (Well, it adds a token one on the spool, but
3676 marks it "old" so that it doesn't get used or transmitted.)
3678 27. Installed eximstats 1.28 (addition of -nt option).
3680 28. There was no check for failure on the call to getsockname() in the daemon
3681 code. This can fail if there is a shortage of resources on the system, with
3682 ENOMEM, for example. A temporary error is now given on failure.
3684 29. Contrary to the C standard, it seems that in some environments, the
3685 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3686 program. Exim now does this explicitly; it affects the formatting of
3687 timestamps using strftime().
3689 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3690 complaints. I've now initialized all the variables, to avoid this.
3692 32. Header lines added by a system filter were not being "seen" during
3693 transport-time rewrites.
3695 33. The info_callback() function passed to OpenSSL is set up with type void
3696 (*)(SSL *, int, int), as described somewhere. However, when calling the
3697 function (actually a macro) that sets it up, the type void(*)() is
3698 expected. I've put in a cast to prevent warnings from picky compilers.
3700 34. If a DNS black list lookup found a CNAME record, but there were no A
3701 records associated with the domain it pointed at, Exim crashed.
3703 35. If a DNS black list lookup returned more than one A record, Exim ignored
3704 all but the first. It now scans all returned addresses if a particular IP
3705 value is being sought. In this situation, the contents of the
3706 $dnslist_value variable are a list of all the addresses, separated by a
3709 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3710 to accept a host name and all its aliases if the forward lookup for any of
3711 them yielded the IP address of the incoming connection. Now it accepts only
3712 those names whose forward lookup yields the correct IP address. Any other
3713 names are discarded. This closes a loophole whereby a rogue DNS
3714 administrator could create reverse DNS records to break through a
3715 wildcarded host restriction in an ACL.
3717 37. If a user filter or a system filter that ran in a subprocess used any of
3718 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3719 the wrong values were passed to the pipe command ($thisaddress had the
3720 value of $0, $0 had the value of $1, etc). This bug was introduced by
3721 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3723 38. Improved the line breaking for long SMTP error messages from ACLs.
3724 Previously, if there was no break point between 40 and 75 characters, Exim
3725 left the rest of the message alone. Two changes have been made: (a) I've
3726 reduced the minimum length to 35 characters; (b) if it can't find a break
3727 point between 35 and 75 characters, it looks ahead and uses the first one
3728 that it finds. This may give the occasional overlong line, but at least the
3729 remaining text gets split now.
3731 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3732 file descriptors might be low, and that setting 1000 would always raise it.
3733 It turns out that in some environments, the limit is already over 1000 and
3734 that lowering it causes trouble. So now Exim takes care not to decrease it.
3736 40. When delivering a message, the value of $return_path is set to $sender_
3737 address at the start of routing (routers may change the value). By an
3738 oversight, this default was not being set up when an address was tested by
3739 -bt or -bv, which affected the outcome if any router or filter referred to
3742 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3743 log only when "message" or "log_message" are set. However, if one of the
3744 conditions was an address verification, or a call to a nested ACL, the
3745 messages generated by the underlying test were being passed through. This
3746 no longer happens. The underlying message is available in $acl_verify_
3747 message for both "message" and "log_message" expansions, so it can be
3748 passed through if needed.
3750 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3751 new expansion $bh_ to give the encoded byte string without charset
3752 translation. Translation happens only if iconv() is available; HAVE_ICONV
3753 indicates this at build time. HEADERS_CHARSET gives the charset to
3754 translate to; headers_charset can change it in the configuration, and
3755 "headers charset" can change it in an individual filter file.
3757 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3758 that creates RFC 2047 encoded "words" labels them as that charset instead
3759 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3760 expansion operator; (ii) when Exim creates a From: line for a local
3761 message; (iii) when a header line is rewritten to include a "phrase" part.
3763 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3764 buggy, causing it to skip the first lines of messages whose message ID
3765 ended in 'D'. This would not have bitten before Exim release 4.14, because
3766 message IDs were unlikely to end in 'D' before then. The effect was to have
3767 incorrect size information for certain domains.
3769 45. #include "config.h" was missing at the start of the crypt16.c module. This
3770 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3773 46. If there was a timeout during a "random" callout check, Exim treated it as
3774 a failure of the random address, and carried on sending RSET and the real
3775 address. If the delay was just some slowness somewhere, the response to the
3776 original RCPT would be taken as a response to RSET and so on, causing
3777 mayhem of various kinds.
3779 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3780 when I implemented it. It didn't allow for the fact that some option values
3781 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3782 the right test for positive values.
3784 48. Domain names in DNS records are case-independent. Exim always looks them up
3785 in lower case. Some resolvers return domain names in exactly the case they
3786 appear in the zone file, that is, they may contain uppercase letters. Not
3787 all resolvers do this - some return always lower case. Exim was treating a
3788 change of case by a resolver as a change of domain, similar to a widening
3789 of a domain abbreviation. This triggered its re-routing code and so it was
3790 trying to route what was effectively the same domain again. This normally
3791 caused routing to fail (because the router wouldn't handle the domain
3792 twice). Now Exim checks for this case specially, and just changes the
3793 casing of the domain that it ultimately uses when it transmits the message
3796 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3799 50. If a filter generated a file delivery with a non-absolute name (possible if
3800 no home directory exists for the router), the forbid_file option was not
3803 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3804 the existing equality matching.
3806 52. Exim was using ints instead of ino_t variables in some places where it was
3807 dealing with inode numbers.
3809 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3810 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3811 called TMPDIR, and if it finds it is different, it changes its value.
3813 54. The smtp_printf() function is now made available to local_scan() so
3814 additional output lines can be written before returning. There is also an
3815 smtp_fflush() function to enable the detection of a dropped connection.
3816 The variables smtp_input and smtp_batched_input are exported to
3819 55. Changed the default runtime configuration: the message "Unknown user"
3820 has been removed from the ACL, and instead placed on the localuser router,
3821 using the cannot_route_message feature. This means that any verification
3822 failures that generate their own messages won't get overridden. Similarly,
3823 the "Unrouteable address" message that was in the ACL for unverifiable
3824 relay addresses has also been removed.
3826 56. Added hosts_avoid_esmtp to the smtp transport.
3828 57. The exicyclog script was not checking for the esoteric option
3829 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3830 will work only if exicyclog is run under the appropriate euid.
3832 58. Following a discussion on the list, the rules by which Exim recognises line
3833 endings on incoming messages have been changed. The -dropcr and drop_cr
3834 options are now no-ops, retained only for backwards compatibility. The
3835 following line terminators are recognized: LF CRLF CR. However, special
3836 processing applies to CR:
3838 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3839 nor a local message in the state where . is a terminator.
3841 (ii) If a bare CR is encountered in a header line, an extra space is added
3842 after the line terminator so as not to end the header. The reasoning
3843 behind this is that bare CRs in header lines are most likely either
3844 to be mistakes, or people trying to play silly games.
3846 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3847 was being incorrectly given as 18 bytes larger than it should have been.
3848 This is a VOB (very old bug).
3850 60. This may never have affected anything current, but just in case it has:
3851 When the local host is found other than at the start of a list of hosts,
3852 the local host, those with the same MX, and any that follow, are discarded.
3853 When the list in question was part of a longer list of hosts, the following
3854 hosts (not currently being processed) were also being discarded. This no
3855 longer happens. I'm not sure if this situation could ever has previously
3858 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3861 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3862 header. This is something that is recommended in a new Internet Draft, and
3863 is something that is documented as being done by Sendmail. There are two
3864 possible values. For messages generated by the autoreply transport, Exim
3867 Auto-Submitted: auto-replied
3869 whereas for all other generated messages (e.g. bounces) it adds
3871 Auto-Submitted: auto-generated
3873 63. The "personal" condition in filters now includes a test for the
3874 Auto-Submitted: header. If it contains the string "auto-" the message it
3875 not considered personal.
3877 64. Added rcpt_include_affixes as a generic transport option.
3879 65. Added queue_only_override (default true).
3881 66. Added the syslog_duplication option.
3883 67. If what should have been the first header line of a message consisted of
3884 a space followed by a colon, Exim was mis-interpreting it as a header line.
3885 It isn't of course - it is syntactically invalid and should therefore be
3886 treated as the start of the message body. The misbehaviour could have
3887 caused a number of strange effects, including loss of data in subsequent
3888 header lines, and spool format errors.
3890 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3891 client host had authenticated. This control can now be exercised by an ACL
3892 for more flexibility.
3894 69. By default, callouts do not happen when testing with -bh. There is now a
3895 variant, -bhc, which does actually run the callout code, including
3896 consulting and updating the callout cache.
3898 70. Added support for saslauthd authentication, courtesy of Alexander
3901 71. If statvfs() failed on the spool or log directories while checking their
3902 size for availability, Exim confusingly gave the error "space shortage".
3903 Furthermore, in debugging mode it crashed with a floating point exception.
3904 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3905 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3906 serious problem, Exim now writes to the main and panic logs when this
3907 happens, with details of the failure. It then refuses to accept the
3908 incoming message, giving the message "spool directory problem" or "log
3909 directory problem" with a 421 code for SMTP messages.
3911 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3912 0, 1, and 2 exist, because some OS complain for execs without them (see
3913 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3914 descriptors. However, the code omitted to check that the open succeeded,
3915 causing mysterious errors if for some reason the permissions on /dev/null
3916 got screwed. Now Exim writes a message to the main and panic logs, and
3917 bombs out if it can't open /dev/null.
3919 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3920 interact so that it is all more flexible. It is supposed to remain
3921 backwards compatible. Also added extra_local_interfaces.
3923 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3924 to bomb out with an assertion failure - to the client this appears as a
3925 connection drop. This problem occurs in the part of the code that was taken
3926 from the Samba project. Fortunately, the assertion is in a very simple
3927 function, so I have fixed this by reproducing the function inline in the
3928 one place where it is called, and arranging for authentication to fail
3929 instead of killing the process with assert().
3931 75. The SPA client code was not working when the server requested OEM rather
3932 than Unicode encoding.
3934 76. Added code to make require_files with a specific uid setting more usable in
3935 the case where statting the file as root fails - usually a non-root-mounted
3936 NFS file system. When this happens and the failure is EACCES, Exim now
3937 forks a subprocess and does the per-uid checking as the relevant uid.
3939 77. Added process_log_path.
3941 78. If log_file_path was not explicitly set, a setting of check_log_space or
3942 check_log_inodes was ignored.
3944 79. If a space check for the spool or log partitions fails, the incident is now
3945 logged. Of course, in the latter case the data may get lost...
3947 80. Added the %p formatting code to string_format() so that it can be used to
3948 print addresses in debug_print(). Adjusted all the address printing in the
3949 debugging in store.c to use %p rather than %d.
3951 81. There was a concern that a line of code in smtp_in.c could overflow a
3952 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3953 initially expressed, the concern was not well-founded, because trailing
3954 spaces are removed early. However, if the trailing spaces were followed by
3955 a NULL, they did not get removed, so the overflow was possible. Two fixes
3958 (a) I re-wrote the offending code in a cleaner fashion.
3959 (b) If an incoming SMTP command contains a NULL character, it is rejected
3962 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3963 runs initgroups(), so that if the Exim user is in any additional groups,
3964 they will be used during message reception.
3970 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob