1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.359 2006/06/27 14:34:26 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 SC/01 Use a glob alias rather than an array ref in eximstats generated
10 parser. This improves both readability and performance.
12 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
13 Don't display local sender or destination tables in eximstats unless
14 there is data to show.
15 Added average volumes into the eximstats top table text output.
17 SC/03 Collect data on the number of addresses (recipients) as well
18 as the number of messages in eximstats.
20 TF/01 Correct an error in the documentation for the redirect router. Exim
21 does (usually) call initgroups() when daemonizing.
23 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs with
24 consistent privilege compared to when running as a daemon.
26 TF/03 Note in the spec that $authenticated_id is not set for local
27 submissions from trusted users.
29 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
30 Thanks to Dean Brooks <dean@iglou.com> for the patch.
32 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
33 by adding some example configuration directives to the default
34 configuration file. A little bit of work is required to uncomment the
35 directives and define how usernames and passwords are checked, but
36 there is now a framework to start from.
38 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
39 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
40 without this. I don't know how relevant this is to other LDAP libraries.
42 PH/02 Add the verb name to the "unknown ACL verb" error.
44 PH/03 Magnus Holmgren's patch for filter_prepend_home.
50 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
51 other effects) broke the use of negated acl sub-conditions.
53 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
56 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
57 "Deny" causes Exim to reject the incoming connection with a 554 error.
58 Unfortunately, if there is a major crisis, such as a disk failure,
59 tcp-wrappers gives "deny", whereas what one would like would be some
60 kind of temporary error. A kludge has been added to help with this.
61 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
62 554 error is used if errno is still zero or contains ENOENT (which occurs
63 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
66 PH/03 Add -lutil to the default FreeBSD LIBS setting.
68 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
69 errors. Otherwise a message that provokes a temporary error (when other
70 messages do not) can cause a whole host to time out.
72 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
73 addresses were routed directly to files or pipes from a redirect router.
74 File deliveries just didn't batch; pipe deliveries might have suffered
77 PH/06 A failure to get a lock for a hints database would erroneously always say
78 "Failed to get write lock", even when it was really a read lock.
80 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
81 of 0600. This has been changed to use the value of the lockfile_mode
82 option (which defaults to 0600).
84 PH/08 Applied small patch from the Sieve maintainer.
86 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
87 folder from quota calculations, a direct delivery into this folder messed
88 up the contents of the maildirsize file. This was because the regex was
89 used only to exclude .Trash (or whatever) when the size of the mailbox
90 was calculated. There was no check that a delivery was happening into an
91 excluded directory. This bug has been fixed by ignoring all quota
92 processing for deliveries into excluded directories.
94 PH/10 Added the maildirfolder_create_regex option to appendfile.
100 PH/01 The code for finding all the local interface addresses on a FreeBSD
101 system running IPv6 was broken. This may well have applied to all BSD
102 systems, as well as to others that have similar system calls. The broken
103 code found IPv4 interfaces correctly, but gave incorrect values for the
104 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
105 that it would not match correctly against @[] and not recognize the IPv6
108 PH/02 The ipliteral router was not recognizing addresses of the form user@
109 [ipv6:....] because it didn't know about the "ipv6:" prefix.
111 PH/03 Added disable_ipv6.
113 PH/04 Changed $reply_address to use the raw form of the headers instead of the
114 decoded form, because it is most often used to construct To: headers
115 lines in autoreplies, and the decoded form may well be syntactically
116 invalid. However, $reply_address has leading white space removed, and all
117 newlines turned into spaces so that the autoreply transport does not
120 PH/05 If group was specified without a user on a router, and no group or user
121 was specified on a transport, the group from the router was ignored.
123 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
124 for visible compile-time settings that can be used to change these
125 numbers, for those that want even more. Backwards compatibility with old
126 spool files has been maintained. However, going back to a previous Exim
127 release will lost any variables that are in spool files.
129 PH/07 Two small changes when running in the test harness: increase delay when
130 passing a TCP/IP connection to a new process, in case the original
131 process has to generate a bounce, and remove special handling of
132 127.0.0.2 (sic), which is no longer necessary.
134 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
135 be the same on different OS.
137 PH/09 Moved a debug statement in filter processing to avoid a race problem when
140 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
141 whether --show-vars was specified or not
143 JJ/02 exipick: Added support for new ACL variable spool format introduced
146 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
147 syntactically invalid From: or Reply-to: line, and a filter used this to
148 generate an autoreply, and therefore failed to obtain an address for the
149 autoreply, Exim could try to deliver to a non-existent relative file
150 name, causing unrelated and misleading errors. What now happens is that
151 it logs this as a hard delivery error, but does not attempt to create a
154 PH/11 The exinext utility has a -C option for testing purposes, but although
155 the given file was scanned by exinext itself; it wasn't being passed on
156 when Exim was called.
158 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
159 an end-of-file indication when reading a command response.
161 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
162 compiled. In many other places in Exim, IPv6 addresses are always
163 recognized, so I have changed this. It also means that IPv4 domain
164 literals of the form [IPV4:n.n.n.n] are now always recognized.
166 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
167 used if the router is not running as root, for example, when verifying at
168 ACL time, or when using -bh. The debugging output from this situation was
169 non-existent - all you got was a failure to exec. I have made two
172 (a) Failures to set uid/gid, the current directory, or a process leader
173 in a subprocess such as that created by queryprogram now generate
174 suitable debugging ouput when -d is set.
176 (b) The queryprogram router detects when it is not running as root,
177 outputs suitable debugging information if -d is set, and then runs
178 the subprocess without attempting to change uid/gid.
180 PH/15 Minor change to Makefile for building test_host (undocumented testing
183 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
184 additional section of a DNS packet that returns MX or SRV records.
185 Instead, it always explicitly searches for A/AAAA records. This avoids
186 major problems that occur when a DNS server includes only records of one
187 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
188 fixed another bug: if SRV records were looked up and the corresponding
189 address records were *not* found in the additional section, the port
190 values from the SRV records were lost.
192 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
193 using the correct key (the original address) when searching the retry
194 rules in order to find which one to use for generating the retry hint.
196 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
197 adding the default one. Similarly, if it contains a Reply-To: header, the
198 errors_reply_to option, if set, is not used.
200 PH/19 When calculating a retry time, Exim used to measure the "time since
201 failure" by looking at the "first failed" field in the retry record. Now
202 it does not use this if it is later than than the arrival time of the
203 message. Instead it uses the arrival time. This makes for better
204 behaviour in cases where some deliveries succeed, thus re-setting the
205 "first failed" field. An example is a quota failure for a huge message
206 when small messages continue to be delivered. Without this change, the
207 "time since failure" will always be short, possible causing more frequent
208 delivery attempts for the huge message than are intended.
209 [Note: This change was subsequently modified - see PH/04 for 4.62.]
211 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
212 $1, $2, $3) because the numerical variables can be reset during some
213 expansion items (e.g. "match"), thereby losing the authentication data.
215 PH/21 Make -bV show the size of off_t variables so that the test suite can
216 decide whether to run tests for quotas > 2G.
218 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
219 mailbox_size, and mailbox_filecount in the appendfile transport. If a
220 filecount value is greater than 2G or if a quota value is greater than 2G
221 on a system where the size of off_t is not greater than 4, a panic error
224 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
225 never match. The debug and -bh output now contains an explicit error
226 message indicating a malformed IPv4 address or mask.
228 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
229 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
232 PH/25 Do not write to syslog when running in the test harness. The only
233 occasion when this arises is a failure to open the main or panic logs
234 (for which there is an explicit test).
236 PH/26 Added the /no_tell option to "control=freeze".
238 PH/27 If a host name lookup failed very early in a connection, for example, if
239 the IP address matched host_lookup and the reverse lookup yielded a name
240 that did not have a forward lookup, an error message of the form "no IP
241 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
242 could be logged. Now it outputs the IP address instead of "NULL".
244 PH/28 An enabling patch from MH: add new function child_open_exim2() which
245 allows the sender and the authenticated sender to be set when
246 submitting a message from within Exim. Since child_open_exim() is
247 documented for local_scan(), the new function should be too.
249 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
250 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
251 results in an empty string is now treated as unset.
253 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
255 PH/31 Added sender_verify_fail logging option.
257 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
258 needed by RFC 822 but not by RFC 2822 was commented out. I have now
259 tidied the source and removed it altogether.
261 PH/33 When a queue run was abandoned because the load average was too high, a
262 log line was always written; now it is written only if the queue_run log
263 selector is set. In addition, the log line for abandonment now contains
264 information about the queue run such as the pid. This is always present
265 in "start" and "stop" lines but was omitted from the "abandon" line.
267 PH/34 Omit spaces between a header name and the colon in the error message that
268 is given when verify = headers_syntax fails (if there are lots of them,
269 the message gets confusing).
271 PH/35 Change the default for dns_check_names_pattern to allow slashes within
272 names, as there are now some PTR records that contain slashes. This check
273 is only to protect against broken name servers that fall over on strange
274 characters, so the fact that it applies to all lookups doesn't matter.
276 PH/36 Now that the new test suite is complete, we can remove some of the
277 special code in Exim that was needed for the old test suite. For example,
278 sorting DNS records because real resolvers return them in an arbitrary
279 order. The new test suite's fake resolver always returns records in the
282 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
283 bounces) except when queue_only is set, to avoid logging races between
284 the different processes.
286 PH/38 Panic-die if .include specifies a non-absolute path.
288 PH/39 A tweak to the "H" retry rule from its user.
290 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
291 a label. They prevented compilation on older perls.
293 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
294 a warning to be raised on newish perls.
296 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
297 on queue. Changes to match documented behaviour of showing count of
298 messages matching specified criteria.
300 PH/40 Changed the default ident timeout from 30s to 5s.
302 PH/41 Added support for the use of login_cap features, on those BSD systems
303 that have them, for controlling the resources used by pipe deliveries.
305 PH/42 The content-scanning code uses fopen() to create files in which to put
306 message data. Previously it was not paying any attention to the mode of
307 the files. Exim runs with umask(0) because the rest of the code creates
308 files with open(), and sets the required mode explicitly. Thus, these
309 files were ending up world-writeable. This was not a big issue, because,
310 being within the spool directory, they were not world-accessible. I have
311 created a function called modefopen, which takes an additional mode
312 argument. It sets umask(777), creates the file, chmods it to the required
313 mode, then resets the umask. All the relevant calls to fopen() in the
314 content scanning code have been changed to use this function.
316 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
317 to 24 hours. This avoids potential overflow problems when processing G
318 and H retry rules. I suspect nobody ever tinkers with this value.
320 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
322 PH/45 When the plaintext authenticator is running as a client, the server's
323 challenges are checked to ensure they are valid base64 strings. By
324 default, the authentication attempt is cancelled if an invalid string is
325 received. Setting client_ignore_invalid_base64 true ignores these errors.
326 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
327 they are received. Thus, the responses can be made to depend on the
328 challenges. If an invalid string is ignored, an empty string is placed in
331 PH/46 Messages that are created by the autoreply transport now contains a
332 References: header, in accordance with RFCs 2822 and 3834.
334 PH/47 Added authenticated_sender_force to the smtp transport.
336 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
338 PH/49 Installed latest patch from the Sieve maintainer.
340 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
341 was also set, the appendfile transport was unnecessarily scanning a
342 directory of message files (e.g. for maildir delivery) to find the count
343 of files (along with the size), even though it did not need this
344 information. It now does the scan only if it needs to find either the
345 size of the count of files.
347 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
349 PH/52 Two bugs concerned with error handling when the smtp transport is
352 (i) Exim was not creating retry information for temporary errors given
353 for individual recipients after the DATA command when the smtp transport
354 was used in LMTP mode. This meant that they could be retried too
355 frequently, and not timed out correctly.
357 (ii) Exim was setting the flag that allows error details to be returned
358 for LMTP errors on RCPT commands, but not for LMTP errors for individual
359 recipients that were returned after the DATA command.
361 PH/53 This is related to PH/52, but is more general: for any failing address,
362 when detailed error information was permitted to be returned to the
363 sender, but the error was temporary, then after the final timeout, only
364 "retry timeout exceeded" was returned. Now it returns the full error as
365 well as "retry timeout exceeded".
367 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
368 do this, and (what is worse) MTAs that accept it.
370 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
371 will now be deprecated.
373 PH/56 New os.c-cygwin from the Cygwin maintainer.
375 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
376 formats (previously only available in exim formats via -bpr, -bpru,
377 and -bpra. Now also available in native and exiqgrep formats)
379 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
380 with very large, slow to parse queues
382 JJ/08 exipick: added ! as generic prefix to negate any criteria format
384 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
386 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
387 responses to authentication challenges, though it was showing the
388 challenges; (ii) I've removed the CR characters from the debug output for
391 PH/58 Allow for the insertion of a newline as well as a space when a string
392 is turned into more than one encoded-word during RFC 2047 encoding. The
393 Sieve code now uses this.
395 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
396 data_4xx, lost_connection, tls_required.
398 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
399 message was being sent as an SMTP response.
401 PH/61 Add -l and -k options to exicyclog.
403 PH/62 When verifying, if an address was redirected to one new address, so that
404 verification continued, and the new address failed or deferred after
405 having set something in $address_data, the value of $address_data was not
406 passed back to the ACL. This was different to the case when no
407 redirection occurred. The value is now passed back in both cases.
409 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
410 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
411 use login_cap.h, so on its own it isn't the distinguishing feature. The
412 new name refers directly to the setclassresources() function.
414 PH/65 Added configuration files for NetBSD3.
416 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
418 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
419 is preferred over IPv4.
421 PH/68 The bounce_return_message and bounce_return_body options were not being
422 honoured for bounces generated during the reception of non-SMTP messages.
423 In particular, this applied to messages rejected by the ACL. This bug has
424 been fixed. However, if bounce_return_message is true and bounce_return_
425 body is false, the headers that are returned for a non-SMTP message
426 include only those that have been read before the error was detected.
427 (In the case of an ACL rejection, they have all been read.)
429 PH/69 The HTML version of the specification is now built in a directory called
430 spec_html instead of spec.html, because the latter looks like a path with
431 a MIME-type, and this confuses some software.
433 PH/70 Catch two compiler warnings in sieve.c.
435 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
436 function verify_get_ident() calls ip_connect() to connect a socket, but
437 if the "connect()" function timed out, ip_connect() used to close the
438 socket. However, verify_get_ident() also closes the socket later, and in
439 between Exim writes to the log, which may get opened at this point. When
440 the socket was closed in ip_connect(), the log could get the same file
441 descriptor number as the socket. This naturally causes chaos. The fix is
442 not to close the socket in ip_connect(); the socket should be closed by
443 the function that creates it. There was only one place in the code where
444 this was missing, in the iplookup router, which I don't think anybody now
445 uses, but I've fixed it anyway.
447 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
448 well as to direct DNS lookups. Otherwise the handling of names in host
449 lists is inconsistent and therefore confusing.
455 PH/01 Two changes to the default runtime configuration:
457 (1) Move the checks for relay_from_hosts and authenticated clients from
458 after to before the (commented out) DNS black list checks.
460 (2) Add control=submission to the relay_from_hosts and authenticated
461 clients checks, on the grounds that messages accepted by these
462 statements are most likely to be submissions.
464 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
466 (1) Generate an error if the third argument for the ${prvs expansion is
469 (2) Treat a missing third argument of ${prvscheck as if it were an empty
472 (3) Reset the variables that are obtained from the first argument of
473 ${prvscheck and used in the second argument before leaving the code,
474 because their memory is reclaimed, so using them afterwards may do
477 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
478 one (it's much easier than Tom thought :-).
480 (5) Because of (4), we can now allow for the use of $prvscheck_result
481 inside the third argument.
483 PH/03 For some reason, the default setting of PATH when running a command from
484 a pipe transport was just "/usr/bin". I have changed it to
487 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
488 anything to be listed in the output from -bV.
490 PH/05 When a filter generated an autoreply, the entire To: header line was
491 quoted in the delivery log line, like this:
493 => >A.N.Other <ano@some.domain> <original@ddress> ...
495 This has been changed so that it extracts the operative address. There
496 may be more than one such address. If so, they are comma-separated, like
499 => >ano@some.domain,ona@other.domain <original@ddress> ...
501 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
502 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
503 not being looked up in the rDNS to get a host name, Exim was showing the
504 IP address twice in Received: lines, even though the IP addresses were
505 identical. For example:
507 Received: from [1.2.3.4] (helo=[1.2.3.4])
509 However, if the real host name was known, it was omitting the HELO data
510 if it matched the actual IP address. This has been tidied up so that it
511 doesn't show the same IP address twice.
513 PH/07 When both +timestamp and +memory debugging was on, the value given by
514 $tod_xxx expansions could be wrong, because the tod_stamp() function was
515 called by the debug printing, thereby overwriting the timestamp buffer.
516 Debugging no longer uses the tod_stamp() function when +timestamp is set.
518 PH/08 When the original message was included in an autoreply transport, it
519 always said "this is a copy of the message, including all the headers",
520 even if body_only or headers_only was set. It now gives an appropriate
523 PH/09 Applied a patch from the Sieve maintainer which:
525 o fixes some comments
526 o adds the (disabled) notify extension core
527 o adds some debug output for the result of if/elsif tests
528 o points to the current vacation draft in the documentation
529 and documents the missing references header update
533 o fixes a bug in processing the envelope test (when testing
534 multiple envelope elements, the last element determinted the
537 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
538 Electronic Mail") by including:
540 Auto-submitted: auto-generated
542 in the messages that it generates (bounce messages and others, such as
543 warnings). In the case of bounce messages for non-SMTP mesages, there was
544 also a typo: it was using "Auto_submitted" (underscore instead of
545 hyphen). Since every message generated by Exim is necessarily in response
546 to another message, thes have all been changed to:
548 Auto-Submitted: auto-replied
550 in accordance with these statements in the RFC:
552 The auto-replied keyword:
554 - SHOULD be used on messages sent in direct response to another
555 message by an automatic process,
557 - MUST NOT be used on manually-generated messages,
559 - MAY be used on Delivery Status Notifications (DSNs) and Message
560 Disposition Notifications (MDNs),
562 - MUST NOT be used on messages generated by automatic or periodic
563 processes, except for messages which are automatic responses to
566 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
567 to the default Received: header definition.
569 PH/12 Added log selector acl_warn_skipped (default on).
571 PH/13 After a successful wildlsearch lookup, discard the values of numeric
572 variables because (a) they are in the wrong storage pool and (b) even if
573 they were copied, it wouldn't work properly because of the caching.
575 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
576 checking when decoding. Apparently there are clients that generate
577 overlong encoded strings. Why am I not surprised?
579 PH/15 If the first argument of "${if match_address" was not empty, but did not
580 contain an "@" character, Exim crashed. Now it writes a panic log message
581 and treats the condition as false.
583 PH/16 In autoreply, treat an empty string for "once" the same as unset.
585 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
586 extension "envelope-auth". The code is finished and in agreement with
587 other implementations, but there is no documentation so far and in fact,
588 nobody wrote the draft yet. This extension is currently #undef'ed, thus
589 not changing the active code.
591 Print executed "if" and "elsif" statements when debugging is used. This
592 helps a great deal to understand what a filter does.
594 Document more things not specified clearly in RFC3028. I had all this
595 sorted out, when out of a sudden new issues came to my mind. Oops."
597 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
600 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
601 canonical form (as documented). However, after a host name lookup from
602 the IP address, check_host() was doing a simple string comparison with
603 addresses acquired from the DNS when checking that the found name did
604 have the original IP as one of its addresses. Since any found IPv6
605 addresses are likely to be in abbreviated form, the comparison could
606 fail. Luckily, there already exists a function for doing the comparison
607 by converting both addresses to binary, so now that is used instead of
610 PH/20 There was another similar case to PH/19, when a complete host name was
611 given in a host list; looking up its IP address could give an abbreviated
612 form, whereas the current host's name might or might not be abbreviated.
613 The same fix has been applied.
619 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
620 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
623 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
624 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
626 PH/03 Typo: missing ".o" in src/pcre/Makefile.
628 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
629 header line, restrict the check to what is listed in RFCs 2369 and 2929.
630 Also, for "Auto-Submitted", treat anything other than "no" as
631 non-personal, in accordance with RFC 3834. (Previously it treated
632 anything starting "auto-" as non-personal.)
634 TF/01 The control=submission/name=... option had a problem with syntax
635 errors if the name included a slash character. The /name= option
636 now slurps the rest of the string, so it can include any characters
637 but it must come last in the list of options (after /sender_retain
640 PH/05 Some modifications to the interface to the fake nameserver for the new
648 TK/01 Added the "success_on_redirect" address verification option. See
649 NewStuff for rationale and an example.
651 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
653 PH/02 Patch to exigrep to allow it to work on syslog lines.
655 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
656 fread() to skip over the body file's header line, because in Cygwin the
657 header line is locked and is inaccessible.
659 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
660 co-exist for some time) to make it clear that it is the Exim ID that is
661 referenced, not the Message-ID: header line.
663 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
664 string_format() function, because snprintf() does not exist on all
667 PH/06 The use of forbid_filter_existstest now also locks out the use of the
668 ${stat: expansion item.
670 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
671 protocol synchronization error", to keep the pedants happy.
673 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
674 well as for IRIX systems, when gcc is being used. See the host.c source
677 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
679 PH/10 Named domain lists were not working if used in a queue_smtp_domains
682 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
683 transport and to the smtp transport in LMTP mode.
685 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
687 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
688 run a filter in a subprocess. This could lead to confusion in subsequent
689 lookups in the parent process. There should also be a search_tidyup() at
690 the end of the subprocess.
692 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
693 only if the host matched helo_try_verify_hosts, which caused the
694 verification to occur when the EHLO/HELO command was issued. The ACL just
695 tested the remembered result. Now, if a previous verification attempt has
696 not happened, "verify = helo" does it there and then.
698 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
700 TK/03 Fix log output including CR from clamd.
702 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
703 exist provoked a memory error which could cause a segfault.
705 PH/15 Installed PCRE 6.2
707 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
709 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
710 of the problem. Specifically, suggested +O2 rather than +O1 for the
713 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
715 PH/20 If a delivery was routed to a non-standard port by means of an SRV
716 record, the port was not correctly logged when the outgoing_port log
717 selector was set (it logged the transort's default port).
719 PH/21 Added support for host-specific ports to manualroute, queryprogram,
720 fallback_hosts, and "hosts" in the smtp transport.
722 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
723 host errors such as "Connection refused".
725 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
726 authentication with radiusclient 0.4.9:
728 - Error returned from rc_read_config was caught wrongly
729 - Username/password not passed on to radius server due to wrong length.
731 The presumption is that some radiusclient API changes for 4.51/PH/17
732 were not taken care of correctly. The code is still untested by me (my
733 Linux distribution still has 0.3.2 of radiusclient), but it was
734 contributed by a Radius user.
736 PH/24 When doing a callout, the value of $domain wasn't set correctly when
737 expanding the "port" option of the smtp transport.
739 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
740 while reading a MIME header. Thanks to Tom Hughes for a patch.
742 PH/24 Include config.h inside local_scan.h so that configuration settings are
745 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
746 that in an ACL for RCPT (for example), you can examine exactly what was
749 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
750 commands, but it was not correctly comparing the address with the actual
751 client host address. Thus, it would show the EHLO address in Received:
752 header lines when this was not necessary.
754 PH/27 Added the % operator to ${eval:}.
756 PH/28 Exim tries to create and chdir to its spool directory when it starts;
757 it should be ignoring failures (because with -C, for example, it has lost
758 privilege). It wasn't ignoring creation failures other than "already
761 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
764 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
765 because an input line was too long, either on its own, or by virtue of
766 too many continuations, the temporary file was not being removed, and the
767 return code was incorrect.
769 PH/31 Missing "BOOL" in function definition in filtertest.c.
771 PH/32 Applied Sieve patches from the maintainer.
773 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
775 PH/33 Added "verify = not_blind".
777 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
778 Local/Makefile (with some defaults set). These are used in built scripts
779 such as exicyclog, but they have never been used in the exim_install
780 script (though there are many overriding facilities there). I have
781 arranged that the exim_install script now takes note of these two
784 PH/35 Installed configuration files for Dragonfly.
786 PH/36 When a locally submitted message by a trusted user did not contain a
787 From: header, and the sender address was obtained from -f or from an SMTP
788 MAIL command, and the trusted user did not use -F to supply a sender
789 name, $originator_name was incorrectly used when constructing a From:
790 header. Furthermore, $originator_name was used for submission mode
791 messages from external hosts without From: headers in a similar way,
792 which is clearly wrong.
794 PH/37 Added control=suppress_local_fixups.
796 PH/38 When log_selector = +received_sender was set, and the addition of the
797 sender made the log line's construction buffer exactly full, or one byte
798 less than full, an overflow happened when the terminating "\n" was
801 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
802 when the result of a list match is failure because a DNS lookup failed.
804 PH/40 RM_COMMAND is now used in the building process.
806 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
807 the "build-* directories that it finds.
809 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
810 address in a domain literal was a prefix of an interface address.
812 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
813 when verifying a sender address, unless rewrite_headers is false.
815 PH/44 Wrote a long comment about why errors_to addresses are verified as
816 recipients, not senders.
818 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
819 the ratelimit ACL was added.
821 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
823 PH/46 Added extra information about PostgreSQL errors to the error string.
825 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
826 avoiding the need to install special zones in a real server. This is
827 backwards compatible; if it can't find the fake resolver, it drops back.
828 Thus, both old and new test suites can be run.
830 TF/02 Added util/ratelimit.pl
832 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
833 clock is set back in time.
835 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
836 Candler <B.Candler@pobox.com>.
838 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
839 OK for addresses that are the result of redirections.
841 PH/48 A number of further additions for the benefit of the new test suite,
842 including a fake gethostbyname() that interfaces to the fake DNS resolver
845 TF/06 The fix for widen_domains has also been applied to qualify_single and
846 search_parents which are the other dnslookup options that can cause
849 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
852 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
854 TF/07 Exim produced the error message "an SRV record indicated no SMTP
855 service" if it encountered an MX record with an empty target hostname.
856 The message is now "an MX or SRV record indicated no SMTP service".
858 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
859 if the DNS of the sending site is misconfigured. This is quite a
860 common situation. This change restores the behaviour of treating a
861 helo verification defer as a failure.
863 PH/51 If self=fail was set on a router, the bounce message did not include the
864 actual error message.
870 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
872 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
873 command itself ended in error, the underlying message about the transport
874 filter timeout was being overwritten with the pipe command error. Now the
875 underlying error message should be appended to the second error message.
877 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
879 PH/02 Reduce the amount of output that "make" produces by default. Full output
880 can still be requested.
882 PH/03 The warning log line about a condition test deferring for a "warn" verb
883 was being output only once per connection, rather than after each
884 occurrence (because it was using the same function as for successful
885 "warn" verbs). This seems wrong, so I have changed it.
887 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
888 it should not have, which might have caused a crash in the right
889 circumstances, but probably never did.
891 PH/04 Installed a modified version of Tony Finch's patch to make submission
892 mode fix the return path as well as the Sender: header line, and to
893 add a /name= option so that you can make the user's friendly name appear
896 TF/03 Added the control = fakedefer ACL modifier.
898 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
899 Mark Lowes for thorough testing.
901 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
903 TK/03 Merged latest SRS patch from Miles Wilton.
905 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
906 with the definition in sysexits.h (which is #included earlier).
907 Fortunately, Exim does not actually use EX_OK. The code used to try to
908 preserve the sysexits.h value, by assumimg that macro definitions were
909 scanned for macro replacements. I have been disabused of this notion,
910 so now the code just undefines EX_OK before #including unistd.h.
912 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
913 in the smtp transport. When a block could not be written in a single
914 write() function, the timeout was being re-applied to each part-write.
915 This seems wrong - if the receiver was accepting one byte at a time it
916 would take for ever. The timeout is now adjusted when this happens. It
917 doesn't have to be particularly precise.
919 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
920 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
922 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
923 without a domain if the check to <postmaster@domain> fails.
925 SC/01 Eximstats: added -xls and the ability to specify output files
926 (patch written by Frank Heydlauf).
928 SC/02 Eximstats: use FileHandles for outputing results.
930 SC/03 Eximstats: allow any combination of xls, txt, and html output.
932 SC/04 Eximstats: fixed display of large numbers with -nvr option
934 SC/05 Eximstats: fixed merging of reports with empty tables.
936 SC/06 Eximstats: added the -include_original_destination flag
938 SC/07 Eximstats: removed tabs and trailing whitespace.
940 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
942 TK/06 MBOX spool code: Add real "From " MBOX separator line
943 so the .eml file is really in mbox format (even though
944 most programs do not really care). Patch from Alex Miller.
946 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
947 The latter is generated from $received_to and is only set if the
948 message has one envelope recipient. SA can use these headers,
949 obviously out-of-the-box. Patch from Alex Miller.
951 PH/08 The ${def test on a variable was returning false if the variable's
952 value was "0", contrary to what the specification has always said!
953 The result should be true unless the variable is empty.
955 PH/09 The syntax error of a character other than { following "${if
956 def:variable_name" (after optional whitespace) was not being diagnosed.
957 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
958 accidental colon was present, for example, could give incorrect results.
960 PH/10 Tidied the code in a number of places where the st_size field of a stat()
961 result is used (not including appendfile, where other changes are about
964 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
965 This involved changing a lot of size variables from int to off_t. It
966 should work with maildirs and everything.
968 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
969 spamd dying while we are connected to it.
971 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
974 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
975 function for BDB 4.3.
977 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
979 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
980 the sources for PCRE have been re-arranged and more files are now
983 PH/15 The code I had for printing potentially long long variables in PH/11
984 above was not the best (it lost precision). The length of off_t variables
985 is now inspected at build time, and an appropriate printing format (%ld
986 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
987 to be "long long int" or "long int". This is needed for the internal
988 formatting function string_vformat().
990 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
991 the configuration file to be ":syslog", then the script "guesses" where
992 the logs files are, rather than using the compiled in default. In our
993 case the guess is not the same as the compiled default, so the script
994 suddenly stopped working when I started to use syslog. The patch checks
995 to see if log_file_path is "". If so, it attempts to read it from exim
996 with no configuration file to get the compiled in version, before it
997 falls back to the previous guessing code."
999 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1000 implementing BATV in an Exim configuration. See NewStuff for the gory
1003 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1004 Makefile that are specific to HP-UX.
1006 PH/18 If the "use_postmaster" option was set for a recipient callout together
1007 with the "random" option, the postmaster address was used as the MAIL
1008 FROM address for the random test, but not for the subsequent recipient
1009 test. It is now used for both.
1011 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1012 patch removes a few documentation additions to RFC 3028, because the
1013 latest draft now contains them. It adds the new en;ascii-case comparator
1014 and a new error check for 8bit text in MIME parts. Comparator and
1015 require names are now matched exactly. I enabled the subaddress
1016 extension, but it is not well tested yet (read: it works for me)."
1018 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1019 rework some of the code of TK/09 above to avoid the hardwired use of
1020 "%lld" and "long long". Replaced the call to snprintf() with a call to
1023 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1024 records point to non-existent hosts", "retry timeout exceeded", and
1025 "retry time not reached for any host after a long failure period".
1027 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1028 experimental DomainKeys support:
1030 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1031 (2) On an error such as an illegally used "control", the wrong name for
1032 the control was given.
1034 These problems did NOT occur unless DomainKeys support was compiled.
1036 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1038 PH/24 Added ${if match_ip condition.
1040 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1041 databases so that it will be absolutely obvious if a crash occurs in the
1042 DB library. This is a regular occurrence (often caused by mis-matched
1045 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1046 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1047 picked up on a user's system that detects such things. There doesn't seem
1048 to be a gcc warning option for this - only an attribute that has to be
1049 put on the function's prototype. It seems that in Fedora Core 4 they have
1050 set this on a number of new functions. No doubt there will be more in due
1053 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1054 specify a transport. However, if an address that was verified by such a
1055 router was the subject of a callout, Exim crashed because it tried to
1056 read the rcpt_include_affixes from the non-existent transport. Now it
1057 just assumes that the setting of that option is false. This bug was
1058 introduced by 4.51/PH/31.
1060 PH/28 Changed -d+all to exclude +memory, because that information is very
1061 rarely of interest, but it makes the output a lot bigger. People tend to
1062 do -d+all out of habit.
1064 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1065 code in os-type was giving problems when libc.so lives in lib64, like on
1068 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1069 aren't the modern standard, and it seems that some systems' include files
1070 don't always have them. Exim was already checking for some of the newer
1071 ones like T_AAAA, and defining it itself. I've added checks for all the
1072 record types that Exim uses.
1074 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1075 not automatically generating a new one, as it is supposed to. This
1076 prevented TLS from working. If the file did exist, but contained invalid
1077 data, a new version was generated, as expected. It was only the case of a
1078 non-existent file that was broken.
1080 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1081 with a change in libdomainkeys > 0.64.
1083 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1084 from DNS. If the selector record carries the flag, it now has
1085 precedence over the domain-wide flag.
1087 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1089 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1090 the use of a port name that isn't defined in /etc/services) occurred, the
1091 message was deferred as in a normal delivery, and thus remained on the
1092 spool, instead of being failed because of the mua_wrapper setting. This
1093 is now fixed, and I tidied up some of the mua_wrapper messages at the
1096 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1097 the messages in a hash of arrays rather than using individual hashes.
1098 This is a bit cleaner and results in dramatic memory savings, albeit
1099 at a slight CPU cost.
1101 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1102 as requested by Marc Sherman.
1104 SC/10 Eximstats: added histograms for user specified patterns as requested
1107 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1109 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1110 fopen() in the content-scanning modules that did not already have it.
1116 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1117 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1119 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1121 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1123 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1124 to test Sieve filters that use "vacation".
1126 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1127 that changes the way the GnuTLS parameters are stored in the cache file.
1128 The new format can be generated externally. For backward compatibility,
1129 if the data in the cache doesn't make sense, Exim assumes it has read an
1130 old-format file, and it generates new data and writes a new file. This
1131 means that you can't go back to an older release without removing the
1134 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1135 work if there are any delivery delays because "one_time" forces the
1136 parent to be marked "delivered", so its unseen clone is never tried
1137 again. For this reason, Exim now forbids the simultaneous setting of
1140 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1141 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1142 entry if you want to know the details. The fix, however, neglected to
1143 consider the case when local delivery batching is involved. The test for
1144 "previously delivered" was not happening when checking to see if an
1145 address could be batched with a previous (undelivered) one; under
1146 certain circumstances this could lead to multiple deliveries to the same
1149 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1150 in its include files, and this causes problems building Exim.
1152 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1153 header_syntax) but Exim was just ignoring anything given after a slash.
1154 In particular, this caused confusion with an attempt to use "verify =
1155 reverse_host_lookup/defer_ok". An error is now given when options are
1156 supplied for verify items that do not have them. (Maybe reverse_host_
1157 lookup should have a defer_ok option, but that's a different point.)
1159 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1160 defined by RFC 821) to 2048, because there were problems with some AUTH
1161 commands, and RFC 1869 says the size should be increased for extended
1162 SMTP commands that take arguments.
1164 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1167 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1168 "unknown" error; now it says that the functionality isn't in the binary.
1170 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1171 an address' error message when a string expansion fails (syntax or
1172 whatever). Otherwise the password may appear in the log. Following change
1173 PH/42 below, there is no longer a chance of it appearing in a bounce
1176 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1178 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1179 of its addresses was ever tried. (Bugzilla bug #2.)
1181 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1182 the result incorrectly in the debug output. (It correctly added a newline
1183 to what was transported.)
1185 TF/01 Added $received_time.
1187 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1188 commented out examples of how to interface to a virus scanner and to
1189 SpamAssassin. Also added commented examples of av_scanner and
1190 spamd_address settings.
1192 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1193 and controls are allowed in which ACLs. There were a couple of minor
1194 errors. Some of the entries in the conditions table (which is a table of
1195 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1196 negation of where the condition IS allowed.
1198 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1200 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1201 header file does not have a version number, so I've had to invent a new
1202 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1203 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1204 radiusclient), but it was contributed by a Radius user.
1206 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1207 files or directories, for OpenSSL.
1209 PH/19 When an Exim process that is running as root has to create an Exim log
1210 file, it does so in a subprocess that runs as exim:exim so as to get the
1211 ownership right at creation (otherwise, other Exim processes might see
1212 the file with the wrong ownership). There was no test for failure of this
1213 fork() call, which would lead to the process getting stuck as it waited
1214 for a non-existent subprocess. Forks do occasionally fail when resources
1215 run out. I reviewed all the other calls to fork(); they all seem to check
1218 PH/20 When checking for unexpected SMTP input at connect time (before writing
1219 the banner), Exim was not dealing correctly with a non-positive return
1220 from the read() function. If the client had disconnected by this time,
1221 the result was a log entry for a synchronization error with an empty
1222 string after "input=" when read() returned zero. If read() returned -1
1223 (an event I could not check), uninitialized data bytes were printed.
1224 There were reports of junk text (parts of files, etc) appearing after
1227 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1229 PH/22 Added support for macro redefinition, and (re)definition in between
1230 driver and ACL definitions.
1232 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1233 forgetting to use the resulting value; it was using the unexpanded value.
1235 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1236 hadn't been configured. The fix is from Juergen Kreileder, who
1237 understands it better than I do:
1239 "Here's what I see happening with three configured cyrus_sasl
1240 authenticators configured (plain, login, cram-md5):
1242 On startup auth_cyrus_sasl_init() gets called for each of these.
1243 This means three calls to sasl_listmech() without a specified mech_list.
1244 => SASL tests which mechs of all available mechs actually work
1245 => three warnings about OTP not working
1246 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1248 With the patch, sasl_listmech() also gets called three times. But now
1249 SASL's mech_list option is set to the server_mech specified in the the
1250 authenticator. Or in other words, the answer from sasl_listmech()
1251 gets limited to just the mech you're testing for (which is different
1253 => the return list contains just 'plain' or 'login', 'cram-md5' or
1254 nothing depending on the value of ob->server_mech.
1256 I've just tested the patch: Authentication still works fine,
1257 unavailable mechs specified in the exim configuration are still
1258 caught, and the auth.log warnings about OTP are gone."
1260 PH/25 When debugging is enabled, the contents of the command line are added
1261 to the debugging output, even when log_selector=+arguments is not
1264 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1265 answer is "GNU", and only if the return is "GNU/something" is the answer
1268 PH/27 $acl_verify_message is now set immediately after the failure of a
1269 verification in an ACL, and so is available in subsequent modifiers. In
1270 particular, the message can be preserved by coding like this:
1272 warn !verify = sender
1273 set acl_m0 = $acl_verify_message
1275 Previously, $acl_verify_message was set only while expanding "message"
1276 and "log_message" when a very denied access.
1278 PH/28 Modified OS/os.c-Linux with
1280 -#ifndef OS_LOAD_AVERAGE
1281 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1283 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1284 nomenclature these days.)
1286 PH/29 Installed patch from the Sieve maintainer that adds the options
1287 sieve_useraddress and sieve_subaddress to the redirect router.
1289 PH/30 In these circumstances:
1290 . Two addresses routed to the same list of hosts;
1291 . First host does not offer TLS;
1292 . First host accepts first address;
1293 . First host gives temporary error to second address;
1294 . Second host offers TLS and a TLS session is established;
1295 . Second host accepts second address.
1296 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1297 and peerdn, if requested) that were in fact used only for the second
1300 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1301 attention to any local part prefix or suffix that was matched by the
1302 router that accepted the address. It now behaves in the same way as it
1303 does for delivery: the affixes are removed from the local part unless
1304 rcpt_include_affixes is set on the transport.
1306 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1307 timeout during the DATA phase of an incoming message.
1309 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1310 applied a patch sanctioned by the Sieve maintainer.
1312 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1313 the uid or gid is negative. A case of a negative gid caused this to be
1314 noticed. The fix allows for either to be negative.
1316 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1317 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1320 PH/36 The change PH/12 above was broken. Fixed it.
1322 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1323 the grounds that routing the same address twice would always produce the
1324 same answer. This might have been true once, but it is certainly no
1325 longer true now. Routing a child address may depend on the previous
1326 routing that produced that child. Some complicated redirection strategies
1327 went wrong when messages had multiple recipients, and made Exim's
1328 behaviour dependent on the order in which the addresses were given.
1330 I have moved the duplicate checking until after the routing is complete.
1331 Exim scans the addresses that are assigned to local and remote
1332 transports, and removes any duplicates. This means that more work will be
1333 done, as duplicates will always all be routed, but duplicates are
1334 presumably rare, so I don't expect this is of any significance.
1336 For deliveries to pipes, files, and autoreplies, the duplicate checking
1337 still happens during the routing process, since they are not going to be
1340 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1341 It corrects a timeout issue with spamd. This is Ian's comment: "The
1342 background is that sometimes spamd either never reads data from a
1343 connection it has accepted, or it never writes response data. The exiscan
1344 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1345 blindly assumes that writes won't block so it may never time out."
1347 PH/39 Allow G after quota size as well as K and M.
1349 PH/40 The value set for $authenticated_id in an authenticator may not contain
1350 binary zeroes or newlines because the value is written to log lines and
1351 to spool files. There was no check on this. Now the value is run through
1352 the string_printing() function so that such characters are converted to
1353 printable escape sequences.
1355 PH/41 $message_linecount is a new variable that contains the total number of
1356 lines in the message. Compare $body_linecount, which is the count for the
1359 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1360 bounce and delay warning messages, except in certain special cases, which
1363 (a) An SMTP error message from a remote host;
1364 (b) A message specified in a :fail: redirection;
1365 (c) A message specified in a "fail" command in a system filter;
1366 (d) A message specified in a FAIL return from the queryprogram router;
1367 (e) A message specified by the cannot_route_message router option.
1369 In these cases only, Exim does include the error details in bounce and
1370 warning messages. There are also a few cases where bland messages such
1371 as "unrouteable address" or "local delivery error" are given.
1373 PH/43 $value is now also set for the "else" part of a ${run expansion.
1375 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1376 being worked on, but at least Exim now implements the latest version to
1379 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1380 process to complete was treated as a delivery failure, a timeout while
1381 writing the message to the pipe was logged, but erroneously treated as a
1382 successful delivery. Such timeouts include transport filter timeouts. For
1383 consistency with the overall process timeout, these timeouts are now
1384 treated as errors, giving rise to delivery failures by default. However,
1385 there is now a new Boolean option for the pipe transport called
1386 timeout_defer, which, if set TRUE, converts the failures into defers for
1387 both kinds of timeout. A transport filter timeout is now identified in
1390 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1391 systems where "make" and "gmake" are different, calling "gmake" at top
1392 level broke things. I've arranged for the value of $(MAKE) to be passed
1393 from the Makefile to this script so that it can call the same version of
1397 A note about Exim versions 4.44 and 4.50
1398 ----------------------------------------
1400 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1401 changes of various kinds. As a consequence, a big documentation update was
1402 needed. This delayed the release for rather longer than seemed good, especially
1403 in the light of a couple of (minor) security issues. Therefore, the changes
1404 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1405 release. So 4.44 and 4.50 are in effect two different branches that both start
1408 I have left the 4.50 change log unchanged; it contains all the changes since
1409 4.43. The change log for 4.44 is below; many of its items are identical to
1410 those for 4.50. This seems to be the most sensible way to preserve the
1411 historical information.
1417 1. Minor wording change to the doc/README.SIEVE file.
1419 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1420 computation of the current number of files was incorrect.
1422 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1423 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1426 4. Give more explanation in the error message when the command for a transport
1427 filter fails to execute.
1429 5. There are several places where Exim runs a non-Exim command in a
1430 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1431 was being done only for the command run by the queryprogram router. It is
1432 now done for all such subprocesses. The other cases are: ${run, transport
1433 filters, and the commands run by the lmtp and pipe transports.
1435 6. Added CONFIGURE_GROUP build-time option.
1437 7. Some older OS have a limit of 256 on the maximum number of file
1438 descriptors. Exim was using setrlimit() to set 1000 as a large value
1439 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1440 systems. I've change it so that if it can't get 1000, it tries for 256.
1442 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1443 was an oversight, and furthermore, ever since the addition of extra
1444 controls (e.g. 4.43/32), the checks on when to allow different forms of
1445 "control" were broken. There should now be diagnostics for all cases when a
1446 control that does not make sense is encountered.
1448 9. Added the /retain_sender option to "control=submission".
1450 10. $recipients is now available in the predata ACL (oversight).
1452 11. Tidy the search cache before the fork to do a delivery from a message
1453 received from the command line. Otherwise the child will trigger a lookup
1454 failure and thereby defer the delivery if it tries to use (for example) a
1455 cached ldap connection that the parent has called unbind on.
1457 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1458 of $address_data from the recipient verification was clobbered by the
1459 sender verification.
1461 13. The value of address_data from a sender verification is now available in
1462 $sender_address_data in subsequent conditions in the ACL statement.
1464 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1466 15. Added a new option "connect=<time>" to callout options, to set a different
1469 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1470 was its contents. (It was OK if the option was not defined at all.)
1472 17. A "Completed" log line is now written for messages that are removed from
1473 the spool by the -Mrm option.
1475 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1476 information about exactly what failed.
1478 19. Added -dd to debug only the daemon process.
1480 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1481 handles timeouts, both on the server side and network timeouts. Renamed the
1482 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1484 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1485 It is now set to "smtps".
1487 22. $host_address is now set to the target address during the checking of
1488 ignore_target_hosts.
1490 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1491 being passed; this would have caused $sender_host_name to have been used if
1492 matching the list had actually called for a host name (not very likely,
1493 since this list is usually IP addresses). A host name is now passed as
1496 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1497 code that specifies a non-restarting handler (typically sigaction() in
1498 modern systems) in an attempt to fix a rare and obscure crash bug.
1500 25. Narrowed the window for a race in the daemon that could cause it to ignore
1501 SIGCHLD signals. This is not a major problem, because they are used only to
1502 wake it up if nothing else does.
1504 26. A malformed maildirsize file could cause Exim to calculate negative values
1505 for the mailbox size or file count. Odd effects could occur as a result.
1506 The maildirsize information is now recalculated if the size or filecount
1509 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1510 support for a long time. Removed HAVE_SYS_VFS_H.
1512 28. Installed the latest version of exipick from John Jetmore.
1514 29. In an address list, if the pattern was not a regular expression, an empty
1515 subject address (from a bounce message) matched only if the pattern was an
1516 empty string. Non-empty patterns were not even tested. This was the wrong
1517 because it is perfectly reasonable to use an empty address as part of a
1518 database query. An empty address is now tested by patterns that are
1519 lookups. However, all the other forms of pattern expect the subject to
1520 contain a local part and a domain, and therefore, for them, an empty
1521 address still always fails if the pattern is not itself empty.
1523 30. Exim went into a mad DNS loop when attempting to do a callout where the
1524 host was specified on an smtp transport, and looking it up yielded more
1525 than one IP address.
1527 31. Re-factored the code for checking spool and log partition space into a
1528 function that finds that data and another that does the check. The former
1529 is then used to implement four new variables: $spool_space, $log_space,
1530 $spool_inodes, and $log_inodes.
1532 32. The RFC2047 encoding function was originally intended for short strings
1533 such as real names; it was not keeping to the 75-character limit for
1534 encoded words that the RFC imposes. It now respects the limit, and
1535 generates multiple encoded words if necessary. To be on the safe side, I
1536 have increased the buffer size for the ${rfc2047: expansion operator from
1539 33. It is now permitted to omit both strings after an "if" condition; if the
1540 condition is true, the result is "true". As before, when the second string
1541 is omitted, a false condition yields an empty string. This makes it less
1542 cumbersome to write custom ACL and router conditions.
1544 34. Failure to deliver a bounce message always caused it to be frozen, even if
1545 there was an errors_to setting on the router. The errors_to setting is now
1548 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1549 canonical form (fully expanded) before being placed in
1550 $sender_host_address.
1552 36. The table in the code that translates DNS record types into text (T_A to
1553 "A" for instance) was missing entries for NS and CNAME. It is just possible
1554 that this could have caused confusion if both these types were looked up
1555 for the same domain, because the text type is used as part of Exim's
1556 per-process caching. But the chance of anyone hitting this buglet seems
1559 37. The dnsdb lookup has been extended in a number of ways.
1561 (1) There is a new type, "zns", which walks up the domain tree until it
1562 finds some nameserver records. It should be used with care.
1564 (2) There is a new type, "mxh", which is like "mx" except that it returns
1565 just the host names, not the priorities.
1567 (3) It is now possible to give a list of domains (or IP addresses) to be
1568 looked up. The behaviour when one of the lookups defers can be
1569 controlled by a keyword.
1571 (4) It is now possible to specify the separator character for use when
1572 multiple records are returned.
1574 38. The dnslists ACL condition has been extended: it is now possible to supply
1575 a list of IP addresses and/or domains to be looked up in a particular DNS
1578 39. Added log_selector=+queue_time_overall.
1580 40. When running the queue in the test harness, wait just a tad after forking a
1581 delivery process, to get repeatability of debugging output.
1583 41. Include certificate and key file names in error message when GnuTLS fails
1584 to set them up, because the GnuTLS error message doesn't include the name
1585 of the failing file when there is a problem reading it.
1587 42. Allow both -bf and -bF in the same test run.
1589 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1591 44. The "Exiscan patch" is now merged into the mainline Exim source.
1593 45. Sometimes the final signoff response after QUIT could fail to get
1594 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1595 < 0 before doing a fflush(). This bug looks as though it goes back to the
1596 introduction of TLS in release 3.20, but "sometimes" must have been rare
1597 because the tests only now provoked it.
1599 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1600 (this can affect the format of dates).
1602 47. exim_tidydb, when checking for the continued existence of a message for
1603 which it has found a message-specific retry record, was not finding
1604 messages that were in split spool directories. Consequently, it was
1605 deleting retry records that should have stayed in existence.
1607 48. Steve fixed some bugs in eximstats.
1609 49. The SPA authentication driver was not abandoning authentication and moving
1610 on to the next authenticator when an expansion was forced to fail,
1611 contradicting the general specification for all authenticators. Instead it
1612 was generating a temporary error. It now behaves as specified.
1614 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1615 (the order specifies the preference for clients). The order is now AES256,
1616 AES128, 3DES, ARCFOUR128.
1618 51. Small patch to Sieve code - explicitly set From: when generating an
1621 52. Exim crashed if a remote delivery caused a very long error message to be
1622 recorded - for instance if somebody sent an entire SpamAssassin report back
1623 as a large number of 550 error lines. This bug was coincidentally fixed by
1624 increasing the size of one of Exim's internal buffers (big_buffer) that
1625 happened as part of the Exiscan merge. However, to be on the safe side, I
1626 have made the code more robust (and fixed the comments that describe what
1629 53. Now that there can be additional text after "Completed" in log lines (if
1630 the queue_time_overall log selector is set), a one-byte patch to exigrep
1631 was needed to allow it to recognize "Completed" as not the last thing in
1634 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1635 patch that reportedly fixes this has been added. I am not expert enough to
1636 create a test for it. This is what the patch creator wrote:
1638 "I found a little strange behaviour of ldap code when working with
1639 Windows 2003 AD Domain, where users was placed in more than one
1640 Organization Units. When I tried to give exim partial DN, the exit code
1641 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1642 But simultaneously result of request was absolutely normal ldap result,
1643 so I produce this patch..."
1645 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1646 REFERENCE, so I have modified the code to exclude the patch when that macro
1649 55. Some experimental protocols are using DNS PTR records for new purposes. The
1650 keys for these records are domain names, not reversed IP addresses. The
1651 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1652 leaves it alone. Component reversal etc. now happens only for IP addresses.
1655 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1657 57. Double the size of the debug message buffer (to 2048) so that more of very
1658 long debug lines gets shown.
1660 58. The exicyclog utility now does better if the number of log files to keep
1661 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1663 59. Two changes related to the smtp_active_hostname option:
1665 (1) $smtp_active_hostname is now available as a variable.
1666 (2) The default for smtp_banner uses $smtp_active_hostname instead
1667 of $primary_hostname.
1669 60. The host_aton() function is supposed to be passed a string that is known
1670 to be a valid IP address. However, in the case of IPv6 addresses, it was
1671 not checking this. This is a hostage to fortune. Exim now panics and dies
1672 if the condition is not met. A case was found where this could be provoked
1673 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1674 components; fortuitously, this particular loophole had already been fixed
1675 by change 4.50/55 above.
1677 If there are any other similar loopholes, the new check in host_aton()
1678 itself should stop them being exploited. The report I received stated that
1679 data on the command line could provoke the exploit when Exim was running as
1680 exim, but did not say which command line option was involved. All I could
1681 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1682 running as the user.
1685 61. There was a buffer overflow vulnerability in the SPA authentication code
1686 (which came originally from the Samba project). I have added a test to the
1687 spa_base64_to_bits() function which I hope fixes it.
1690 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1691 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1693 63. The daemon start-up calls getloadavg() while still root for those OS that
1694 need the first call to be done as root, but it missed one case: when
1695 deliver_queue_load_max is set with deliver_drop_privilege. This is
1696 necessary for the benefit of the queue runner, because there is no re-exec
1697 when deliver_drop_privilege is set.
1699 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1700 This has been fixed.
1702 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1703 was in use, was not putting the data itself into the right store pool;
1704 consequently, it could be overwritten for a subsequent message in the same
1705 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1708 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1710 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1711 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1712 the function were treating the return as a boolean value, which happened to
1713 work because 0=false and not-0=true, but is not correct code.
1715 68. The host_aton() function was not handling scoped IPv6 addresses (those
1716 with, for example, "%eth0" on the end) correctly.
1718 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1719 negated items (that is, ~something) in unsigned ints. Some compilers
1720 apparently mutter when there is no cast.
1722 70. If an address verification called from an ACL failed, and did not produce a
1723 user-specific message (i.e. there was only a "system" message), nothing was
1724 put in $acl_verify_message. In this situation, it now puts the system
1727 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1728 session; change 4.31/43 added the unwanted input to the log line - except
1729 that it did not do this in the start of session case. It now does.
1731 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1732 This is wrong and can cause the other end to generate a synchronization
1733 error if it is another Exim or anything else that does the synchronization
1734 check. A QUIT command is no longer sent after a timeout.
1736 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1737 during host lookups.
1739 74. The defer_ok option of callout verification was not working if it was used
1740 when verifying addresses in header lines, that is, for this case:
1742 verify = header_sender/callout=defer_ok
1744 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1745 those file descriptors could be used for SMTP connections. If anything
1746 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1747 could be sent to the SMTP client, causing chaos. The daemon now opens
1748 stdin, stdout, and stderr to /dev/null when it puts itself into the
1751 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1752 log by default. The user can override this with suitable Perl magic.
1754 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1755 the log message when discard triggers, was not working for the DATA ACL or
1756 for the non-SMTP ACL.
1758 78. Error message wording change in sieve.c.
1760 79. If smtp_accept_max_per_host was set, the number of connections could be
1761 restricted to fewer than expected, because the daemon was trying to set up
1762 a new connection before checking whether the processes handling previous
1763 connections had finished. The check for completed processes is now done
1764 earlier. On busy systems, this bug wouldn't be noticed because something
1765 else would have woken the daemon, and it would have reaped the completed
1768 80. If a message was submitted locally by a user whose login name contained one
1769 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1770 It caused a spool format error. I have fixed the spool reading code. A
1771 related problem was that the "from" clause in the Received: line became
1772 illegal because of the space(s). It is now covered by ${quote_local_part.
1774 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1777 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1779 83. Patch from Sieve maintainer to fix unterminated string problem in
1780 "vacation" handling.
1782 84. Some minor changes to the Linux configuration files to help with other
1783 OS variants using glibc.
1785 85. One more patch for Sieve to update vacation handling to latest spec.
1788 ----------------------------------------------------
1789 See the note above about the 4.44 and 4.50 releases.
1790 ----------------------------------------------------
1796 1. Change 4.43/35 introduced a bug that caused file counts to be
1797 incorrectly computed when quota_filecount was set in an appendfile
1800 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1801 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1804 3. Give more explanation in the error message when the command for a transport
1805 filter fails to execute.
1807 4. There are several places where Exim runs a non-Exim command in a
1808 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1809 was being done only for the command run by the queryprogram router. It is
1810 now done for all such subprocesses. The other cases are: ${run, transport
1811 filters, and the commands run by the lmtp and pipe transports.
1813 5. Some older OS have a limit of 256 on the maximum number of file
1814 descriptors. Exim was using setrlimit() to set 1000 as a large value
1815 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1816 systems. I've change it so that if it can't get 1000, it tries for 256.
1818 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1819 was an oversight, and furthermore, ever since the addition of extra
1820 controls (e.g. 4.43/32), the checks on when to allow different forms of
1821 "control" were broken. There should now be diagnostics for all cases when a
1822 control that does not make sense is encountered.
1824 7. $recipients is now available in the predata ACL (oversight).
1826 8. Tidy the search cache before the fork to do a delivery from a message
1827 received from the command line. Otherwise the child will trigger a lookup
1828 failure and thereby defer the delivery if it tries to use (for example) a
1829 cached ldap connection that the parent has called unbind on.
1831 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1832 of $address_data from the recipient verification was clobbered by the
1833 sender verification.
1835 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1836 was its contents. (It was OK if the option was not defined at all.)
1838 11. A "Completed" log line is now written for messages that are removed from
1839 the spool by the -Mrm option.
1841 12. $host_address is now set to the target address during the checking of
1842 ignore_target_hosts.
1844 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1845 being passed; this would have caused $sender_host_name to have been used if
1846 matching the list had actually called for a host name (not very likely,
1847 since this list is usually IP addresses). A host name is now passed as
1850 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1851 code that specifies a non-restarting handler (typically sigaction() in
1852 modern systems) in an attempt to fix a rare and obscure crash bug.
1854 15. Narrowed the window for a race in the daemon that could cause it to ignore
1855 SIGCHLD signals. This is not a major problem, because they are used only to
1856 wake it up if nothing else does.
1858 16. A malformed maildirsize file could cause Exim to calculate negative values
1859 for the mailbox size or file count. Odd effects could occur as a result.
1860 The maildirsize information is now recalculated if the size or filecount
1863 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1864 support for a long time. Removed HAVE_SYS_VFS_H.
1866 18. Updated exipick to current release from John Jetmore.
1868 19. Allow an empty sender to be matched against a lookup in an address list.
1869 Previously the only cases considered were a regular expression, or an
1872 20. Exim went into a mad DNS lookup loop when doing a callout where the
1873 host was specified on the transport, if the DNS lookup yielded more than
1876 21. The RFC2047 encoding function was originally intended for short strings
1877 such as real names; it was not keeping to the 75-character limit for
1878 encoded words that the RFC imposes. It now respects the limit, and
1879 generates multiple encoded words if necessary. To be on the safe side, I
1880 have increased the buffer size for the ${rfc2047: expansion operator from
1883 22. Failure to deliver a bounce message always caused it to be frozen, even if
1884 there was an errors_to setting on the router. The errors_to setting is now
1887 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1888 canonical form (fully expanded) before being placed in
1889 $sender_host_address.
1891 24. Updated eximstats to version 1.33
1893 25. Include certificate and key file names in error message when GnuTLS fails
1894 to set them up, because the GnuTLS error message doesn't include the name
1895 of the failing file when there is a problem reading it.
1897 26. Expand error message when OpenSSL has problems setting up cert/key files.
1900 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1901 (this can affect the format of dates).
1903 28. exim_tidydb, when checking for the continued existence of a message for
1904 which it has found a message-specific retry record, was not finding
1905 messages that were in split spool directories. Consequently, it was
1906 deleting retry records that should have stayed in existence.
1908 29. eximstats updated to version 1.35
1909 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1910 1.35 - bugfix such that pie charts by volume are generated correctly
1912 30. The SPA authentication driver was not abandoning authentication and moving
1913 on to the next authenticator when an expansion was forced to fail,
1914 contradicting the general specification for all authenticators. Instead it
1915 was generating a temporary error. It now behaves as specified.
1917 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1918 (the order specifies the preference for clients). The order is now AES256,
1919 AES128, 3DES, ARCFOUR128.
1921 31. Small patch to Sieve code - explicitly set From: when generating an
1924 32. Exim crashed if a remote delivery caused a very long error message to be
1925 recorded - for instance if somebody sent an entire SpamAssassin report back
1926 as a large number of 550 error lines. This bug was coincidentally fixed by
1927 increasing the size of one of Exim's internal buffers (big_buffer) that
1928 happened as part of the Exiscan merge. However, to be on the safe side, I
1929 have made the code more robust (and fixed the comments that describe what
1932 33. Some experimental protocols are using DNS PTR records for new purposes. The
1933 keys for these records are domain names, not reversed IP addresses. The
1934 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1935 leaves it alone. Component reversal etc. now happens only for IP addresses.
1938 34. The host_aton() function is supposed to be passed a string that is known
1939 to be a valid IP address. However, in the case of IPv6 addresses, it was
1940 not checking this. This is a hostage to fortune. Exim now panics and dies
1941 if the condition is not met. A case was found where this could be provoked
1942 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1943 components; fortuitously, this particular loophole had already been fixed
1944 by change 4.50/55 or 4.44/33 above.
1946 If there are any other similar loopholes, the new check in host_aton()
1947 itself should stop them being exploited. The report I received stated that
1948 data on the command line could provoke the exploit when Exim was running as
1949 exim, but did not say which command line option was involved. All I could
1950 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1951 running as the user.
1954 35. There was a buffer overflow vulnerability in the SPA authentication code
1955 (which came originally from the Samba project). I have added a test to the
1956 spa_base64_to_bits() function which I hope fixes it.
1959 36. The daemon start-up calls getloadavg() while still root for those OS that
1960 need the first call to be done as root, but it missed one case: when
1961 deliver_queue_load_max is set with deliver_drop_privilege. This is
1962 necessary for the benefit of the queue runner, because there is no re-exec
1963 when deliver_drop_privilege is set.
1965 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1966 was in use, was not putting the data itself into the right store pool;
1967 consequently, it could be overwritten for a subsequent message in the same
1968 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1971 38. Sometimes the final signoff response after QUIT could fail to get
1972 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1973 < 0 before doing a fflush(). This bug looks as though it goes back to the
1974 introduction of TLS in release 3.20, but "sometimes" must have been rare
1975 because the tests only now provoked it.
1981 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1982 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1983 Now it returns an integer. A number of calls were still expecting a T/F
1984 return. Fortuitously, in all cases, the tests worked in OK situations,
1985 which is the norm. However, things would have gone wrong on any write
1986 failures on the smtp file descriptor. This function is used when sending
1987 messages over SMTP and also when doing verify callouts.
1989 2. When Exim is called to do synchronous delivery of a locally submitted
1990 message (the -odf or -odi options), it no longer closes stderr before doing
1993 3. Implemented the mua_wrapper option.
1995 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1997 5. Implemented the functions header_remove(), header_testname(),
1998 header_add_at_position(), and receive_remove_recipient(), and exported them
2001 6. If an ACL "warn" statement specified the addition of headers, Exim already
2002 inserted X-ACL-Warn: at the start if there was no header name. However, it
2003 was not making this test for the second and subsequent header lines if
2004 there were newlines in the string. This meant that an invalid header could
2005 be inserted if Exim was badly configured.
2007 7. Allow an ACL "warn" statement to add header lines at the start or after all
2008 the Received: headers, as well as at the end.
2010 8. Added the rcpt_4xx retry error code.
2012 9. Added postmaster_mailfrom=xxx to callout verification option.
2014 10. Added mailfrom=xxxx to the callout verification option, for verify=
2017 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2018 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2020 12. Inserted some casts to stop certain compilers warning when using pointer
2021 differences as field lengths or precisions in printf-type calls (mostly
2022 affecting debugging statements).
2024 13. Added optional readline() support for -be (dynamically loaded).
2026 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2027 same clock tick as a message's arrival, so that its received time was the
2028 same as the "first fail" time on the retry record, and that message
2029 remained on the queue past the ultimate address timeout, every queue runner
2030 would try a delivery (because it was past the ultimate address timeout) but
2031 after another failure, the ultimate address timeout, which should have then
2032 bounced the address, did not kick in. This was a "< instead of <=" error;
2033 in most cases the first failure would have been in the next clock tick
2034 after the received time, and all would be well.
2036 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2037 being recognized when the domain list was tested by the match_domain
2038 condition in an expansion string.
2040 16. Added the ${str2b64: operator.
2042 17. Exim was always calling setrlimit() to set a large limit for the number of
2043 processes, without checking whether the existing limit was already
2044 adequate. (It did check for the limit on file descriptors.) Furthermore,
2045 errors from getrlimit() and setrlimit() were being ignored. Now they are
2046 logged to the main and panic logs, but Exim does carry on, to try to do its
2047 job under whatever limits there are.
2049 18. Imported PCRE 5.0.
2051 19. Trivial typo in log message " temporarily refused connection" (the leading
2054 20. If the log selector return_path_on_delivery was set and an address was
2055 redirected to /dev/null, the delivery process crashed because it assumed
2056 that a return path would always be set for a "successful" delivery. In this
2057 case, the whole delivery is bypassed as an optimization, and therefore no
2060 21. Internal re-arrangement: the function for sending a challenge and reading
2061 a response while authentication was assuming a zero-terminated challenge
2062 string. It's now changed to take a pointer and a length, to allow for
2063 binary data in such strings.
2065 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2067 23. Exim was not respecting finduser_retries when seeking the login of the
2068 uid under which it was called; it was always trying 10 times. (The default
2069 setting of finduser_retries is zero.) Also, it was sleeping after the final
2070 failure, which is pointless.
2072 24. Implemented tls_on_connect_ports.
2074 25. Implemented acl_smtp_predata.
2076 26. If the domain in control=submission is set empty, Exim assumes that the
2077 authenticated id is a complete email address when it generates From: or
2078 Sender: header lines.
2080 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2081 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2082 chown and chgrp in /bin and hostname in /usr/bin.
2084 28. Exim was keeping the "process log" file open after each use, just as it
2085 does for the main log. This opens the possibility of it remaining open for
2086 long periods when the USR1 signal hits a daemon. Occasional processlog
2087 errors were reported, that could have been caused by this. Anyway, it seems
2088 much more sensible not to leave this file open at all, so that is what now
2091 29. The long-running daemon process does not normally write to the log once it
2092 has entered its main loop, and it closes the log before doing so. This is
2093 so that log files can straightforwardly be renamed and moved. However,
2094 there are a couple of unusual error situations where the daemon does write
2095 log entries, and I had neglected to close the log afterwards.
2097 30. The text of an SMTP error response that was received during a remote
2098 delivery was being truncated at 512 bytes. This is too short for some of
2099 the long messages that one sometimes sees. I've increased the limit to
2102 31. It is now possible to make retry rules that apply only when a message has a
2103 specific sender, in particular, an empty sender.
2105 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2106 it possible to be selective about when SMTP synchronization is enforced.
2108 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2110 32. Implemented hosts_connection_nolog.
2112 33. Added an ACL for QUIT.
2114 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2117 35. Added mailbox_size and mailbox_filecount to appendfile.
2119 36. Added control = no_multiline_responses to ACLs.
2121 37. There was a bug in the logic of the code that waits for the clock to tick
2122 in the case where the clock went backwards by a substantial amount such
2123 that the microsecond fraction of "now" was more than the microsecond
2124 fraction of "then" (but the whole seconds number was less).
2126 38. Added support for the libradius Radius client library this is found on
2127 FreeBSD (previously only the radiusclient library was supported).
2133 1. When certain lookups returned multiple values in the form name=value, the
2134 quoting of the values was not always being done properly. Specifically:
2135 (a) If the value started with a double quote, but contained no whitespace,
2137 (b) If the value contained whitespace other than a space character (i.e.
2138 tabs or newlines or carriage returns) it was not quoted.
2139 This fix has been applied to the mysql and pgsql lookups by writing a
2140 separate quoting function and calling it from the lookup code. The fix
2141 should probably also be applied to nisplus, ibase and oracle lookups, but
2142 since I cannot test any of those, I have not disturbed their existing code.
2144 2. A hit in the callout cache for a specific address caused a log line with no
2145 reason for rejecting RCPT. Now it says "Previous (cached) callout
2146 verification failure".
2148 3. There was an off-by-one bug in the queryprogram router. An over-long
2149 return line was truncated at 256 instead of 255 characters, thereby
2150 overflowing its buffer with the terminating zero. As well as fixing this, I
2151 have increased the buffer size to 1024 (and made a note to document this).
2153 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2154 when Exim is waiting for an SMTP response from a remote server, Exim
2155 restarts its select() call on the socket, thereby resetting its timeout.
2156 This is not a problem when such interrupts are rare. Somebody set up a cron
2157 job to run exiwhat every 2 minutes, which is less than the normal select()
2158 timeout (5 or 10 minutes). This meant that the select() timeout never
2159 kicked in because it was always reset. I have fixed this by comparing the
2160 time when an interrupt arrives with the time at the start of the first call
2161 to select(). If more time than the timeout has elapsed, the interrupt is
2162 treated as a timeout.
2164 5. Some internal re-factoring in preparation for the addition of Sieve
2165 extensions (by MH). In particular, the "personal" test is moved to a
2166 separate function, and given an option for scanning Cc: and Bcc: (which is
2167 not set for Exim filters).
2169 6. When Exim created an email address using the login of the caller as the
2170 local part (e.g. when creating a From: or Sender: header line), it was not
2171 quoting the local part when it contained special characters such as @.
2173 7. Installed new OpenBSD configuration files.
2175 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2176 try to make them clearer.
2178 9. Callout options, other than the timeout value, were being ignored when
2179 verifying sender addresses in header lines. For example, when using
2181 verify = header_sender/callout=no_cache
2183 the cache was (incorrectly) being used.
2185 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2186 only the Cygwin environment.
2188 11. When return_path_on_delivery was set as a log selector, if different remote
2189 addresses in the same message used different return paths and parallel
2190 remote delivery occurred, the wrong values would sometimes be logged.
2191 (Whenever a remote delivery process finished, the return path value from
2192 the most recently started remote delivery process was logged.)
2194 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2195 lines when AUTH and/or TLS are in use. This is the "received protocol"
2196 field. Exim used to use "asmtp" for authenticated SMTP, without any
2197 indication (in the protocol name) for TLS use. Now it follows the RFC and
2198 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2199 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2200 names appear in log lines as well as in Received: header lines.
2202 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2203 extensions, and comparison tests, and to fix some bugs.
2205 14. Changes to the "personal" filter test:
2207 (1) The test was buggy in that it was just doing the equivalent of
2208 "contains" tests on header lines. For example, if a user's address was
2209 anne@some.where, the "personal" test would incorrectly be true for
2211 To: susanne@some.where
2213 This test is now done by extracting each address from the header in turn,
2214 and checking the entire address. Other tests that are part of "personal"
2215 are now done using regular expressions (for example, to check local parts
2216 of addresses in From: header lines).
2218 (2) The list of non-personal local parts in From: addresses has been
2219 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2220 taken from the Sieve specification recommendations.
2222 (3) If the message contains any header line starting with "List-" it is
2223 treated as non-personal.
2225 (4) The test for "circular" in the Subject: header line has been removed
2226 because it now seems ill-conceived.
2228 15. Minor typos in src/EDITME comments corrected.
2230 16. Installed latest exipick from John Jetmore.
2232 17. If headers_add on a router specified a text string that was too long for
2233 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2234 of string_sprintf() is now avoided.
2236 18. $message_body_size was not set (it was always zero) when running the DATA
2237 ACL and the local_scan() function.
2239 19. For the "mail" command in an Exim filter, no default was being set for
2240 the once_repeat time, causing a random time value to be used if "once" was
2241 specified. (If the value happened to be <= 0, no repeat happened.) The
2242 default is now 0s, meaning "never repeat". The "vacation" command was OK
2243 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2244 (I found it when inspecting the code).
2246 20. There is now an overall timeout for performing a callout verification. It
2247 defaults to 4 times the callout timeout, which applies to individual SMTP
2248 commands during the callout. The overall timeout applies when there is more
2249 than one host that can be tried. The timeout is checked before trying the
2250 next host. This prevents very long delays if there are a large number of
2251 hosts and all are timing out (e.g. when the network connections are timing
2252 out). The value of the overall timeout can be changed by specifying an
2253 additional sub-option for "callout", called "maxwait". For example:
2255 verify = sender/callout=5s,maxwait=20s
2257 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2258 to the end before writing, but this should make it even safer).
2260 22. Exim was forgetting that it had advertised PIPELINING for the second and
2261 subsequent messages on an SMTP connection. It was also not resetting its
2262 memory on STARTTLS and an internal HELO.
2264 23. When Exim logs an SMTP synchronization error within a session, it now
2265 records whether PIPELINING has been advertised or not.
2267 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2268 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2269 rather than long int.
2271 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2273 26. Added the never_mail option to autoreply.
2279 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2280 crash if the getsockname() call failed; this can happen if a connection is
2281 closed very soon after it is established. The problem was simply in the
2282 order in which certain operations were done, causing Exim to try to write
2283 to the SMTP stream before it had set up the file descriptor. The bug has
2284 been fixed by making things happen in the correct order.
2290 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2291 before the connection was closed, thus losing the rejection response.
2293 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2294 some early Solaris releases, but causes trouble in current releases where
2295 socklen_t is defined.
2297 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2300 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2303 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2304 that releasing the top of it at the end releases what was used for sub-
2305 expansions (unless the block got too big). However, discard this block if
2306 the first thing is a variable or header, so that we can use its block when
2307 it is dynamic (useful for very large $message_headers, for example).
2309 6. Lookups now cache *every* query, not just the most recent. A new, separate
2310 store pool is used for this. It can be recovered when all lookup caches are
2311 flushed. Lookups now release memory at the end of their result strings.
2312 This has involved some general refactoring of the lookup sources.
2314 7. Some code has been added to the store_xxx() functions to reduce the amount
2315 of flapping under certain conditions.
2317 8. log_incoming_interface used to affect only the <= reception log lines. Now
2318 it causes the local interface and port to be added to several more SMTP log
2319 lines, for example "SMTP connection from", and rejection lines.
2321 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2323 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2325 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2326 could be overwritten at the end of the current message (or the start of a
2327 new message if it was set in a HELO ACL). The value is now preserved for
2328 the duration of the SMTP connection.
2330 12. If a transport had a headers_rewrite setting, and a matching header line
2331 contained an unqualified address, that address was qualified, even if it
2332 did not match any rewriting rules. The underlying bug was that the values
2333 of the flags that permit the existence of unqualified sender and recipient
2334 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2335 non-local messages, and by -bnq for local messages) were not being
2336 preserved with the message after it was received.
2338 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2339 "next input=" as part of the text comprising the host identity instead of
2340 the correct text. The code was using the same buffer for two different
2341 strings. However, depending on which order the printing function evaluated
2342 its arguments, the bug did not always show up. Under Linux, for example, my
2343 test suite worked just fine.
2345 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2346 doesn't work with some older versions of Perl. It has been changed to "my",
2347 which in any case is probably the better facility to use.
2349 15. A really picky compiler found some instances of statements for creating
2350 error messages that either had too many or two few arguments for the format
2353 16. The size of the buffer for calls to the DNS resolver has been increased
2354 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2355 for addresses that have a lot of PTR records. This alleviates a problem; it
2356 does not fully solve it.
2358 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2359 buffer now truncates the list and logs the incident, which is the same
2360 action as happens when Exim is looking up a host name and its aliases.
2361 Previously in this situation something unpredictable would happen;
2362 sometimes it was "internal error: store_reset failed".
2364 18. If a server dropped the connection unexpectedly when an Exim client was
2365 using GnuTLS and trying to read a response, the client delivery process
2366 crashed while trying to generate an error log message.
2368 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2369 string, for example:
2371 warn message = H1: something\nH2: something
2373 the text was added as a single header line from Exim's point of view
2374 though it ended up OK in the delivered message. However, searching for the
2375 second and subsequent header lines using $h_h2: did not work. This has been
2376 fixed. Similarly, if a system filter added multiple headers in this way,
2377 the routers could not see them.
2379 20. Expanded the error message when iplsearch is called with an invalid key to
2380 suggest using net-iplsearch in a host list.
2382 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2383 ACLs are no longer actually imposed (and a message to that effect is
2386 22. If a "gecos" field in a passwd entry contained escaped characters, in
2387 particular, if it contained a \" sequence, Exim got it wrong when building
2388 a From: or a Sender: header from that name. A second bug also caused
2389 incorrect handling when an unquoted " was present following a character
2390 that needed quoting.
2392 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2393 was not being matched caselessly.
2395 24. Arranged for all hyphens in the exim.8 source to be escaped with
2398 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2399 properly. Recipient callout cache records were still being keyed to include
2400 the sender, even when use_sender was set false. This led to far more
2401 callouts that were necessary. The sender is no longer included in the key
2402 when use_sender is false.
2404 26. Added "control = submission" modifier to ACLs.
2406 27. Added the ${base62d: operator to decode base 62 numbers.
2408 28. dnsdb lookups can now access SRV records.
2410 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2411 the configuration file.
2413 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2414 (-v) mode. This makes the output for a verbose queue run more intelligible.
2416 31. Added a use_postmaster feature to recipient callouts.
2418 32. Added the $body_zerocount variable, containing the number of binary zero
2419 bytes in the message body.
2421 33. The time of last modification of the "new" subdirectory is now used as the
2422 "mailbox time last read" when there is a quota error for a maildir
2425 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2427 35. Added +ignore_unknown as a special item in host lists.
2429 36. Code for decoding IPv6 addresses in host lists is now included, even if
2430 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2431 address was recognized as an IP address, but was then not correctly decoded
2432 into binary, causing unexpected and incorrect effects when compared with
2439 1. Very minor rewording of debugging text in manualroute to say "list of
2440 hosts" instead of "hostlist".
2442 2. If verify=header_syntax was set, and a header line with an unqualified
2443 address (no domain) and a large number of spaces between the end of the
2444 name and the colon was received, the reception process suffered a buffer
2445 overflow, and (when I tested it) crashed. This was caused by some obsolete
2446 code that should have been removed. The fix is to remove it!
2448 3. When running in the test harness, delay a bit after writing a bounce
2449 message to get a bit more predictability in the log output.
2451 4. Added a call to search_tidyup() just before forking a reception process. In
2452 theory, someone could use a lookup in the expansion of smtp_accept_max_
2453 per_host which, without the tidyup, could leave open a database connection.
2455 5. Added the variables $recipient_data and $sender_data which get set from a
2456 lookup success in an ACL "recipients" or "senders" condition, or a router
2457 "senders" option, similar to $domain_data and $local_part_data.
2459 6. Moved the writing of debug_print from before to after the "senders" test
2462 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2463 problems for message scanning, either using a data ACL, or using
2464 local_scan() because the Received: header was not generated till after they
2465 were called (in order to set the time as the time of reception completion).
2466 I have revised the way this works. The header is now generated after the
2467 body is received, but before the ACL or local_scan() are called. After they
2468 are run, the timestamp in the header is updated.
2474 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2475 before starting a queue runner without re-exec. This happened only when
2476 deliver_drop_privilege was set or when the Exim user was set to root. The
2477 effect of the bug was that timeouts during subsequent deliveries caused
2478 crashes instead of being properly handled. The handler is now left at its
2479 default (and expected) setting.
2481 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2482 message, again when deliver_drop_privilege is set or Exim is run as root.
2483 The bug described in (1) was not present in this case, but the tidying up
2484 of the other signals was missing. I have made the two cases consistent.
2486 3. The ignore_target_hosts setting on a manualroute router was being ignored
2487 for hosts that were looked up using the /MX notation.
2489 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2492 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2493 operated on the sender address. After changing the $sender_address to <>
2494 for the sender address verify, Exim was re-instated it as the original
2495 (before rewriting) address, but remembering that it had rewritten it, so it
2496 wasn't rewriting it again. This bug also had the effect of breaking the
2497 sender address verification caching when the sender address was rewritten.
2499 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2500 This has been changed so that if the ip literal address matches
2501 ignore_target_hosts, the router declines.
2503 7. Added expansion conditions match_domain, match_address, and match_local_
2504 part (NOT match_host).
2506 8. The placeholder for the Received: header didn't have a length field set.
2508 9. Added code to Exim itself and to exim_lock to test for a specific race
2509 condition that could lead to file corruption when using MBX delivery. The
2510 issue is with the lockfile that is created in /tmp. If this file is removed
2511 after a process has opened it but before that process has acquired a lock,
2512 there is the potential for a second process to recreate the file and also
2513 acquire a lock. This could lead to two Exim processes writing to the file
2514 at the same time. The added code performs the same test as UW imapd; it
2515 checks after acquiring the lock that its file descriptor still refers to
2516 the same named file.
2518 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2519 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2522 11. Added the smtp_active_hostname option. If used, this will typically be made
2523 to depend on the incoming interface address. Because $interface_address is
2524 not set up until the daemon has forked a reception process, error responses
2525 that can happen earlier (such as "too many connections") no longer contain
2528 12. If an expansion in a condition on a "warn" statement fails because a lookup
2529 defers, the "warn" statement is abandoned, and the next ACL statement is
2530 processed. Previously this caused the whole ACL to be aborted.
2532 13. Added the iplsearch lookup type.
2534 14. Added ident_timeout as a log selector.
2536 15. Added tls_certificate_verified as a log selector.
2538 16. Added a global option tls_require_ciphers (compare the smtp transport
2539 option of the same name). This controls incoming TLS connections.
2541 17. I finally figured out how to make tls_require_ciphers do a similar thing
2542 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2543 before starting the TLS session.
2545 18. Tabs are now shown as \t in -bP output.
2547 19. If the log selector return_path_on_delivery was set, Exim crashed when
2548 bouncing a message because it had too many Received: header lines.
2550 20. If two routers both had headers_remove settings, and the first one included
2551 a superfluous trailing colon, the final name in the first list and the
2552 first name in the second list were incorrectly joined into one item (with a
2553 colon in the middle).
2559 1. Added -C and -D options to the exinext utility, mainly to make it easier
2560 to include in the automated testing, but these could be helpful when
2561 multiple configurations are in use.
2563 2. The exinext utility was not formatting the output nicely when there was
2564 an alternate port involved in the retry record key, nor when there was a
2565 message id as well (for retries that were specific to a specific message
2566 and a specific host). It was also confused by IPv6 addresses, because of
2567 the additional colons they contain. I have fixed the IPv4 problem, and
2568 patched it up to do a reasonable job for IPv6.
2570 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2571 delivery, the log line now contains "pipelined" if PIPELINING was used.
2573 4. An SMTP transport process used to panic and die if the bind() call to set
2574 an explicit outgoing interface failed. This has been changed; it is now
2575 treated in the same way as a connect() failure.
2577 5. A reference to $sender_host_name in the part of a conditional expansion
2578 that was being skipped was still causing a DNS lookup. This no longer
2581 6. The def: expansion condition was not recognizing references to header lines
2582 that used bh_ and bheader_.
2584 7. Added the _cache feature to named lists.
2586 8. The code for checking quota_filecount in the appendfile transport was
2587 allowing one more file than it should have been.
2589 9. For compatibility with Sendmail, the command line option
2597 and sets the incoming protocol and host name (for trusted callers). The
2598 host name and its colon can be omitted when only the protocol is to be set.
2599 Note the Exim already has two private options, -pd and -ps, that refer to
2600 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2601 "s", but I don't think that's a major issue.
2603 10. A number of refactoring changes to the code, none of which should affect
2606 (a) The number of logging options was getting close to filling up the
2607 32-bit word that was used as a bit map. I have split them into two classes:
2608 those that are passed in the argument to log_write(), and those that are
2609 only ever tested independently outside of that function. These are now in
2610 separate 32-bit words, so there is plenty of room for expansion again.
2611 There is no change in the user interface or the logging behaviour.
2613 (b) When building, for example, log lines, the code previously used a
2614 macro that called string_cat() twice, in order to add two strings. This is
2615 not really sufficiently general. Furthermore, there was one instance where
2616 it was actually wrong because one of the argument was used twice, and in
2617 one call a function was used. (As it happened, calling the function twice
2618 did not affect the overall behaviour.) The macro has been replaced by a
2619 function that can join an arbitrary number of extra strings onto a growing
2622 (c) The code for expansion conditions now uses a table and a binary chop
2623 instead of a serial search (which was left over from when there were very
2624 few conditions). Also, it now recognizes conditions like "pam" even when
2625 the relevant support is not compiled in: a suitably worded error message is
2626 given if an attempt is made to use such a condition.
2628 11. Added ${time_interval:xxxxx}.
2630 12. A bug was causing one of the ddress fields not to be passed back correctly
2631 from remote delivery subprocesses. The field in question was not being
2632 subsequently used, so this caused to problems in practice.
2634 13. Added new log selectors queue_time and deliver_time.
2636 14. Might have fixed a bug in maildirsizefile handling that threw up
2637 "unexpected character" debug warnings, and recalculated the data
2638 unnecessarily. In any case, I expanded the warning message to give more
2641 15. Added the message "Restricted characters in address" to the statements in
2642 the default ACL that block characters like @ and % in local parts.
2644 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2645 Three changes have been made:
2647 (a) There was a serious bug; a negative response to MAIL caused the whole
2648 recipient domain to be cached as invalid, thereby blocking all messages
2649 to all local parts at the same domain, from all senders. This bug has
2650 been fixed. The domain is no longer cached after a negative response to
2651 MAIL if the sender used is not empty.
2653 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2656 (c) A new callout option, "use_sender" has been added for people who want
2657 the modified behaviour.
2663 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2666 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2667 indeed breaks things for older releases.
2669 3. Added additional logging to the case where there is a problem reading data
2670 from a filter that is running in a subprocess using a pipe, in order to
2671 try to track down a specific problem.
2673 4. Testing facility fudge: when running in the test harness and attempting
2674 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2675 getting "No route to host". Convert this to a timeout.
2677 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2680 6. Some OS don't have socklen_t but use size_t instead. This affects the
2681 fifth argument of getsockopt() amongst other things. This is now
2682 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2683 can be set for individual OS. I have set it for SunOS5, OSF1, and
2684 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2685 some earlier ones do not.
2687 7. Change 4.30/15 was not doing the test caselessly.
2689 8. The standard form for an IPv6 address literal was being rejected by address
2690 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2691 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2692 this, as well as the form without the "IPv6" on the front (but only when
2693 address literals are enabled, of course).
2695 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2697 10. Exim crashed if a message with an empty sender address specified by -f
2698 encountered a router with an errors_to setting. This could be provoked only
2699 by a command such as
2703 where an empty string was supplied; "<>" did not hit this bug.
2705 11. Installed PCRE release 4.5.
2707 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2708 remained set. It is now erased.
2710 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2711 times from message ids (which are base 36 rather than the normal 62).
2713 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2714 were being counted as actual protocol errors, and logged if the log
2715 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2716 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2717 and DATA following a set of rejected RCPTs do not count as protocol errors.
2718 In other words, Exim assumes they were pipelined, though this may not
2719 actually be the case. Of course, in all cases the client gets an
2720 appropriate error code.
2722 15. If a lookup fails in an ACL condition, a message about the failure may
2723 be available; it is used if testing the ACL cannot continue, because most
2724 such messages specify what the cause of the deferral is. However, some
2725 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2726 that caused an old message to be retained and used if a later statement
2727 caused a defer, replacing the real cause of the deferral.
2729 16. If an IP address had so many PTR records that the DNS lookup buffer
2730 was not large enough to hold them, Exim could crash while trying to process
2731 the truncated data. It now detects and logs this case.
2733 17. Further to 4.21/58, another change has been made: if (and only if) the
2734 first line of a message (the first header line) ends with CRLF, a bare LF
2735 in a subsequent header line has a space inserted after it, so as not to
2736 terminate the header.
2738 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2739 unnecessarily, used atoi() instead of strtol(), and didn't check the
2740 termination when getting file sizes from file names by regex.
2742 19. Completely re-implemented the support for maildirsize files, in the light
2743 of a number of problems with the previous contributed implementation
2744 (4.30/29). In particular:
2746 . If the quota is zero, the maildirsize file is maintained, but no quota is
2749 . If the maildir directory does not exist, it is created before any attempt
2750 to write a maildirsize file.
2752 . The quota value in the file is just a cache; if the quota is changed in
2753 the transport, the new value overrides.
2755 . A regular expression is available for excluding directories from the
2758 20. The autoreply transport checks the characters in options that define the
2759 message's headers; it allows continued headers, but it was checking with
2760 isspace() after an embedded newline instead of explicitly looking for a
2763 21. If all the "regular" hosts to which an address was routed had passed their
2764 expiry times, and had not reached their retry times, the address was
2765 bounced, even if fallback hosts were defined. Now Exim should go on to try
2768 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2769 equivalent code in the SMTP transport. Some hosts send humungous responses
2770 to HELO/EHLO, more than 1024 it seems.
2772 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2773 gives compiler warnings in some environments. I've now done it "properly",
2776 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2777 (because of problems with the built-in one) was declared to return uschar *
2778 instead of char *, causing compiler failure.
2780 25. Fixed a file descriptor leak when processing alias/forward files.
2782 26. Fixed a minor format string issue in dbfn.c.
2784 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2786 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2787 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2790 29. When a sender address is verified, it is cached, to save repeating the test
2791 when there is more than one recipient in a message. However, when the
2792 verification involves a callout, it is possible for different callout
2793 options to be set for different recipients. It is too complicated to keep
2794 track of this in the cache, so now Exim always runs a verification when a
2795 callout is required, relying on the callout cache for the optimization.
2796 The overhead is duplication of the address routing, but this should not be
2799 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2800 to be verified with callout=postmaster, and the main callout worked but the
2801 postmaster check failed, the verification correctly failed. However, if a
2802 subsequent RCPT command asked for sender verification *without* the
2803 postmaster check, incorrect caching caused this verification also to fail,
2806 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2807 it was not caching the DNS options (qualify_single, search_parents) that
2808 were used when the lookup failed. A subsequent lookup with different
2809 options therefore always gave the same answer, though there were cases
2810 where it should not have. (Example: a "domains = !$mx_any" option on a
2811 dnslookup router: the "domains" option is always processed without any
2812 widening, but the router might have qualify_single set.) Now Exim uses the
2813 cached value only when the same options are set.
2815 32. Added John Jetmore's "exipick" utility to the distribution.
2817 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2818 than a timeout (e.g. a certificate is required, and is not provided), an
2819 Exim server now closes the connection immediately. Previously it waited for
2820 the client to close - but if the client is SSL, it seems that they each
2821 wait for each other, leading to a delay before one of them times out.
2823 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2824 maintained 0.8.x compatibility because I don't think many are using it, and
2825 it is clearly obsolete.
2827 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2830 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2831 client certificate was expired. A simple patch fixes this, though I don't
2832 understand the full logic of why the verify callback is called multiple
2835 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2836 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2837 which causes problems with some clients (such as the Certicom SSL Plus
2838 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2839 disables the coutermeasure allowing Eudora to connect."
2841 38. Exim was not checking that a write() to a log file succeeded. This could
2842 lead to Bad Things if a log got too big, in particular if it hit a file
2843 size limit. Exim now panics and dies if it cannot write to a log file, just
2844 as it does if it cannot open a log file.
2846 39. Modified OS/Makefile-Linux so that it now contains
2848 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2850 The two -D definitions ensure that Exim is compiled with large file
2851 support, which makes it possible to handle log files that are bigger than
2854 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2855 instance) a domain was checked against a named list that involved a lookup,
2856 causing $domain_data to be set, then another domain was checked against the
2857 same list, then the first domain was re-checked, the value of $domain_data
2858 after the final check could be wrong. In particular, if the second check
2859 failed, it could be set empty. This bug probably also applied to
2862 41. The strip_trailing_dot option was not being applied to the address given
2863 with the -f command-line option.
2865 42. The code for reading a message's header from the spool was incrementing
2866 $received_count, but never initializing it. This meant that the value was
2867 incorrect (doubled) while delivering a message in the same process in which
2868 it was received. In the most common configuration of Exim, this never
2869 happens - a fresh exec is done - but it can happen when
2870 deliver_drop_privilege is set.
2872 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2873 it now includes up to 150 characters of the unexpected data in the log
2876 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2877 and building data strings. The size of both of these buffers was 10 000
2878 bytes - far larger than anybody would *ever* want, thought I. Needless to
2879 say, somebody hit the limit. I have increased the maximum line length to
2880 20 000 and the maximum data length of concatenated lines to 100 000. I have
2881 also fixed two bugs, because there was no checking on these buffers. Tsk,
2882 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2883 buffer is too small.
2885 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2886 lsearch lookups. Now it does.
2888 46. When parsing a route_list item in a manualroute router, a fixed-length
2889 buffer was used for the list of hosts. I made this 1024 bytes long,
2890 thinking that nobody would ever have a list of hosts that long. Wrong.
2891 Somebody had a whole pile of complicated expansion conditions, and the
2892 string was silently truncated, leading to an expansion error. It turns out
2893 that it is easier to change to an unlimited length (owing to other changes
2894 that have happened since this code was originally written) than to build
2895 structure for giving a limitation error. The length of the item that
2896 expands into the list of hosts is now unlimited.
2898 47. The lsearch lookup could not handle data where the length of text line was
2899 more than 4095 characters. Such lines were truncated, leading to shortened
2900 data being returned. It should now handle lines of any length.
2902 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2903 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2906 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2907 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2908 no longer happens in comment lines. A list of replacements is now placed
2909 at the head of all of the source files, except those whose only change is
2910 to replace PERL_COMMAND in the very first #! line.
2912 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2913 messages on the queue, with a bottom-up merge sort, using code contributed
2914 by Michael Haardt. This should make operations like -bp somewhat faster on
2915 large queues. It won't affect queue runners, except when queue_run_in_order
2918 51. Installed eximstats 1.31 in the distribution.
2920 52. Added support for SRV lookups to the dnslookup router.
2922 53. If an ACL referred to $message_body or $message_body_end, the value was not
2923 reset for any messages that followed in the same SMTP session.
2925 54. The store-handling optimization for building very long strings was not
2926 differentiating between the different store pools. I don't think this
2927 actually made any difference in practice, but I've tidied it.
2929 55. While running the routers to verify a sender address, $sender_address
2930 was still set to the sender address. This is wrong, because when routing to
2931 send a bounce to the sender, it would be empty. Therefore, I have changed
2932 it so that, while verifying a sender address, $sender_address is set to <>.
2933 (There is no change to what happens when verifying a recipient address.)
2935 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2936 target A or AAAA records (if not already returned) without resetting the
2937 qualify_single or search_parents options of the DNS resolver. These are
2938 inappropriate in this case because the targets of MX and SRV records must
2939 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2940 target that, when qualified, matched something in the local domain. These
2941 two options are now turned off when doing these lookups.
2943 57. It seems that at least some releases of Reiserfs (which does not have the
2944 concept of a fixed number of inodes) returns zero and not -1 for the
2945 number of available inodes. This interacted badly with check_spool_inodes,
2946 which assumed that -1 was the "no such thing" setting. What I have done is
2947 to check that the total number of inodes is greater than zero before doing
2948 the test of how many are available.
2950 58. When a "warn" ACL statement has a log_message modifier, the message is
2951 remembered, and not repeated. This is to avoid a lot of repetition when a
2952 message has many recipients that cause the same warning to be written.
2953 Howewer, Exim was preserving the list of already written lines for an
2954 entire SMTP session, which doesn't seem right. The memory is now reset if a
2955 new message is started.
2957 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2958 debugging output unless log_rewrite was also set.
2960 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2961 of (int)(handle) when we know that handle contains (void *)(-1).
2963 61. The Exim daemon panic-logs an error return when it closes the incoming
2964 connection. However "connection reset by peer" seems to be common, and
2965 isn't really an error worthy of noting specially, so that particular error
2968 62. When Exim is trying to find all the local interfaces, it used to panic and
2969 die if the ioctl to get the interface flags failed. However, it seems that
2970 on at least one OS (Solaris 9) it is possible to have an interface that is
2971 included in the list of interfaces, but for which you get a failure error
2972 for this call. This happens when the interface is not "plumbed" into a
2973 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2974 failure of the "get flags" call assumes that the interface is down.
2976 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2977 makes life easier for people who are doing arithmetic on fields extracted
2978 from dates, where you often get leading zeros that should not be
2979 interpreted as octal.
2981 64. Added qualify_domain to the redirect router, to override the global
2984 65. If a pathologically long header line contained very many addresses (the
2985 report of this problem mentioned 10 000) and each of them was rewritten,
2986 Exim could use up a very large amount of memory. (It kept on making new
2987 copies of the header line as it rewrote, and never released the old ones.)
2988 At the expense of a bit more processing, the header rewriting function has
2989 been changed so that it no longer eats memory in this way.
2991 66. The generation of the Received: header has been moved from the time that a
2992 message starts to be received, to the time that it finishes. The timestamp
2993 in the Received: header should now be very close to that of the <= log
2994 line. There are two side-effects of this change:
2996 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2997 logged header lines no longer include the local Received: line, because
2998 it has not yet been created. The same applies to a copy of the message
2999 that is returned to a non-SMTP sender when a message is rejected.
3001 (b) When a filter file is tested using -bf, no additional Received: header
3002 is added to the test message. After some thought, I decided that this
3005 This change does not affect the value of $received_for. It is still set
3006 after address rewriting, but before local_scan() is called.
3008 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3010 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3011 gave an unhelpful panic error message, and a defer error. I have managed to
3012 change this behaviour so that it now rejects any supplied certificate,
3013 which seems right, as the list of acceptable certificates is empty.
3015 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3016 gave an unhelpful defer error. I have not managed to make this reject any
3017 supplied certificates, but the error message it gives is "no certificate
3018 supplied", which is not helpful.
3020 70. exigrep's output now also includes lines that are not associated with any
3021 message, but which match the given pattern. Implemented by a patch from
3022 Martin Sluka, which also tidied up the Perl a bit.
3024 71. Recipient callout verification, like sender verification, was using <> in
3025 the MAIL FROM command. This isn't really the right thing, since the actual
3026 sender may affect whether the remote host accepts the recipient or not. I
3027 have changed it to use the actual sender in the callout; this means that
3028 the cache record is now keyed on a recipient/sender pair, not just the
3029 recipient address. There doesn't seem to be a real danger of callout loops,
3030 since a callout by the remote host to check the sender would use <>.
3031 [SEE ABOVE: changed after hitting problems.]
3033 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3034 temporary errors. However, in the case of such a code being given after
3035 the end of a data transmission (i.e. after ".") Exim was failing to write
3036 a retry record for the message. (Yes, there was some broken host that was
3037 actually sending 8xx at this point.)
3039 73. An unknown lookup type in a host list could cause Exim to panic-die when
3040 the list was checked. (An example that provoked this was putting <; in the
3041 middle of a list instead of at the start.) If this happened during a DATA
3042 ACL check, a -D file could be left lying around. This kind of configuration
3043 error no longer causes Exim to die; instead it causes a defer errror. The
3044 incident is still logged to the main and panic logs.
3046 74. Buglet left over from Exim 3 conversion. The message "too many messages
3047 in one connection" was written to the rejectlog but not the mainlog, except
3048 when address rewriting (yes!) was being logged.
3050 75. Added write_rejectlog option.
3052 76. When a system filter was run not as root (that is, when system_filter_user
3053 was set), the values of the $n variables were not being returned to the
3054 main process; thus, they were not subsequently available in the $sn
3057 77. Added +return_path_on_delivery log selector.
3059 78. A connection timeout was being treated differently from recipients deferred
3060 when testing hosts_max_try with a message that was older than the host's
3061 retry timeout. (The host should not be counted, thus allowing all hosts to
3062 be tried at least once before bouncing.) This may have been the cause of an
3063 occasionally reported bug whereby a message would remain on the queue
3064 longer than the retry timeout, but would be bounced if a delivery was
3065 forced. I say "may" because I never totally pinned down the problem;
3066 setting up timeout/retry tests is difficult. See also the next item.
3068 79. The ultimate address timeout was not being applied to errors that involved
3069 a combination of host plus message (for example, a timeout on a MAIL
3070 command). When an address resolved to a number of possible hosts, and they
3071 were not all tried for each delivery (e.g. because of hosts_max_try), a
3072 message could remain on the queue longer than the retry timeout.
3074 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3077 81. Fixed an obscure SMTP outgoing bug which required at least the following
3078 conditions: (a) there was another message waiting for the same server;
3079 (b) the server returned 5xx to all RCPT commands in the first message so
3080 that the message was not completed; (c) the server dropped the connection
3081 or gave a negative response to the RSET that Exim sends to abort the
3082 transaction. The observed case was a dropped connection after DATA that had
3083 been sent in pipelining mode. That is, the server had advertised PIPELINING
3084 but was not implementing it correctly. The effect of the bug was incorrect
3085 behaviour, such as trying another host, and this could lead to a crash.
3091 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3092 and daemon.c were passed as pointers to ints; they should have been
3093 pointers to socklen_t variables (which are typically unsigned ints).
3095 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3098 3. Fixed a really odd bug that affected only the testing scheme; patching a
3099 certain fixed string in the binary changed the value of another string that
3100 happened to be identical to the end of the original first string.
3102 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3103 name", it returns that address as the IP address. On some operating
3104 systems (e.g. Solaris), it also passes back the IP address string as the
3105 "host name". However, on others (e.g. Linux), it passes back an empty
3106 string. Exim wasn't checking for this, and was changing the host name to an
3107 empty string, assuming it had been canonicized.
3109 5. Although rare, it is permitted to have more than one PTR record for a given
3110 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3111 all the names associated with an address, because they do in Solaris.
3112 However, it seems that they do not in Linux for data that comes from the
3113 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3114 I found this out when I moved to a new Linux workstation and tried to run
3115 the Exim test suite.
3117 To get round this problem I have changed the code so that it now does its
3118 own call to the DNS to look up PTR records when searching for a host name.
3119 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3120 addresses that are only in /etc/hosts are still found.
3122 This behaviour is, however, controlled by an option called host_lookup_
3123 order, which defaults to "bydns:byaddr". If people want to use the other
3124 order, or indeed, just use one or the other means of lookup, they can
3125 specify it in this variable.
3127 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3128 some operating systems, this comes back from gethostbyaddr() as an empty
3129 string, and this is what Exim used to test for. However, it seems that in
3130 other systems, "." is yielded. Exim now tests for this case too.
3132 7. The values of check_spool_space and check_log_space are now held internally
3133 as a number of kilobytes instead of an absolute number of bytes. If a
3134 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3135 kilobyte. This means that much larger values can be stored.
3137 8. Exim monitor: an attempt to get the action menu when not actually pointing
3138 at a message produces an empty menu entitled "No message selected". This
3139 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3140 no entries in it ("Shell widget menu has zero width and/or height"). So I
3141 have added a single, blank menu entry in this case.
3143 9. Added ${quote_local_part.
3145 10. MIME decoding is now applied to the contents of Subject: header lines when
3148 11. Now that a reference to $sender_host_address automatically causes a reverse
3149 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3150 host lookup before query-style lookups in lists that might use this
3151 variable. This has therefore been abolished, and the "net-" prefix is no
3152 longer necessary for query-style lookups.
3154 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3155 have been a typo for LFLAGS, so it has been changed.
3157 13. The install script calls Exim with "-C /dev/null" in order to find the
3158 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3159 to be output. Howeve, since Exim outputs its version number before the
3160 error, it didn't break the script. It just looked ugly. I fixed this by
3161 always allowing "-C /dev/null" if the caller is root.
3163 14. Ignore overlarge ACL variable number when reading spool file - insurance
3164 against a later release with more variables having written the file.
3166 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3167 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3168 this, as well as the form without the "IPv6" on the front.
3170 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3171 OS/Makefile-Darwin file.
3173 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3174 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3176 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3177 variable. However, some OS use ints for inodes. I've added cast to long int
3178 to get rid of the compiler warning.
3180 19. I had forgotten to lock out "/../" in configuration file names when
3181 ALT_CONFIG_PREFIX was set.
3183 20. Routers used for verification do not need to specify transports. However,
3184 if such a router generated a host list, and callout was configured, Exim
3185 crashed, because it could not find a port number from the (non-existent)
3186 transport. It now assumes port 25 in this circumstance.
3188 21. Added the -t option to exigrep.
3190 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3191 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3192 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3193 them is set, LOOKUP_LSEARCH is forced.
3195 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3197 24. Added sender and host information to the "rejected by local_scan()" log
3198 line; previously there was no indication of these.
3200 25. Added .include_if_exists.
3202 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3203 Linux. It turns out that not all file systems support this. Apparently some
3204 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3205 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3206 not supported on the file descriptor, is now ignored when Exim is trying to
3207 sync a directory. This applies only to Linux.
3209 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3211 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3212 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3215 29. Added support for maildirsize files from supplied patch (modified a bit).
3217 30. The use of :fail: followed by an empty string could lead Exim to respond to
3218 sender verification failures with (e.g.):
3220 550 Verification failed for <xxx>
3221 550 Sender verify failed
3223 where the first response line was missing the '-' that indicates it is not
3224 the final line of the response.
3226 31. The loop for finding the name of the user that called Exim had a hardwired
3227 limit of 10; it now uses the value of finduser_retries, which is used for
3228 all other user lookups.
3230 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3233 33. Exim was neglecting to zero errno before one call of strtol() when
3234 expanding a string and expecting an integer value. On some systems this
3235 resulted in spurious "integer overflow" errors. Also, it was casting the
3236 result into an int without checking.
3238 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3239 did not work. The code looks as if it has *never* worked, though it appears
3240 to have been documented since at least releast 1.62. I have made it work.
3242 35. The "timeout_DNS" error in retry rules, also documented since at least
3243 1.62, also never worked. As it isn't clear exactly what this means, and
3244 clearly it isn't a major issue, I have abolished the feature by treating it
3245 as "timeout", and writing a warning to the main and panic logs.
3247 36. The display of retry rules for -brt wasn't always showing the error code
3250 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3251 timeout_connect_A, timeout_connect_MX.
3253 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3254 to the empty sender.
3256 39. The daemon was not analysing the content of -oX till after it had closed
3257 stderr and disconnected from the controlling terminal. This meant that any
3258 syntax errors were only noted on the panic log, and the return code from
3259 the command was 0. By re-arranging the code a little, I've made the
3260 decoding happen first, so such errors now appear on stderr, and the return
3261 code is 1. However, the actual setting up of the sockets still happens in
3262 the disconnected process, so errors there are still only recorded on the
3265 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3266 connections (as happens on some IP stacks) was logged at start up time as
3267 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3268 it from "IPv6 and IPv4", which means that two separate sockets are being
3271 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3272 says whether AF_INET or AF_INET6 was passed as an argument.
3274 42. Exiwhat output was messed up when time zones were included in log
3277 43. Exiwhat now gives more information about the daemon's listening ports,
3278 and whether -tls-on-connect was used.
3280 44. The "port" option of the smtp transport is now expanded.
3282 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3283 silently ignored. Now an error message is written to the main and panic
3286 46. There's a new ACL modifier called "logwrite" which writes to a log file
3287 as soon as it is encountered.
3289 47. Added $local_user_uid and $local_user_gid at routing time.
3291 48. Exim crashed when trying to verify a sender address that was being
3294 49. Exim was recognizing only a space character after ".include". It now also
3295 recognizes a tab character.
3297 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3298 extracting the relevant information from the specification. The man page no
3299 longer contains scrambled data for the -d option, and I've added a section
3300 at the front about calling Exim under different names.
3302 51. Added "extra_headers" argument to the "mail" command in filter files.
3304 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3307 53. Installed eximstats 1.29.
3309 54. Added transport_filter_timeout as a generic transport option.
3311 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3312 Cc: header lines. This was required by RFC 822, but it not required by RFC
3315 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3316 incoming messages that did not have them. Now it does so only if the
3317 message originates locally, that is, if there is no associated remote host
3318 address. When Resent- header lines are present, this applies to the Resent-
3319 lines rather than the non-Resent- lines.
3321 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3322 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3324 58. Messages for configuration errors now include the name of the main
3325 configuration file - useful now that there may be more than one file in a
3326 list (.included file names were always shown).
3328 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3329 for those rare installations that do not start the daemon as root or run it
3330 setuid root. I've cut out the call to initgroups() if the daemon is not
3333 60. The Exim user and group can now be bound into the binary as text strings
3334 that are looked up at the start of Exim's processing.
3336 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3338 62. Added $mailstore_basename variable.
3340 63. Installed patch to sieve.c from Michael Haardt.
3342 64. When Exim failed to open the panic log after failing to open the main log,
3343 the original message it was trying to log was written to stderr and debug
3344 output, but if they were not available (the usual case in production), it
3345 was lost. Now it is written to syslog before the two lines that record the
3346 failures to open the logs.
3348 65. Users' Exim filters run in subprocesses under the user's uid. It is
3349 possible for a "deliver" command or an alias in a "personal" command to
3350 provoke an address rewrite. If logging of address rewriting is configured,
3351 this fails because the process is not running as root or exim. There may be
3352 a better way of dealing with this, but for the moment (because 4.30 needs
3353 to be released), I have disabled address rewrite logging when running a
3354 filter in a non-root, non-exim process.
3360 1. The buildconfig auxiliary program wasn't quoting the value set for
3361 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3362 not defined. This bug was masked in 4.22 by the effect that was fixed in
3365 2. Some messages that were rejected after a message id was allocated were
3366 shown as "incomplete" by exigrep. It no longer does this for messages that
3367 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3369 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3370 have allow_domain_literals set, the ID did not get logged in the <= line.
3371 Domain literals are now always recognized in Message-ID: header lines.
3373 4. The first argument for a ${extract expansion item is the key name or field
3374 number. Leading and trailing spaces in this item were not being ignored,
3375 causing some misleading effects.
3377 5. When deliver_drop_privilege was set, single queue runner processes started
3378 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3379 same command in the process it spins off) were not dropping privilege.
3381 6. When the daemon running as "exim" started a queue runner, it always
3382 re-executed Exim in the spun-off process. This is a waste of effort when
3383 deliver_drop_privilege is set. The new process now just calls the
3384 queue-runner function directly.
3390 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3393 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3394 ignored. Though the use of -oP was forcing the writing of a pid file, it
3395 was always written to the default place.
3397 3. If the message "no IP address found for host xxxx" is generated during
3398 incoming verification, it is now followed by identification of the incoming
3399 connection (so you can more easily find what provoked it).
3401 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3403 5. Added some features to "harden" Exim a bit more against certain attacks:
3405 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3406 be put in Local/Makefile. This is like the never_users runtime option,
3407 but it cannot be overridden. The default setting is "root".
3409 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3410 prefix string with which any file named in a -C command line option
3413 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3414 is retained for -C and -D only if the caller of Exim is root. Without
3415 it, the exim user may also use -C and -D and retain privilege.
3417 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3418 command line option is disabled.
3420 6. Macro names set by the -D option must start with an upper case letter, just
3421 like macro names defined in the configuration file.
3423 7. Added "dereference=" facility to LDAP.
3425 8. Two instances of the typo "uknown" in the source files are fixed.
3427 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3428 the Configure-Makefile script screwed up while processing it.
3430 10. Incorporated PCRE 4.4.
3432 11. The SMTP synchronization check was not operating right at the start of an
3433 SMTP session. For example, it could not catch a HELO sent before the client
3434 waited for the greeting. There is now a check for outstanding input at the
3435 point when the greeting is written. Because of the duplex, asynchronous
3436 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3437 way, but not yet received, when the check is performed.
3439 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3440 on TCP/IP sockets, because this apparently causes some broken clients to
3443 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3444 unchanged) from the Cygwin maintainer.
3446 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3447 was supported instead of testing for mbx. Effectively a typo.
3449 15. The spa authenticator server code was not checking that the input it
3450 received was valid base64.
3452 16. The debug output line for the "set" modifier in ACLs was not showing the
3453 name of the variable that was being set.
3455 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3457 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3458 lookup on its own. Something else was needed to trigger the lookup. For
3459 example, a match in host_lookup or the need for a host name in a host list.
3460 Now, if $sender_host_name is referenced and the host name has not yet been
3461 looked up, a lookup is performed. If the lookup fails, the variable remains
3462 empty, and $host_lookup_failed is set to "1".
3464 19. Added "eqi" as a case-independent comparison operator.
3466 20. The saslauthd authentication condition could segfault if neither service
3467 nor realm was specified.
3469 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3470 error message that was logged was misleading, and incoming SMTP
3471 connections were dropped. The message is now more accurate, and temporary
3472 errors are given to SMTP connections.
3474 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3475 (except RSET) with a 421 error, until QUIT is received. However, it was
3476 failing to send a response to QUIT.
3478 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3479 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3480 ACL is now run after the helo_try_verify_hosts code.
3482 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3483 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3484 case-independent, so other case variants are also recognized). Apparently
3485 some systems use these upper case variants.
3487 25. If more than two messages were waiting for the same host, and a transport
3488 filter was specified for the transport, Exim sent two messages over the
3489 same TCP/IP connection, and then failed with "socket operation on non-
3490 socket" when it tried to send the third.
3492 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3494 27. The extern definition of crypt16() in expand.c was not being excluded when
3495 the OS had its own crypt16() function.
3497 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3498 as a preferred synonym for return_size_limit, both as an option and as an
3501 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3503 30. Changed the default configuration ACL to relax the local part checking rule
3504 for addresses that are not in any local domains. For these addresses,
3505 slashes and pipe symbols are allowed within local parts, but the sequence
3506 /../ is explicitly forbidden.
3508 31. SPA server authentication was not clearing the challenge buffer before
3511 32. log_message in a "warn" ACL statement was writing to the reject log as
3512 well as to the main log, which contradicts the documentation and doesn't
3513 seem right (because no rejection is happening). So I have stopped it.
3515 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3516 However, I am unable to do any testing of this.
3518 34. Fixed an infelicity in the appendfile transport. When checking directories
3519 for a mailbox, to see if any needed to be created, it was accidentally
3520 using path names with one or more superfluous leading slashes; tracing
3521 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3523 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3524 contents are added to the log line that is written for every discarded
3525 recipient. (Previously a log_message setting was ignored.)
3527 36. The ${quote: operator now quotes the string if it is empty.
3529 37. The install script runs exim in order to find its version number. If for
3530 some reason other than non-existence or emptiness, which it checks, it
3531 could not run './exim', it was installing it with an empty version number,
3532 i.e. as "exim-". This error state is now caught, and the installation is
3535 38. An argument was missing from the function that creates an error message
3536 when Exim fails to connect to the socket for saslauthd authentication.
3537 This could cause Exim to crash, or give a corrupted message.
3539 39. Added isip, isip4, and isip6 to ${if conditions.
3541 40. The ACL variables $acl_xx are now saved with the message, and can be
3542 accessed later in routers, transports, and filters.
3544 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3545 strings in the file are not string-expanded.
3547 42. If a MAIL command specified a SIZE value that was too large to fit into an
3548 int variable, the check against message_size_limit failed. Such values are
3549 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3550 day this will have to be increased, but I don't think I want to be around
3551 when emails are that large.
3558 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3559 iconv() is not standard in FreeBSD.
3561 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3562 IPv6 enabled. The observed symptom was a segmentation fault on return from
3563 the function os_common_find_running_interfaces() in src/os.c.
3565 3. In the check_special_case() function in daemon.c I had used "errno" as an
3566 argument name, which causes warnings on some systems. This was basically a
3567 typo, since it was named "eno" in the comments!
3569 4. The code that waits for the clock to tick (at a resolution of some fraction
3570 of a second) so as to ensure message-id uniqueness was always waiting for
3571 at least one whole tick, when it could have waited for less. [This is
3572 almost certainly not relevant at current processor speeds, where it is
3573 unlikely to ever wait at all. But we try to future-proof.]
3575 5. The function that sleeps for a time interval that includes fractions of a
3576 second contained a race. It did not block SIGALRM between setting the
3577 timer, and suspending (a couple of lines later). If the interval was short
3578 and the sigsuspend() was delayed until after it had expired, the suspension
3579 never ended. On busy systems this could lead to processes getting stuck for
3582 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3583 process, before it forks any delivery processes. The open lookup caching
3584 mechanism meant that the open file or database connection was passed into
3585 the delivery process. The problem was that delivery processes always tidy
3586 up cached lookup data. This could cause a problem for the next delivery
3587 process started by the queue runner, because the external queue runner
3588 process does not know about the closure. So the next delivery process
3589 still has data in the lookup cache. In the case of a file lookup, there was
3590 no problem because closing a file descriptor in a subprocess doesn't affect
3591 the parent. However, if the lookup was caching a connection to a database,
3592 the connection was closed, and the second delivery process was likely to
3593 see errors such as "PGSQL: query failed: server closed the connection
3594 unexpectedly". The problem has been fixed by closing all cached lookups
3595 in a queue runner before running a delivery process.
3597 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3598 seem to have the "const" qualifier which it has on other OS. I've
3601 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3602 *of the same type* (client or server) with the same public name that an
3603 error should be diagnosed.
3605 9. When Exim looked up a host name for an IP address, but failed to find the
3606 original IP address when looking up the host name (a safety check), it
3607 output the message "<ip address> does not match any IP for NULL", which was
3608 confusing, to say the least. The bug was that the host name should have
3609 appeared instead of "NULL".
3611 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3612 user that is built into the binary, and the -C or -D options is used, root
3613 privilege is dropped before the configuration file is read. In addition,
3614 logging is switched to stderr instead of the normal log files. If the
3615 configuration then re-defines the Exim user, the unprivileged environment
3616 is probably not what is expected, so Exim logs a panic warning message (but
3619 However, if deliver_drop_privilege is set, the unprivileged state may well
3620 be exactly what is intended, so the warning has been cut out in that case,
3621 and Exim is allowed to try to write to its normal log files.
3627 1. smtp_return_error_details was not giving details for temporary sender
3628 or receiver verification errors.
3630 2. Diagnose a configuration error if two authenticators have the same public
3633 3. Exim used not to create the message log file for a message until the first
3634 delivery attempt. This could be confusing when incoming messages were held
3635 for policy or load reasons. The message log file is now created at the time
3636 the message is received, and an initial "Received" line is written to it.
3638 4. The automatically generated man page for command line options had a minor
3639 bug that caused no ill effects; however, a more serious problem was that
3640 the procedure for building the man page automatically didn't always
3641 operate. Consequently, release 4.20 contains an out-of-date version. This
3642 shouldn't happen again.
3644 5. When building Exim with embedded Perl support, the script that builds the
3645 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3646 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3648 6. The freeze_tell option was not being used for messages that were frozen on
3649 arrival, either by an ACL or by local_scan().
3651 7. Added the smtp_incomplete_transaction log selector.
3653 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3654 was accepting AUTH without a new EHLO.
3656 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3657 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3660 10. Logging of TCP/IP connections (when configured) now happens in the main
3661 daemon process instead of the child process, so that the TCP/IP connection
3662 count is more accurate (but it can never be perfect).
3664 11. The use of "drop" in a nested ACL was not being handled correctly in the
3665 outer ACL. Now, if condition failure induced by the nested "drop" causes
3666 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3667 or "require"), the connection is dropped.
3669 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3670 that yield "discard" can now be used with an "accept" or a "discard" verb,
3671 but an error is generated for any others (because I can't see a useful way
3672 to define what should happen).
3674 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3675 are now processed in the same way as lines in the Exim configuration file.
3676 In particular, continuation lines are supported.
3678 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3680 15. Added -ti meaning -t -i.
3682 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3683 domains, and warn by logging if others are found.
3685 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3686 returned from GIFCONF when Exim is trying to find the list of interfaces on
3687 a host. The code in os.c has been modified to copy each ifreq to an aligned
3688 structure in all cases.
3690 Also, in some cases, the returned ifreq's were being copied to a 'struct
3691 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3692 means the last couple of bytes of an IPv6 address could be chopped if the
3693 ifreq contained only a normal sockaddr (14 bytes storage).
3695 18. Named domain lists were not supported in the hosts_treat_as_local option.
3696 An entry such as +xxxx was not recognized, and was treated as a literal
3699 19. Ensure that header lines added by a DATA ACL are included in the reject log
3700 if the ACL subsequently rejects the message.
3702 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3703 MD5 (which is deprecated).
3705 21. When testing a filter file using -bf, Exim was writing a message when it
3706 took the sender from a "From " line in the message, but it was not doing so
3707 when it took $return_path from a Return-Path: header line. It now does.
3709 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3710 with a valid header line field name (a series of printing characters
3711 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3713 23. Changed "disc" in the source to "disk" to conform to the documentation and
3714 the book and for uniformity.
3716 24. Ignore Sendmail's -Ooption=value command line item.
3718 25. When execve() failed while trying to run a command in a pipe transport,
3719 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3720 could be confused with a return value of 69 from the command itself. This
3721 has been changed to 127, the value the shell returns if it is asked to run
3722 a non-existent command. The wording for the related log line suggests a
3723 non-existent command as the problem.
3725 26. If received_header_text expands to an empty string, do not add a Received:
3726 header line to the message. (Well, it adds a token one on the spool, but
3727 marks it "old" so that it doesn't get used or transmitted.)
3729 27. Installed eximstats 1.28 (addition of -nt option).
3731 28. There was no check for failure on the call to getsockname() in the daemon
3732 code. This can fail if there is a shortage of resources on the system, with
3733 ENOMEM, for example. A temporary error is now given on failure.
3735 29. Contrary to the C standard, it seems that in some environments, the
3736 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3737 program. Exim now does this explicitly; it affects the formatting of
3738 timestamps using strftime().
3740 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3741 complaints. I've now initialized all the variables, to avoid this.
3743 32. Header lines added by a system filter were not being "seen" during
3744 transport-time rewrites.
3746 33. The info_callback() function passed to OpenSSL is set up with type void
3747 (*)(SSL *, int, int), as described somewhere. However, when calling the
3748 function (actually a macro) that sets it up, the type void(*)() is
3749 expected. I've put in a cast to prevent warnings from picky compilers.
3751 34. If a DNS black list lookup found a CNAME record, but there were no A
3752 records associated with the domain it pointed at, Exim crashed.
3754 35. If a DNS black list lookup returned more than one A record, Exim ignored
3755 all but the first. It now scans all returned addresses if a particular IP
3756 value is being sought. In this situation, the contents of the
3757 $dnslist_value variable are a list of all the addresses, separated by a
3760 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3761 to accept a host name and all its aliases if the forward lookup for any of
3762 them yielded the IP address of the incoming connection. Now it accepts only
3763 those names whose forward lookup yields the correct IP address. Any other
3764 names are discarded. This closes a loophole whereby a rogue DNS
3765 administrator could create reverse DNS records to break through a
3766 wildcarded host restriction in an ACL.
3768 37. If a user filter or a system filter that ran in a subprocess used any of
3769 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3770 the wrong values were passed to the pipe command ($thisaddress had the
3771 value of $0, $0 had the value of $1, etc). This bug was introduced by
3772 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3774 38. Improved the line breaking for long SMTP error messages from ACLs.
3775 Previously, if there was no break point between 40 and 75 characters, Exim
3776 left the rest of the message alone. Two changes have been made: (a) I've
3777 reduced the minimum length to 35 characters; (b) if it can't find a break
3778 point between 35 and 75 characters, it looks ahead and uses the first one
3779 that it finds. This may give the occasional overlong line, but at least the
3780 remaining text gets split now.
3782 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3783 file descriptors might be low, and that setting 1000 would always raise it.
3784 It turns out that in some environments, the limit is already over 1000 and
3785 that lowering it causes trouble. So now Exim takes care not to decrease it.
3787 40. When delivering a message, the value of $return_path is set to $sender_
3788 address at the start of routing (routers may change the value). By an
3789 oversight, this default was not being set up when an address was tested by
3790 -bt or -bv, which affected the outcome if any router or filter referred to
3793 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3794 log only when "message" or "log_message" are set. However, if one of the
3795 conditions was an address verification, or a call to a nested ACL, the
3796 messages generated by the underlying test were being passed through. This
3797 no longer happens. The underlying message is available in $acl_verify_
3798 message for both "message" and "log_message" expansions, so it can be
3799 passed through if needed.
3801 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3802 new expansion $bh_ to give the encoded byte string without charset
3803 translation. Translation happens only if iconv() is available; HAVE_ICONV
3804 indicates this at build time. HEADERS_CHARSET gives the charset to
3805 translate to; headers_charset can change it in the configuration, and
3806 "headers charset" can change it in an individual filter file.
3808 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3809 that creates RFC 2047 encoded "words" labels them as that charset instead
3810 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3811 expansion operator; (ii) when Exim creates a From: line for a local
3812 message; (iii) when a header line is rewritten to include a "phrase" part.
3814 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3815 buggy, causing it to skip the first lines of messages whose message ID
3816 ended in 'D'. This would not have bitten before Exim release 4.14, because
3817 message IDs were unlikely to end in 'D' before then. The effect was to have
3818 incorrect size information for certain domains.
3820 45. #include "config.h" was missing at the start of the crypt16.c module. This
3821 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3824 46. If there was a timeout during a "random" callout check, Exim treated it as
3825 a failure of the random address, and carried on sending RSET and the real
3826 address. If the delay was just some slowness somewhere, the response to the
3827 original RCPT would be taken as a response to RSET and so on, causing
3828 mayhem of various kinds.
3830 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3831 when I implemented it. It didn't allow for the fact that some option values
3832 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3833 the right test for positive values.
3835 48. Domain names in DNS records are case-independent. Exim always looks them up
3836 in lower case. Some resolvers return domain names in exactly the case they
3837 appear in the zone file, that is, they may contain uppercase letters. Not
3838 all resolvers do this - some return always lower case. Exim was treating a
3839 change of case by a resolver as a change of domain, similar to a widening
3840 of a domain abbreviation. This triggered its re-routing code and so it was
3841 trying to route what was effectively the same domain again. This normally
3842 caused routing to fail (because the router wouldn't handle the domain
3843 twice). Now Exim checks for this case specially, and just changes the
3844 casing of the domain that it ultimately uses when it transmits the message
3847 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3850 50. If a filter generated a file delivery with a non-absolute name (possible if
3851 no home directory exists for the router), the forbid_file option was not
3854 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3855 the existing equality matching.
3857 52. Exim was using ints instead of ino_t variables in some places where it was
3858 dealing with inode numbers.
3860 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3861 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3862 called TMPDIR, and if it finds it is different, it changes its value.
3864 54. The smtp_printf() function is now made available to local_scan() so
3865 additional output lines can be written before returning. There is also an
3866 smtp_fflush() function to enable the detection of a dropped connection.
3867 The variables smtp_input and smtp_batched_input are exported to
3870 55. Changed the default runtime configuration: the message "Unknown user"
3871 has been removed from the ACL, and instead placed on the localuser router,
3872 using the cannot_route_message feature. This means that any verification
3873 failures that generate their own messages won't get overridden. Similarly,
3874 the "Unrouteable address" message that was in the ACL for unverifiable
3875 relay addresses has also been removed.
3877 56. Added hosts_avoid_esmtp to the smtp transport.
3879 57. The exicyclog script was not checking for the esoteric option
3880 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3881 will work only if exicyclog is run under the appropriate euid.
3883 58. Following a discussion on the list, the rules by which Exim recognises line
3884 endings on incoming messages have been changed. The -dropcr and drop_cr
3885 options are now no-ops, retained only for backwards compatibility. The
3886 following line terminators are recognized: LF CRLF CR. However, special
3887 processing applies to CR:
3889 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3890 nor a local message in the state where . is a terminator.
3892 (ii) If a bare CR is encountered in a header line, an extra space is added
3893 after the line terminator so as not to end the header. The reasoning
3894 behind this is that bare CRs in header lines are most likely either
3895 to be mistakes, or people trying to play silly games.
3897 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3898 was being incorrectly given as 18 bytes larger than it should have been.
3899 This is a VOB (very old bug).
3901 60. This may never have affected anything current, but just in case it has:
3902 When the local host is found other than at the start of a list of hosts,
3903 the local host, those with the same MX, and any that follow, are discarded.
3904 When the list in question was part of a longer list of hosts, the following
3905 hosts (not currently being processed) were also being discarded. This no
3906 longer happens. I'm not sure if this situation could ever has previously
3909 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3912 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3913 header. This is something that is recommended in a new Internet Draft, and
3914 is something that is documented as being done by Sendmail. There are two
3915 possible values. For messages generated by the autoreply transport, Exim
3918 Auto-Submitted: auto-replied
3920 whereas for all other generated messages (e.g. bounces) it adds
3922 Auto-Submitted: auto-generated
3924 63. The "personal" condition in filters now includes a test for the
3925 Auto-Submitted: header. If it contains the string "auto-" the message it
3926 not considered personal.
3928 64. Added rcpt_include_affixes as a generic transport option.
3930 65. Added queue_only_override (default true).
3932 66. Added the syslog_duplication option.
3934 67. If what should have been the first header line of a message consisted of
3935 a space followed by a colon, Exim was mis-interpreting it as a header line.
3936 It isn't of course - it is syntactically invalid and should therefore be
3937 treated as the start of the message body. The misbehaviour could have
3938 caused a number of strange effects, including loss of data in subsequent
3939 header lines, and spool format errors.
3941 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3942 client host had authenticated. This control can now be exercised by an ACL
3943 for more flexibility.
3945 69. By default, callouts do not happen when testing with -bh. There is now a
3946 variant, -bhc, which does actually run the callout code, including
3947 consulting and updating the callout cache.
3949 70. Added support for saslauthd authentication, courtesy of Alexander
3952 71. If statvfs() failed on the spool or log directories while checking their
3953 size for availability, Exim confusingly gave the error "space shortage".
3954 Furthermore, in debugging mode it crashed with a floating point exception.
3955 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3956 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3957 serious problem, Exim now writes to the main and panic logs when this
3958 happens, with details of the failure. It then refuses to accept the
3959 incoming message, giving the message "spool directory problem" or "log
3960 directory problem" with a 421 code for SMTP messages.
3962 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3963 0, 1, and 2 exist, because some OS complain for execs without them (see
3964 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3965 descriptors. However, the code omitted to check that the open succeeded,
3966 causing mysterious errors if for some reason the permissions on /dev/null
3967 got screwed. Now Exim writes a message to the main and panic logs, and
3968 bombs out if it can't open /dev/null.
3970 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3971 interact so that it is all more flexible. It is supposed to remain
3972 backwards compatible. Also added extra_local_interfaces.
3974 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3975 to bomb out with an assertion failure - to the client this appears as a
3976 connection drop. This problem occurs in the part of the code that was taken
3977 from the Samba project. Fortunately, the assertion is in a very simple
3978 function, so I have fixed this by reproducing the function inline in the
3979 one place where it is called, and arranging for authentication to fail
3980 instead of killing the process with assert().
3982 75. The SPA client code was not working when the server requested OEM rather
3983 than Unicode encoding.
3985 76. Added code to make require_files with a specific uid setting more usable in
3986 the case where statting the file as root fails - usually a non-root-mounted
3987 NFS file system. When this happens and the failure is EACCES, Exim now
3988 forks a subprocess and does the per-uid checking as the relevant uid.
3990 77. Added process_log_path.
3992 78. If log_file_path was not explicitly set, a setting of check_log_space or
3993 check_log_inodes was ignored.
3995 79. If a space check for the spool or log partitions fails, the incident is now
3996 logged. Of course, in the latter case the data may get lost...
3998 80. Added the %p formatting code to string_format() so that it can be used to
3999 print addresses in debug_print(). Adjusted all the address printing in the
4000 debugging in store.c to use %p rather than %d.
4002 81. There was a concern that a line of code in smtp_in.c could overflow a
4003 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4004 initially expressed, the concern was not well-founded, because trailing
4005 spaces are removed early. However, if the trailing spaces were followed by
4006 a NULL, they did not get removed, so the overflow was possible. Two fixes
4009 (a) I re-wrote the offending code in a cleaner fashion.
4010 (b) If an incoming SMTP command contains a NULL character, it is rejected
4013 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4014 runs initgroups(), so that if the Exim user is in any additional groups,
4015 they will be used during message reception.
4021 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob