1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.369 2006/07/06 14:28:03 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
24 with 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.
46 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
48 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
49 directory not expanded when it should be if an expanded home directory
50 was set for the address (which is overridden by the transport).
52 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
55 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
56 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
57 because it is too late at that time, and has no effect.
59 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
60 security issue with \' (bugzilla #107). I could not use the
61 PQescapeStringConn() function, because it needs a PGconn value as one of
64 PH/08 When testing addresses using -bt, indicate those final addresses that
65 are duplicates that would not cause an additional delivery. At least one
66 person was confused, thinking that -bt output corresponded to deliveries.
67 (Suppressing duplicates isn't a good idea as you lose the information
68 about possibly different redirections that led to the duplicates.)
70 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
71 systems where poll() doesn't work, in particular OS X.
73 PH/10 Added more information to debugging output for retry time not reached.
75 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
76 operations in malware.c.
78 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
85 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
86 other effects) broke the use of negated acl sub-conditions.
88 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
91 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
92 "Deny" causes Exim to reject the incoming connection with a 554 error.
93 Unfortunately, if there is a major crisis, such as a disk failure,
94 tcp-wrappers gives "deny", whereas what one would like would be some
95 kind of temporary error. A kludge has been added to help with this.
96 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
97 554 error is used if errno is still zero or contains ENOENT (which occurs
98 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
101 PH/03 Add -lutil to the default FreeBSD LIBS setting.
103 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
104 errors. Otherwise a message that provokes a temporary error (when other
105 messages do not) can cause a whole host to time out.
107 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
108 addresses were routed directly to files or pipes from a redirect router.
109 File deliveries just didn't batch; pipe deliveries might have suffered
112 PH/06 A failure to get a lock for a hints database would erroneously always say
113 "Failed to get write lock", even when it was really a read lock.
115 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
116 of 0600. This has been changed to use the value of the lockfile_mode
117 option (which defaults to 0600).
119 PH/08 Applied small patch from the Sieve maintainer.
121 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
122 folder from quota calculations, a direct delivery into this folder messed
123 up the contents of the maildirsize file. This was because the regex was
124 used only to exclude .Trash (or whatever) when the size of the mailbox
125 was calculated. There was no check that a delivery was happening into an
126 excluded directory. This bug has been fixed by ignoring all quota
127 processing for deliveries into excluded directories.
129 PH/10 Added the maildirfolder_create_regex option to appendfile.
135 PH/01 The code for finding all the local interface addresses on a FreeBSD
136 system running IPv6 was broken. This may well have applied to all BSD
137 systems, as well as to others that have similar system calls. The broken
138 code found IPv4 interfaces correctly, but gave incorrect values for the
139 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
140 that it would not match correctly against @[] and not recognize the IPv6
143 PH/02 The ipliteral router was not recognizing addresses of the form user@
144 [ipv6:....] because it didn't know about the "ipv6:" prefix.
146 PH/03 Added disable_ipv6.
148 PH/04 Changed $reply_address to use the raw form of the headers instead of the
149 decoded form, because it is most often used to construct To: headers
150 lines in autoreplies, and the decoded form may well be syntactically
151 invalid. However, $reply_address has leading white space removed, and all
152 newlines turned into spaces so that the autoreply transport does not
155 PH/05 If group was specified without a user on a router, and no group or user
156 was specified on a transport, the group from the router was ignored.
158 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
159 for visible compile-time settings that can be used to change these
160 numbers, for those that want even more. Backwards compatibility with old
161 spool files has been maintained. However, going back to a previous Exim
162 release will lost any variables that are in spool files.
164 PH/07 Two small changes when running in the test harness: increase delay when
165 passing a TCP/IP connection to a new process, in case the original
166 process has to generate a bounce, and remove special handling of
167 127.0.0.2 (sic), which is no longer necessary.
169 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
170 be the same on different OS.
172 PH/09 Moved a debug statement in filter processing to avoid a race problem when
175 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
176 whether --show-vars was specified or not
178 JJ/02 exipick: Added support for new ACL variable spool format introduced
181 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
182 syntactically invalid From: or Reply-to: line, and a filter used this to
183 generate an autoreply, and therefore failed to obtain an address for the
184 autoreply, Exim could try to deliver to a non-existent relative file
185 name, causing unrelated and misleading errors. What now happens is that
186 it logs this as a hard delivery error, but does not attempt to create a
189 PH/11 The exinext utility has a -C option for testing purposes, but although
190 the given file was scanned by exinext itself; it wasn't being passed on
191 when Exim was called.
193 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
194 an end-of-file indication when reading a command response.
196 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
197 compiled. In many other places in Exim, IPv6 addresses are always
198 recognized, so I have changed this. It also means that IPv4 domain
199 literals of the form [IPV4:n.n.n.n] are now always recognized.
201 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
202 used if the router is not running as root, for example, when verifying at
203 ACL time, or when using -bh. The debugging output from this situation was
204 non-existent - all you got was a failure to exec. I have made two
207 (a) Failures to set uid/gid, the current directory, or a process leader
208 in a subprocess such as that created by queryprogram now generate
209 suitable debugging ouput when -d is set.
211 (b) The queryprogram router detects when it is not running as root,
212 outputs suitable debugging information if -d is set, and then runs
213 the subprocess without attempting to change uid/gid.
215 PH/15 Minor change to Makefile for building test_host (undocumented testing
218 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
219 additional section of a DNS packet that returns MX or SRV records.
220 Instead, it always explicitly searches for A/AAAA records. This avoids
221 major problems that occur when a DNS server includes only records of one
222 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
223 fixed another bug: if SRV records were looked up and the corresponding
224 address records were *not* found in the additional section, the port
225 values from the SRV records were lost.
227 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
228 using the correct key (the original address) when searching the retry
229 rules in order to find which one to use for generating the retry hint.
231 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
232 adding the default one. Similarly, if it contains a Reply-To: header, the
233 errors_reply_to option, if set, is not used.
235 PH/19 When calculating a retry time, Exim used to measure the "time since
236 failure" by looking at the "first failed" field in the retry record. Now
237 it does not use this if it is later than than the arrival time of the
238 message. Instead it uses the arrival time. This makes for better
239 behaviour in cases where some deliveries succeed, thus re-setting the
240 "first failed" field. An example is a quota failure for a huge message
241 when small messages continue to be delivered. Without this change, the
242 "time since failure" will always be short, possible causing more frequent
243 delivery attempts for the huge message than are intended.
244 [Note: This change was subsequently modified - see PH/04 for 4.62.]
246 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
247 $1, $2, $3) because the numerical variables can be reset during some
248 expansion items (e.g. "match"), thereby losing the authentication data.
250 PH/21 Make -bV show the size of off_t variables so that the test suite can
251 decide whether to run tests for quotas > 2G.
253 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
254 mailbox_size, and mailbox_filecount in the appendfile transport. If a
255 filecount value is greater than 2G or if a quota value is greater than 2G
256 on a system where the size of off_t is not greater than 4, a panic error
259 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
260 never match. The debug and -bh output now contains an explicit error
261 message indicating a malformed IPv4 address or mask.
263 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
264 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
267 PH/25 Do not write to syslog when running in the test harness. The only
268 occasion when this arises is a failure to open the main or panic logs
269 (for which there is an explicit test).
271 PH/26 Added the /no_tell option to "control=freeze".
273 PH/27 If a host name lookup failed very early in a connection, for example, if
274 the IP address matched host_lookup and the reverse lookup yielded a name
275 that did not have a forward lookup, an error message of the form "no IP
276 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
277 could be logged. Now it outputs the IP address instead of "NULL".
279 PH/28 An enabling patch from MH: add new function child_open_exim2() which
280 allows the sender and the authenticated sender to be set when
281 submitting a message from within Exim. Since child_open_exim() is
282 documented for local_scan(), the new function should be too.
284 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
285 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
286 results in an empty string is now treated as unset.
288 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
290 PH/31 Added sender_verify_fail logging option.
292 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
293 needed by RFC 822 but not by RFC 2822 was commented out. I have now
294 tidied the source and removed it altogether.
296 PH/33 When a queue run was abandoned because the load average was too high, a
297 log line was always written; now it is written only if the queue_run log
298 selector is set. In addition, the log line for abandonment now contains
299 information about the queue run such as the pid. This is always present
300 in "start" and "stop" lines but was omitted from the "abandon" line.
302 PH/34 Omit spaces between a header name and the colon in the error message that
303 is given when verify = headers_syntax fails (if there are lots of them,
304 the message gets confusing).
306 PH/35 Change the default for dns_check_names_pattern to allow slashes within
307 names, as there are now some PTR records that contain slashes. This check
308 is only to protect against broken name servers that fall over on strange
309 characters, so the fact that it applies to all lookups doesn't matter.
311 PH/36 Now that the new test suite is complete, we can remove some of the
312 special code in Exim that was needed for the old test suite. For example,
313 sorting DNS records because real resolvers return them in an arbitrary
314 order. The new test suite's fake resolver always returns records in the
317 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
318 bounces) except when queue_only is set, to avoid logging races between
319 the different processes.
321 PH/38 Panic-die if .include specifies a non-absolute path.
323 PH/39 A tweak to the "H" retry rule from its user.
325 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
326 a label. They prevented compilation on older perls.
328 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
329 a warning to be raised on newish perls.
331 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
332 on queue. Changes to match documented behaviour of showing count of
333 messages matching specified criteria.
335 PH/40 Changed the default ident timeout from 30s to 5s.
337 PH/41 Added support for the use of login_cap features, on those BSD systems
338 that have them, for controlling the resources used by pipe deliveries.
340 PH/42 The content-scanning code uses fopen() to create files in which to put
341 message data. Previously it was not paying any attention to the mode of
342 the files. Exim runs with umask(0) because the rest of the code creates
343 files with open(), and sets the required mode explicitly. Thus, these
344 files were ending up world-writeable. This was not a big issue, because,
345 being within the spool directory, they were not world-accessible. I have
346 created a function called modefopen, which takes an additional mode
347 argument. It sets umask(777), creates the file, chmods it to the required
348 mode, then resets the umask. All the relevant calls to fopen() in the
349 content scanning code have been changed to use this function.
351 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
352 to 24 hours. This avoids potential overflow problems when processing G
353 and H retry rules. I suspect nobody ever tinkers with this value.
355 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
357 PH/45 When the plaintext authenticator is running as a client, the server's
358 challenges are checked to ensure they are valid base64 strings. By
359 default, the authentication attempt is cancelled if an invalid string is
360 received. Setting client_ignore_invalid_base64 true ignores these errors.
361 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
362 they are received. Thus, the responses can be made to depend on the
363 challenges. If an invalid string is ignored, an empty string is placed in
366 PH/46 Messages that are created by the autoreply transport now contains a
367 References: header, in accordance with RFCs 2822 and 3834.
369 PH/47 Added authenticated_sender_force to the smtp transport.
371 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
373 PH/49 Installed latest patch from the Sieve maintainer.
375 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
376 was also set, the appendfile transport was unnecessarily scanning a
377 directory of message files (e.g. for maildir delivery) to find the count
378 of files (along with the size), even though it did not need this
379 information. It now does the scan only if it needs to find either the
380 size of the count of files.
382 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
384 PH/52 Two bugs concerned with error handling when the smtp transport is
387 (i) Exim was not creating retry information for temporary errors given
388 for individual recipients after the DATA command when the smtp transport
389 was used in LMTP mode. This meant that they could be retried too
390 frequently, and not timed out correctly.
392 (ii) Exim was setting the flag that allows error details to be returned
393 for LMTP errors on RCPT commands, but not for LMTP errors for individual
394 recipients that were returned after the DATA command.
396 PH/53 This is related to PH/52, but is more general: for any failing address,
397 when detailed error information was permitted to be returned to the
398 sender, but the error was temporary, then after the final timeout, only
399 "retry timeout exceeded" was returned. Now it returns the full error as
400 well as "retry timeout exceeded".
402 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
403 do this, and (what is worse) MTAs that accept it.
405 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
406 will now be deprecated.
408 PH/56 New os.c-cygwin from the Cygwin maintainer.
410 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
411 formats (previously only available in exim formats via -bpr, -bpru,
412 and -bpra. Now also available in native and exiqgrep formats)
414 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
415 with very large, slow to parse queues
417 JJ/08 exipick: added ! as generic prefix to negate any criteria format
419 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
421 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
422 responses to authentication challenges, though it was showing the
423 challenges; (ii) I've removed the CR characters from the debug output for
426 PH/58 Allow for the insertion of a newline as well as a space when a string
427 is turned into more than one encoded-word during RFC 2047 encoding. The
428 Sieve code now uses this.
430 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
431 data_4xx, lost_connection, tls_required.
433 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
434 message was being sent as an SMTP response.
436 PH/61 Add -l and -k options to exicyclog.
438 PH/62 When verifying, if an address was redirected to one new address, so that
439 verification continued, and the new address failed or deferred after
440 having set something in $address_data, the value of $address_data was not
441 passed back to the ACL. This was different to the case when no
442 redirection occurred. The value is now passed back in both cases.
444 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
445 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
446 use login_cap.h, so on its own it isn't the distinguishing feature. The
447 new name refers directly to the setclassresources() function.
449 PH/65 Added configuration files for NetBSD3.
451 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
453 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
454 is preferred over IPv4.
456 PH/68 The bounce_return_message and bounce_return_body options were not being
457 honoured for bounces generated during the reception of non-SMTP messages.
458 In particular, this applied to messages rejected by the ACL. This bug has
459 been fixed. However, if bounce_return_message is true and bounce_return_
460 body is false, the headers that are returned for a non-SMTP message
461 include only those that have been read before the error was detected.
462 (In the case of an ACL rejection, they have all been read.)
464 PH/69 The HTML version of the specification is now built in a directory called
465 spec_html instead of spec.html, because the latter looks like a path with
466 a MIME-type, and this confuses some software.
468 PH/70 Catch two compiler warnings in sieve.c.
470 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
471 function verify_get_ident() calls ip_connect() to connect a socket, but
472 if the "connect()" function timed out, ip_connect() used to close the
473 socket. However, verify_get_ident() also closes the socket later, and in
474 between Exim writes to the log, which may get opened at this point. When
475 the socket was closed in ip_connect(), the log could get the same file
476 descriptor number as the socket. This naturally causes chaos. The fix is
477 not to close the socket in ip_connect(); the socket should be closed by
478 the function that creates it. There was only one place in the code where
479 this was missing, in the iplookup router, which I don't think anybody now
480 uses, but I've fixed it anyway.
482 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
483 well as to direct DNS lookups. Otherwise the handling of names in host
484 lists is inconsistent and therefore confusing.
490 PH/01 Two changes to the default runtime configuration:
492 (1) Move the checks for relay_from_hosts and authenticated clients from
493 after to before the (commented out) DNS black list checks.
495 (2) Add control=submission to the relay_from_hosts and authenticated
496 clients checks, on the grounds that messages accepted by these
497 statements are most likely to be submissions.
499 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
501 (1) Generate an error if the third argument for the ${prvs expansion is
504 (2) Treat a missing third argument of ${prvscheck as if it were an empty
507 (3) Reset the variables that are obtained from the first argument of
508 ${prvscheck and used in the second argument before leaving the code,
509 because their memory is reclaimed, so using them afterwards may do
512 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
513 one (it's much easier than Tom thought :-).
515 (5) Because of (4), we can now allow for the use of $prvscheck_result
516 inside the third argument.
518 PH/03 For some reason, the default setting of PATH when running a command from
519 a pipe transport was just "/usr/bin". I have changed it to
522 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
523 anything to be listed in the output from -bV.
525 PH/05 When a filter generated an autoreply, the entire To: header line was
526 quoted in the delivery log line, like this:
528 => >A.N.Other <ano@some.domain> <original@ddress> ...
530 This has been changed so that it extracts the operative address. There
531 may be more than one such address. If so, they are comma-separated, like
534 => >ano@some.domain,ona@other.domain <original@ddress> ...
536 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
537 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
538 not being looked up in the rDNS to get a host name, Exim was showing the
539 IP address twice in Received: lines, even though the IP addresses were
540 identical. For example:
542 Received: from [1.2.3.4] (helo=[1.2.3.4])
544 However, if the real host name was known, it was omitting the HELO data
545 if it matched the actual IP address. This has been tidied up so that it
546 doesn't show the same IP address twice.
548 PH/07 When both +timestamp and +memory debugging was on, the value given by
549 $tod_xxx expansions could be wrong, because the tod_stamp() function was
550 called by the debug printing, thereby overwriting the timestamp buffer.
551 Debugging no longer uses the tod_stamp() function when +timestamp is set.
553 PH/08 When the original message was included in an autoreply transport, it
554 always said "this is a copy of the message, including all the headers",
555 even if body_only or headers_only was set. It now gives an appropriate
558 PH/09 Applied a patch from the Sieve maintainer which:
560 o fixes some comments
561 o adds the (disabled) notify extension core
562 o adds some debug output for the result of if/elsif tests
563 o points to the current vacation draft in the documentation
564 and documents the missing references header update
568 o fixes a bug in processing the envelope test (when testing
569 multiple envelope elements, the last element determinted the
572 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
573 Electronic Mail") by including:
575 Auto-submitted: auto-generated
577 in the messages that it generates (bounce messages and others, such as
578 warnings). In the case of bounce messages for non-SMTP mesages, there was
579 also a typo: it was using "Auto_submitted" (underscore instead of
580 hyphen). Since every message generated by Exim is necessarily in response
581 to another message, thes have all been changed to:
583 Auto-Submitted: auto-replied
585 in accordance with these statements in the RFC:
587 The auto-replied keyword:
589 - SHOULD be used on messages sent in direct response to another
590 message by an automatic process,
592 - MUST NOT be used on manually-generated messages,
594 - MAY be used on Delivery Status Notifications (DSNs) and Message
595 Disposition Notifications (MDNs),
597 - MUST NOT be used on messages generated by automatic or periodic
598 processes, except for messages which are automatic responses to
601 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
602 to the default Received: header definition.
604 PH/12 Added log selector acl_warn_skipped (default on).
606 PH/13 After a successful wildlsearch lookup, discard the values of numeric
607 variables because (a) they are in the wrong storage pool and (b) even if
608 they were copied, it wouldn't work properly because of the caching.
610 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
611 checking when decoding. Apparently there are clients that generate
612 overlong encoded strings. Why am I not surprised?
614 PH/15 If the first argument of "${if match_address" was not empty, but did not
615 contain an "@" character, Exim crashed. Now it writes a panic log message
616 and treats the condition as false.
618 PH/16 In autoreply, treat an empty string for "once" the same as unset.
620 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
621 extension "envelope-auth". The code is finished and in agreement with
622 other implementations, but there is no documentation so far and in fact,
623 nobody wrote the draft yet. This extension is currently #undef'ed, thus
624 not changing the active code.
626 Print executed "if" and "elsif" statements when debugging is used. This
627 helps a great deal to understand what a filter does.
629 Document more things not specified clearly in RFC3028. I had all this
630 sorted out, when out of a sudden new issues came to my mind. Oops."
632 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
635 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
636 canonical form (as documented). However, after a host name lookup from
637 the IP address, check_host() was doing a simple string comparison with
638 addresses acquired from the DNS when checking that the found name did
639 have the original IP as one of its addresses. Since any found IPv6
640 addresses are likely to be in abbreviated form, the comparison could
641 fail. Luckily, there already exists a function for doing the comparison
642 by converting both addresses to binary, so now that is used instead of
645 PH/20 There was another similar case to PH/19, when a complete host name was
646 given in a host list; looking up its IP address could give an abbreviated
647 form, whereas the current host's name might or might not be abbreviated.
648 The same fix has been applied.
654 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
655 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
658 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
659 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
661 PH/03 Typo: missing ".o" in src/pcre/Makefile.
663 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
664 header line, restrict the check to what is listed in RFCs 2369 and 2929.
665 Also, for "Auto-Submitted", treat anything other than "no" as
666 non-personal, in accordance with RFC 3834. (Previously it treated
667 anything starting "auto-" as non-personal.)
669 TF/01 The control=submission/name=... option had a problem with syntax
670 errors if the name included a slash character. The /name= option
671 now slurps the rest of the string, so it can include any characters
672 but it must come last in the list of options (after /sender_retain
675 PH/05 Some modifications to the interface to the fake nameserver for the new
683 TK/01 Added the "success_on_redirect" address verification option. See
684 NewStuff for rationale and an example.
686 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
688 PH/02 Patch to exigrep to allow it to work on syslog lines.
690 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
691 fread() to skip over the body file's header line, because in Cygwin the
692 header line is locked and is inaccessible.
694 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
695 co-exist for some time) to make it clear that it is the Exim ID that is
696 referenced, not the Message-ID: header line.
698 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
699 string_format() function, because snprintf() does not exist on all
702 PH/06 The use of forbid_filter_existstest now also locks out the use of the
703 ${stat: expansion item.
705 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
706 protocol synchronization error", to keep the pedants happy.
708 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
709 well as for IRIX systems, when gcc is being used. See the host.c source
712 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
714 PH/10 Named domain lists were not working if used in a queue_smtp_domains
717 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
718 transport and to the smtp transport in LMTP mode.
720 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
722 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
723 run a filter in a subprocess. This could lead to confusion in subsequent
724 lookups in the parent process. There should also be a search_tidyup() at
725 the end of the subprocess.
727 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
728 only if the host matched helo_try_verify_hosts, which caused the
729 verification to occur when the EHLO/HELO command was issued. The ACL just
730 tested the remembered result. Now, if a previous verification attempt has
731 not happened, "verify = helo" does it there and then.
733 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
735 TK/03 Fix log output including CR from clamd.
737 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
738 exist provoked a memory error which could cause a segfault.
740 PH/15 Installed PCRE 6.2
742 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
744 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
745 of the problem. Specifically, suggested +O2 rather than +O1 for the
748 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
750 PH/20 If a delivery was routed to a non-standard port by means of an SRV
751 record, the port was not correctly logged when the outgoing_port log
752 selector was set (it logged the transort's default port).
754 PH/21 Added support for host-specific ports to manualroute, queryprogram,
755 fallback_hosts, and "hosts" in the smtp transport.
757 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
758 host errors such as "Connection refused".
760 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
761 authentication with radiusclient 0.4.9:
763 - Error returned from rc_read_config was caught wrongly
764 - Username/password not passed on to radius server due to wrong length.
766 The presumption is that some radiusclient API changes for 4.51/PH/17
767 were not taken care of correctly. The code is still untested by me (my
768 Linux distribution still has 0.3.2 of radiusclient), but it was
769 contributed by a Radius user.
771 PH/24 When doing a callout, the value of $domain wasn't set correctly when
772 expanding the "port" option of the smtp transport.
774 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
775 while reading a MIME header. Thanks to Tom Hughes for a patch.
777 PH/24 Include config.h inside local_scan.h so that configuration settings are
780 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
781 that in an ACL for RCPT (for example), you can examine exactly what was
784 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
785 commands, but it was not correctly comparing the address with the actual
786 client host address. Thus, it would show the EHLO address in Received:
787 header lines when this was not necessary.
789 PH/27 Added the % operator to ${eval:}.
791 PH/28 Exim tries to create and chdir to its spool directory when it starts;
792 it should be ignoring failures (because with -C, for example, it has lost
793 privilege). It wasn't ignoring creation failures other than "already
796 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
799 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
800 because an input line was too long, either on its own, or by virtue of
801 too many continuations, the temporary file was not being removed, and the
802 return code was incorrect.
804 PH/31 Missing "BOOL" in function definition in filtertest.c.
806 PH/32 Applied Sieve patches from the maintainer.
808 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
810 PH/33 Added "verify = not_blind".
812 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
813 Local/Makefile (with some defaults set). These are used in built scripts
814 such as exicyclog, but they have never been used in the exim_install
815 script (though there are many overriding facilities there). I have
816 arranged that the exim_install script now takes note of these two
819 PH/35 Installed configuration files for Dragonfly.
821 PH/36 When a locally submitted message by a trusted user did not contain a
822 From: header, and the sender address was obtained from -f or from an SMTP
823 MAIL command, and the trusted user did not use -F to supply a sender
824 name, $originator_name was incorrectly used when constructing a From:
825 header. Furthermore, $originator_name was used for submission mode
826 messages from external hosts without From: headers in a similar way,
827 which is clearly wrong.
829 PH/37 Added control=suppress_local_fixups.
831 PH/38 When log_selector = +received_sender was set, and the addition of the
832 sender made the log line's construction buffer exactly full, or one byte
833 less than full, an overflow happened when the terminating "\n" was
836 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
837 when the result of a list match is failure because a DNS lookup failed.
839 PH/40 RM_COMMAND is now used in the building process.
841 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
842 the "build-* directories that it finds.
844 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
845 address in a domain literal was a prefix of an interface address.
847 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
848 when verifying a sender address, unless rewrite_headers is false.
850 PH/44 Wrote a long comment about why errors_to addresses are verified as
851 recipients, not senders.
853 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
854 the ratelimit ACL was added.
856 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
858 PH/46 Added extra information about PostgreSQL errors to the error string.
860 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
861 avoiding the need to install special zones in a real server. This is
862 backwards compatible; if it can't find the fake resolver, it drops back.
863 Thus, both old and new test suites can be run.
865 TF/02 Added util/ratelimit.pl
867 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
868 clock is set back in time.
870 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
871 Candler <B.Candler@pobox.com>.
873 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
874 OK for addresses that are the result of redirections.
876 PH/48 A number of further additions for the benefit of the new test suite,
877 including a fake gethostbyname() that interfaces to the fake DNS resolver
880 TF/06 The fix for widen_domains has also been applied to qualify_single and
881 search_parents which are the other dnslookup options that can cause
884 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
887 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
889 TF/07 Exim produced the error message "an SRV record indicated no SMTP
890 service" if it encountered an MX record with an empty target hostname.
891 The message is now "an MX or SRV record indicated no SMTP service".
893 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
894 if the DNS of the sending site is misconfigured. This is quite a
895 common situation. This change restores the behaviour of treating a
896 helo verification defer as a failure.
898 PH/51 If self=fail was set on a router, the bounce message did not include the
899 actual error message.
905 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
907 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
908 command itself ended in error, the underlying message about the transport
909 filter timeout was being overwritten with the pipe command error. Now the
910 underlying error message should be appended to the second error message.
912 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
914 PH/02 Reduce the amount of output that "make" produces by default. Full output
915 can still be requested.
917 PH/03 The warning log line about a condition test deferring for a "warn" verb
918 was being output only once per connection, rather than after each
919 occurrence (because it was using the same function as for successful
920 "warn" verbs). This seems wrong, so I have changed it.
922 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
923 it should not have, which might have caused a crash in the right
924 circumstances, but probably never did.
926 PH/04 Installed a modified version of Tony Finch's patch to make submission
927 mode fix the return path as well as the Sender: header line, and to
928 add a /name= option so that you can make the user's friendly name appear
931 TF/03 Added the control = fakedefer ACL modifier.
933 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
934 Mark Lowes for thorough testing.
936 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
938 TK/03 Merged latest SRS patch from Miles Wilton.
940 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
941 with the definition in sysexits.h (which is #included earlier).
942 Fortunately, Exim does not actually use EX_OK. The code used to try to
943 preserve the sysexits.h value, by assumimg that macro definitions were
944 scanned for macro replacements. I have been disabused of this notion,
945 so now the code just undefines EX_OK before #including unistd.h.
947 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
948 in the smtp transport. When a block could not be written in a single
949 write() function, the timeout was being re-applied to each part-write.
950 This seems wrong - if the receiver was accepting one byte at a time it
951 would take for ever. The timeout is now adjusted when this happens. It
952 doesn't have to be particularly precise.
954 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
955 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
957 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
958 without a domain if the check to <postmaster@domain> fails.
960 SC/01 Eximstats: added -xls and the ability to specify output files
961 (patch written by Frank Heydlauf).
963 SC/02 Eximstats: use FileHandles for outputing results.
965 SC/03 Eximstats: allow any combination of xls, txt, and html output.
967 SC/04 Eximstats: fixed display of large numbers with -nvr option
969 SC/05 Eximstats: fixed merging of reports with empty tables.
971 SC/06 Eximstats: added the -include_original_destination flag
973 SC/07 Eximstats: removed tabs and trailing whitespace.
975 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
977 TK/06 MBOX spool code: Add real "From " MBOX separator line
978 so the .eml file is really in mbox format (even though
979 most programs do not really care). Patch from Alex Miller.
981 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
982 The latter is generated from $received_to and is only set if the
983 message has one envelope recipient. SA can use these headers,
984 obviously out-of-the-box. Patch from Alex Miller.
986 PH/08 The ${def test on a variable was returning false if the variable's
987 value was "0", contrary to what the specification has always said!
988 The result should be true unless the variable is empty.
990 PH/09 The syntax error of a character other than { following "${if
991 def:variable_name" (after optional whitespace) was not being diagnosed.
992 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
993 accidental colon was present, for example, could give incorrect results.
995 PH/10 Tidied the code in a number of places where the st_size field of a stat()
996 result is used (not including appendfile, where other changes are about
999 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1000 This involved changing a lot of size variables from int to off_t. It
1001 should work with maildirs and everything.
1003 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1004 spamd dying while we are connected to it.
1006 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1009 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1010 function for BDB 4.3.
1012 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1014 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1015 the sources for PCRE have been re-arranged and more files are now
1018 PH/15 The code I had for printing potentially long long variables in PH/11
1019 above was not the best (it lost precision). The length of off_t variables
1020 is now inspected at build time, and an appropriate printing format (%ld
1021 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1022 to be "long long int" or "long int". This is needed for the internal
1023 formatting function string_vformat().
1025 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1026 the configuration file to be ":syslog", then the script "guesses" where
1027 the logs files are, rather than using the compiled in default. In our
1028 case the guess is not the same as the compiled default, so the script
1029 suddenly stopped working when I started to use syslog. The patch checks
1030 to see if log_file_path is "". If so, it attempts to read it from exim
1031 with no configuration file to get the compiled in version, before it
1032 falls back to the previous guessing code."
1034 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1035 implementing BATV in an Exim configuration. See NewStuff for the gory
1038 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1039 Makefile that are specific to HP-UX.
1041 PH/18 If the "use_postmaster" option was set for a recipient callout together
1042 with the "random" option, the postmaster address was used as the MAIL
1043 FROM address for the random test, but not for the subsequent recipient
1044 test. It is now used for both.
1046 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1047 patch removes a few documentation additions to RFC 3028, because the
1048 latest draft now contains them. It adds the new en;ascii-case comparator
1049 and a new error check for 8bit text in MIME parts. Comparator and
1050 require names are now matched exactly. I enabled the subaddress
1051 extension, but it is not well tested yet (read: it works for me)."
1053 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1054 rework some of the code of TK/09 above to avoid the hardwired use of
1055 "%lld" and "long long". Replaced the call to snprintf() with a call to
1058 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1059 records point to non-existent hosts", "retry timeout exceeded", and
1060 "retry time not reached for any host after a long failure period".
1062 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1063 experimental DomainKeys support:
1065 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1066 (2) On an error such as an illegally used "control", the wrong name for
1067 the control was given.
1069 These problems did NOT occur unless DomainKeys support was compiled.
1071 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1073 PH/24 Added ${if match_ip condition.
1075 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1076 databases so that it will be absolutely obvious if a crash occurs in the
1077 DB library. This is a regular occurrence (often caused by mis-matched
1080 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1081 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1082 picked up on a user's system that detects such things. There doesn't seem
1083 to be a gcc warning option for this - only an attribute that has to be
1084 put on the function's prototype. It seems that in Fedora Core 4 they have
1085 set this on a number of new functions. No doubt there will be more in due
1088 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1089 specify a transport. However, if an address that was verified by such a
1090 router was the subject of a callout, Exim crashed because it tried to
1091 read the rcpt_include_affixes from the non-existent transport. Now it
1092 just assumes that the setting of that option is false. This bug was
1093 introduced by 4.51/PH/31.
1095 PH/28 Changed -d+all to exclude +memory, because that information is very
1096 rarely of interest, but it makes the output a lot bigger. People tend to
1097 do -d+all out of habit.
1099 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1100 code in os-type was giving problems when libc.so lives in lib64, like on
1103 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1104 aren't the modern standard, and it seems that some systems' include files
1105 don't always have them. Exim was already checking for some of the newer
1106 ones like T_AAAA, and defining it itself. I've added checks for all the
1107 record types that Exim uses.
1109 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1110 not automatically generating a new one, as it is supposed to. This
1111 prevented TLS from working. If the file did exist, but contained invalid
1112 data, a new version was generated, as expected. It was only the case of a
1113 non-existent file that was broken.
1115 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1116 with a change in libdomainkeys > 0.64.
1118 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1119 from DNS. If the selector record carries the flag, it now has
1120 precedence over the domain-wide flag.
1122 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1124 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1125 the use of a port name that isn't defined in /etc/services) occurred, the
1126 message was deferred as in a normal delivery, and thus remained on the
1127 spool, instead of being failed because of the mua_wrapper setting. This
1128 is now fixed, and I tidied up some of the mua_wrapper messages at the
1131 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1132 the messages in a hash of arrays rather than using individual hashes.
1133 This is a bit cleaner and results in dramatic memory savings, albeit
1134 at a slight CPU cost.
1136 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1137 as requested by Marc Sherman.
1139 SC/10 Eximstats: added histograms for user specified patterns as requested
1142 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1144 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1145 fopen() in the content-scanning modules that did not already have it.
1151 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1152 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1154 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1156 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1158 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1159 to test Sieve filters that use "vacation".
1161 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1162 that changes the way the GnuTLS parameters are stored in the cache file.
1163 The new format can be generated externally. For backward compatibility,
1164 if the data in the cache doesn't make sense, Exim assumes it has read an
1165 old-format file, and it generates new data and writes a new file. This
1166 means that you can't go back to an older release without removing the
1169 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1170 work if there are any delivery delays because "one_time" forces the
1171 parent to be marked "delivered", so its unseen clone is never tried
1172 again. For this reason, Exim now forbids the simultaneous setting of
1175 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1176 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1177 entry if you want to know the details. The fix, however, neglected to
1178 consider the case when local delivery batching is involved. The test for
1179 "previously delivered" was not happening when checking to see if an
1180 address could be batched with a previous (undelivered) one; under
1181 certain circumstances this could lead to multiple deliveries to the same
1184 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1185 in its include files, and this causes problems building Exim.
1187 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1188 header_syntax) but Exim was just ignoring anything given after a slash.
1189 In particular, this caused confusion with an attempt to use "verify =
1190 reverse_host_lookup/defer_ok". An error is now given when options are
1191 supplied for verify items that do not have them. (Maybe reverse_host_
1192 lookup should have a defer_ok option, but that's a different point.)
1194 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1195 defined by RFC 821) to 2048, because there were problems with some AUTH
1196 commands, and RFC 1869 says the size should be increased for extended
1197 SMTP commands that take arguments.
1199 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1202 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1203 "unknown" error; now it says that the functionality isn't in the binary.
1205 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1206 an address' error message when a string expansion fails (syntax or
1207 whatever). Otherwise the password may appear in the log. Following change
1208 PH/42 below, there is no longer a chance of it appearing in a bounce
1211 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1213 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1214 of its addresses was ever tried. (Bugzilla bug #2.)
1216 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1217 the result incorrectly in the debug output. (It correctly added a newline
1218 to what was transported.)
1220 TF/01 Added $received_time.
1222 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1223 commented out examples of how to interface to a virus scanner and to
1224 SpamAssassin. Also added commented examples of av_scanner and
1225 spamd_address settings.
1227 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1228 and controls are allowed in which ACLs. There were a couple of minor
1229 errors. Some of the entries in the conditions table (which is a table of
1230 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1231 negation of where the condition IS allowed.
1233 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1235 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1236 header file does not have a version number, so I've had to invent a new
1237 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1238 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1239 radiusclient), but it was contributed by a Radius user.
1241 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1242 files or directories, for OpenSSL.
1244 PH/19 When an Exim process that is running as root has to create an Exim log
1245 file, it does so in a subprocess that runs as exim:exim so as to get the
1246 ownership right at creation (otherwise, other Exim processes might see
1247 the file with the wrong ownership). There was no test for failure of this
1248 fork() call, which would lead to the process getting stuck as it waited
1249 for a non-existent subprocess. Forks do occasionally fail when resources
1250 run out. I reviewed all the other calls to fork(); they all seem to check
1253 PH/20 When checking for unexpected SMTP input at connect time (before writing
1254 the banner), Exim was not dealing correctly with a non-positive return
1255 from the read() function. If the client had disconnected by this time,
1256 the result was a log entry for a synchronization error with an empty
1257 string after "input=" when read() returned zero. If read() returned -1
1258 (an event I could not check), uninitialized data bytes were printed.
1259 There were reports of junk text (parts of files, etc) appearing after
1262 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1264 PH/22 Added support for macro redefinition, and (re)definition in between
1265 driver and ACL definitions.
1267 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1268 forgetting to use the resulting value; it was using the unexpanded value.
1270 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1271 hadn't been configured. The fix is from Juergen Kreileder, who
1272 understands it better than I do:
1274 "Here's what I see happening with three configured cyrus_sasl
1275 authenticators configured (plain, login, cram-md5):
1277 On startup auth_cyrus_sasl_init() gets called for each of these.
1278 This means three calls to sasl_listmech() without a specified mech_list.
1279 => SASL tests which mechs of all available mechs actually work
1280 => three warnings about OTP not working
1281 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1283 With the patch, sasl_listmech() also gets called three times. But now
1284 SASL's mech_list option is set to the server_mech specified in the the
1285 authenticator. Or in other words, the answer from sasl_listmech()
1286 gets limited to just the mech you're testing for (which is different
1288 => the return list contains just 'plain' or 'login', 'cram-md5' or
1289 nothing depending on the value of ob->server_mech.
1291 I've just tested the patch: Authentication still works fine,
1292 unavailable mechs specified in the exim configuration are still
1293 caught, and the auth.log warnings about OTP are gone."
1295 PH/25 When debugging is enabled, the contents of the command line are added
1296 to the debugging output, even when log_selector=+arguments is not
1299 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1300 answer is "GNU", and only if the return is "GNU/something" is the answer
1303 PH/27 $acl_verify_message is now set immediately after the failure of a
1304 verification in an ACL, and so is available in subsequent modifiers. In
1305 particular, the message can be preserved by coding like this:
1307 warn !verify = sender
1308 set acl_m0 = $acl_verify_message
1310 Previously, $acl_verify_message was set only while expanding "message"
1311 and "log_message" when a very denied access.
1313 PH/28 Modified OS/os.c-Linux with
1315 -#ifndef OS_LOAD_AVERAGE
1316 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1318 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1319 nomenclature these days.)
1321 PH/29 Installed patch from the Sieve maintainer that adds the options
1322 sieve_useraddress and sieve_subaddress to the redirect router.
1324 PH/30 In these circumstances:
1325 . Two addresses routed to the same list of hosts;
1326 . First host does not offer TLS;
1327 . First host accepts first address;
1328 . First host gives temporary error to second address;
1329 . Second host offers TLS and a TLS session is established;
1330 . Second host accepts second address.
1331 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1332 and peerdn, if requested) that were in fact used only for the second
1335 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1336 attention to any local part prefix or suffix that was matched by the
1337 router that accepted the address. It now behaves in the same way as it
1338 does for delivery: the affixes are removed from the local part unless
1339 rcpt_include_affixes is set on the transport.
1341 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1342 timeout during the DATA phase of an incoming message.
1344 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1345 applied a patch sanctioned by the Sieve maintainer.
1347 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1348 the uid or gid is negative. A case of a negative gid caused this to be
1349 noticed. The fix allows for either to be negative.
1351 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1352 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1355 PH/36 The change PH/12 above was broken. Fixed it.
1357 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1358 the grounds that routing the same address twice would always produce the
1359 same answer. This might have been true once, but it is certainly no
1360 longer true now. Routing a child address may depend on the previous
1361 routing that produced that child. Some complicated redirection strategies
1362 went wrong when messages had multiple recipients, and made Exim's
1363 behaviour dependent on the order in which the addresses were given.
1365 I have moved the duplicate checking until after the routing is complete.
1366 Exim scans the addresses that are assigned to local and remote
1367 transports, and removes any duplicates. This means that more work will be
1368 done, as duplicates will always all be routed, but duplicates are
1369 presumably rare, so I don't expect this is of any significance.
1371 For deliveries to pipes, files, and autoreplies, the duplicate checking
1372 still happens during the routing process, since they are not going to be
1375 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1376 It corrects a timeout issue with spamd. This is Ian's comment: "The
1377 background is that sometimes spamd either never reads data from a
1378 connection it has accepted, or it never writes response data. The exiscan
1379 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1380 blindly assumes that writes won't block so it may never time out."
1382 PH/39 Allow G after quota size as well as K and M.
1384 PH/40 The value set for $authenticated_id in an authenticator may not contain
1385 binary zeroes or newlines because the value is written to log lines and
1386 to spool files. There was no check on this. Now the value is run through
1387 the string_printing() function so that such characters are converted to
1388 printable escape sequences.
1390 PH/41 $message_linecount is a new variable that contains the total number of
1391 lines in the message. Compare $body_linecount, which is the count for the
1394 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1395 bounce and delay warning messages, except in certain special cases, which
1398 (a) An SMTP error message from a remote host;
1399 (b) A message specified in a :fail: redirection;
1400 (c) A message specified in a "fail" command in a system filter;
1401 (d) A message specified in a FAIL return from the queryprogram router;
1402 (e) A message specified by the cannot_route_message router option.
1404 In these cases only, Exim does include the error details in bounce and
1405 warning messages. There are also a few cases where bland messages such
1406 as "unrouteable address" or "local delivery error" are given.
1408 PH/43 $value is now also set for the "else" part of a ${run expansion.
1410 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1411 being worked on, but at least Exim now implements the latest version to
1414 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1415 process to complete was treated as a delivery failure, a timeout while
1416 writing the message to the pipe was logged, but erroneously treated as a
1417 successful delivery. Such timeouts include transport filter timeouts. For
1418 consistency with the overall process timeout, these timeouts are now
1419 treated as errors, giving rise to delivery failures by default. However,
1420 there is now a new Boolean option for the pipe transport called
1421 timeout_defer, which, if set TRUE, converts the failures into defers for
1422 both kinds of timeout. A transport filter timeout is now identified in
1425 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1426 systems where "make" and "gmake" are different, calling "gmake" at top
1427 level broke things. I've arranged for the value of $(MAKE) to be passed
1428 from the Makefile to this script so that it can call the same version of
1432 A note about Exim versions 4.44 and 4.50
1433 ----------------------------------------
1435 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1436 changes of various kinds. As a consequence, a big documentation update was
1437 needed. This delayed the release for rather longer than seemed good, especially
1438 in the light of a couple of (minor) security issues. Therefore, the changes
1439 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1440 release. So 4.44 and 4.50 are in effect two different branches that both start
1443 I have left the 4.50 change log unchanged; it contains all the changes since
1444 4.43. The change log for 4.44 is below; many of its items are identical to
1445 those for 4.50. This seems to be the most sensible way to preserve the
1446 historical information.
1452 1. Minor wording change to the doc/README.SIEVE file.
1454 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1455 computation of the current number of files was incorrect.
1457 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1458 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1461 4. Give more explanation in the error message when the command for a transport
1462 filter fails to execute.
1464 5. There are several places where Exim runs a non-Exim command in a
1465 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1466 was being done only for the command run by the queryprogram router. It is
1467 now done for all such subprocesses. The other cases are: ${run, transport
1468 filters, and the commands run by the lmtp and pipe transports.
1470 6. Added CONFIGURE_GROUP build-time option.
1472 7. Some older OS have a limit of 256 on the maximum number of file
1473 descriptors. Exim was using setrlimit() to set 1000 as a large value
1474 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1475 systems. I've change it so that if it can't get 1000, it tries for 256.
1477 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1478 was an oversight, and furthermore, ever since the addition of extra
1479 controls (e.g. 4.43/32), the checks on when to allow different forms of
1480 "control" were broken. There should now be diagnostics for all cases when a
1481 control that does not make sense is encountered.
1483 9. Added the /retain_sender option to "control=submission".
1485 10. $recipients is now available in the predata ACL (oversight).
1487 11. Tidy the search cache before the fork to do a delivery from a message
1488 received from the command line. Otherwise the child will trigger a lookup
1489 failure and thereby defer the delivery if it tries to use (for example) a
1490 cached ldap connection that the parent has called unbind on.
1492 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1493 of $address_data from the recipient verification was clobbered by the
1494 sender verification.
1496 13. The value of address_data from a sender verification is now available in
1497 $sender_address_data in subsequent conditions in the ACL statement.
1499 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1501 15. Added a new option "connect=<time>" to callout options, to set a different
1504 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1505 was its contents. (It was OK if the option was not defined at all.)
1507 17. A "Completed" log line is now written for messages that are removed from
1508 the spool by the -Mrm option.
1510 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1511 information about exactly what failed.
1513 19. Added -dd to debug only the daemon process.
1515 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1516 handles timeouts, both on the server side and network timeouts. Renamed the
1517 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1519 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1520 It is now set to "smtps".
1522 22. $host_address is now set to the target address during the checking of
1523 ignore_target_hosts.
1525 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1526 being passed; this would have caused $sender_host_name to have been used if
1527 matching the list had actually called for a host name (not very likely,
1528 since this list is usually IP addresses). A host name is now passed as
1531 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1532 code that specifies a non-restarting handler (typically sigaction() in
1533 modern systems) in an attempt to fix a rare and obscure crash bug.
1535 25. Narrowed the window for a race in the daemon that could cause it to ignore
1536 SIGCHLD signals. This is not a major problem, because they are used only to
1537 wake it up if nothing else does.
1539 26. A malformed maildirsize file could cause Exim to calculate negative values
1540 for the mailbox size or file count. Odd effects could occur as a result.
1541 The maildirsize information is now recalculated if the size or filecount
1544 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1545 support for a long time. Removed HAVE_SYS_VFS_H.
1547 28. Installed the latest version of exipick from John Jetmore.
1549 29. In an address list, if the pattern was not a regular expression, an empty
1550 subject address (from a bounce message) matched only if the pattern was an
1551 empty string. Non-empty patterns were not even tested. This was the wrong
1552 because it is perfectly reasonable to use an empty address as part of a
1553 database query. An empty address is now tested by patterns that are
1554 lookups. However, all the other forms of pattern expect the subject to
1555 contain a local part and a domain, and therefore, for them, an empty
1556 address still always fails if the pattern is not itself empty.
1558 30. Exim went into a mad DNS loop when attempting to do a callout where the
1559 host was specified on an smtp transport, and looking it up yielded more
1560 than one IP address.
1562 31. Re-factored the code for checking spool and log partition space into a
1563 function that finds that data and another that does the check. The former
1564 is then used to implement four new variables: $spool_space, $log_space,
1565 $spool_inodes, and $log_inodes.
1567 32. The RFC2047 encoding function was originally intended for short strings
1568 such as real names; it was not keeping to the 75-character limit for
1569 encoded words that the RFC imposes. It now respects the limit, and
1570 generates multiple encoded words if necessary. To be on the safe side, I
1571 have increased the buffer size for the ${rfc2047: expansion operator from
1574 33. It is now permitted to omit both strings after an "if" condition; if the
1575 condition is true, the result is "true". As before, when the second string
1576 is omitted, a false condition yields an empty string. This makes it less
1577 cumbersome to write custom ACL and router conditions.
1579 34. Failure to deliver a bounce message always caused it to be frozen, even if
1580 there was an errors_to setting on the router. The errors_to setting is now
1583 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1584 canonical form (fully expanded) before being placed in
1585 $sender_host_address.
1587 36. The table in the code that translates DNS record types into text (T_A to
1588 "A" for instance) was missing entries for NS and CNAME. It is just possible
1589 that this could have caused confusion if both these types were looked up
1590 for the same domain, because the text type is used as part of Exim's
1591 per-process caching. But the chance of anyone hitting this buglet seems
1594 37. The dnsdb lookup has been extended in a number of ways.
1596 (1) There is a new type, "zns", which walks up the domain tree until it
1597 finds some nameserver records. It should be used with care.
1599 (2) There is a new type, "mxh", which is like "mx" except that it returns
1600 just the host names, not the priorities.
1602 (3) It is now possible to give a list of domains (or IP addresses) to be
1603 looked up. The behaviour when one of the lookups defers can be
1604 controlled by a keyword.
1606 (4) It is now possible to specify the separator character for use when
1607 multiple records are returned.
1609 38. The dnslists ACL condition has been extended: it is now possible to supply
1610 a list of IP addresses and/or domains to be looked up in a particular DNS
1613 39. Added log_selector=+queue_time_overall.
1615 40. When running the queue in the test harness, wait just a tad after forking a
1616 delivery process, to get repeatability of debugging output.
1618 41. Include certificate and key file names in error message when GnuTLS fails
1619 to set them up, because the GnuTLS error message doesn't include the name
1620 of the failing file when there is a problem reading it.
1622 42. Allow both -bf and -bF in the same test run.
1624 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1626 44. The "Exiscan patch" is now merged into the mainline Exim source.
1628 45. Sometimes the final signoff response after QUIT could fail to get
1629 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1630 < 0 before doing a fflush(). This bug looks as though it goes back to the
1631 introduction of TLS in release 3.20, but "sometimes" must have been rare
1632 because the tests only now provoked it.
1634 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1635 (this can affect the format of dates).
1637 47. exim_tidydb, when checking for the continued existence of a message for
1638 which it has found a message-specific retry record, was not finding
1639 messages that were in split spool directories. Consequently, it was
1640 deleting retry records that should have stayed in existence.
1642 48. Steve fixed some bugs in eximstats.
1644 49. The SPA authentication driver was not abandoning authentication and moving
1645 on to the next authenticator when an expansion was forced to fail,
1646 contradicting the general specification for all authenticators. Instead it
1647 was generating a temporary error. It now behaves as specified.
1649 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1650 (the order specifies the preference for clients). The order is now AES256,
1651 AES128, 3DES, ARCFOUR128.
1653 51. Small patch to Sieve code - explicitly set From: when generating an
1656 52. Exim crashed if a remote delivery caused a very long error message to be
1657 recorded - for instance if somebody sent an entire SpamAssassin report back
1658 as a large number of 550 error lines. This bug was coincidentally fixed by
1659 increasing the size of one of Exim's internal buffers (big_buffer) that
1660 happened as part of the Exiscan merge. However, to be on the safe side, I
1661 have made the code more robust (and fixed the comments that describe what
1664 53. Now that there can be additional text after "Completed" in log lines (if
1665 the queue_time_overall log selector is set), a one-byte patch to exigrep
1666 was needed to allow it to recognize "Completed" as not the last thing in
1669 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1670 patch that reportedly fixes this has been added. I am not expert enough to
1671 create a test for it. This is what the patch creator wrote:
1673 "I found a little strange behaviour of ldap code when working with
1674 Windows 2003 AD Domain, where users was placed in more than one
1675 Organization Units. When I tried to give exim partial DN, the exit code
1676 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1677 But simultaneously result of request was absolutely normal ldap result,
1678 so I produce this patch..."
1680 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1681 REFERENCE, so I have modified the code to exclude the patch when that macro
1684 55. Some experimental protocols are using DNS PTR records for new purposes. The
1685 keys for these records are domain names, not reversed IP addresses. The
1686 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1687 leaves it alone. Component reversal etc. now happens only for IP addresses.
1690 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1692 57. Double the size of the debug message buffer (to 2048) so that more of very
1693 long debug lines gets shown.
1695 58. The exicyclog utility now does better if the number of log files to keep
1696 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1698 59. Two changes related to the smtp_active_hostname option:
1700 (1) $smtp_active_hostname is now available as a variable.
1701 (2) The default for smtp_banner uses $smtp_active_hostname instead
1702 of $primary_hostname.
1704 60. The host_aton() function is supposed to be passed a string that is known
1705 to be a valid IP address. However, in the case of IPv6 addresses, it was
1706 not checking this. This is a hostage to fortune. Exim now panics and dies
1707 if the condition is not met. A case was found where this could be provoked
1708 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1709 components; fortuitously, this particular loophole had already been fixed
1710 by change 4.50/55 above.
1712 If there are any other similar loopholes, the new check in host_aton()
1713 itself should stop them being exploited. The report I received stated that
1714 data on the command line could provoke the exploit when Exim was running as
1715 exim, but did not say which command line option was involved. All I could
1716 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1717 running as the user.
1720 61. There was a buffer overflow vulnerability in the SPA authentication code
1721 (which came originally from the Samba project). I have added a test to the
1722 spa_base64_to_bits() function which I hope fixes it.
1725 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1726 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1728 63. The daemon start-up calls getloadavg() while still root for those OS that
1729 need the first call to be done as root, but it missed one case: when
1730 deliver_queue_load_max is set with deliver_drop_privilege. This is
1731 necessary for the benefit of the queue runner, because there is no re-exec
1732 when deliver_drop_privilege is set.
1734 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1735 This has been fixed.
1737 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1738 was in use, was not putting the data itself into the right store pool;
1739 consequently, it could be overwritten for a subsequent message in the same
1740 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1743 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1745 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1746 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1747 the function were treating the return as a boolean value, which happened to
1748 work because 0=false and not-0=true, but is not correct code.
1750 68. The host_aton() function was not handling scoped IPv6 addresses (those
1751 with, for example, "%eth0" on the end) correctly.
1753 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1754 negated items (that is, ~something) in unsigned ints. Some compilers
1755 apparently mutter when there is no cast.
1757 70. If an address verification called from an ACL failed, and did not produce a
1758 user-specific message (i.e. there was only a "system" message), nothing was
1759 put in $acl_verify_message. In this situation, it now puts the system
1762 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1763 session; change 4.31/43 added the unwanted input to the log line - except
1764 that it did not do this in the start of session case. It now does.
1766 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1767 This is wrong and can cause the other end to generate a synchronization
1768 error if it is another Exim or anything else that does the synchronization
1769 check. A QUIT command is no longer sent after a timeout.
1771 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1772 during host lookups.
1774 74. The defer_ok option of callout verification was not working if it was used
1775 when verifying addresses in header lines, that is, for this case:
1777 verify = header_sender/callout=defer_ok
1779 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1780 those file descriptors could be used for SMTP connections. If anything
1781 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1782 could be sent to the SMTP client, causing chaos. The daemon now opens
1783 stdin, stdout, and stderr to /dev/null when it puts itself into the
1786 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1787 log by default. The user can override this with suitable Perl magic.
1789 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1790 the log message when discard triggers, was not working for the DATA ACL or
1791 for the non-SMTP ACL.
1793 78. Error message wording change in sieve.c.
1795 79. If smtp_accept_max_per_host was set, the number of connections could be
1796 restricted to fewer than expected, because the daemon was trying to set up
1797 a new connection before checking whether the processes handling previous
1798 connections had finished. The check for completed processes is now done
1799 earlier. On busy systems, this bug wouldn't be noticed because something
1800 else would have woken the daemon, and it would have reaped the completed
1803 80. If a message was submitted locally by a user whose login name contained one
1804 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1805 It caused a spool format error. I have fixed the spool reading code. A
1806 related problem was that the "from" clause in the Received: line became
1807 illegal because of the space(s). It is now covered by ${quote_local_part.
1809 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1812 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1814 83. Patch from Sieve maintainer to fix unterminated string problem in
1815 "vacation" handling.
1817 84. Some minor changes to the Linux configuration files to help with other
1818 OS variants using glibc.
1820 85. One more patch for Sieve to update vacation handling to latest spec.
1823 ----------------------------------------------------
1824 See the note above about the 4.44 and 4.50 releases.
1825 ----------------------------------------------------
1831 1. Change 4.43/35 introduced a bug that caused file counts to be
1832 incorrectly computed when quota_filecount was set in an appendfile
1835 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1836 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1839 3. Give more explanation in the error message when the command for a transport
1840 filter fails to execute.
1842 4. There are several places where Exim runs a non-Exim command in a
1843 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1844 was being done only for the command run by the queryprogram router. It is
1845 now done for all such subprocesses. The other cases are: ${run, transport
1846 filters, and the commands run by the lmtp and pipe transports.
1848 5. Some older OS have a limit of 256 on the maximum number of file
1849 descriptors. Exim was using setrlimit() to set 1000 as a large value
1850 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1851 systems. I've change it so that if it can't get 1000, it tries for 256.
1853 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1854 was an oversight, and furthermore, ever since the addition of extra
1855 controls (e.g. 4.43/32), the checks on when to allow different forms of
1856 "control" were broken. There should now be diagnostics for all cases when a
1857 control that does not make sense is encountered.
1859 7. $recipients is now available in the predata ACL (oversight).
1861 8. Tidy the search cache before the fork to do a delivery from a message
1862 received from the command line. Otherwise the child will trigger a lookup
1863 failure and thereby defer the delivery if it tries to use (for example) a
1864 cached ldap connection that the parent has called unbind on.
1866 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1867 of $address_data from the recipient verification was clobbered by the
1868 sender verification.
1870 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1871 was its contents. (It was OK if the option was not defined at all.)
1873 11. A "Completed" log line is now written for messages that are removed from
1874 the spool by the -Mrm option.
1876 12. $host_address is now set to the target address during the checking of
1877 ignore_target_hosts.
1879 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1880 being passed; this would have caused $sender_host_name to have been used if
1881 matching the list had actually called for a host name (not very likely,
1882 since this list is usually IP addresses). A host name is now passed as
1885 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1886 code that specifies a non-restarting handler (typically sigaction() in
1887 modern systems) in an attempt to fix a rare and obscure crash bug.
1889 15. Narrowed the window for a race in the daemon that could cause it to ignore
1890 SIGCHLD signals. This is not a major problem, because they are used only to
1891 wake it up if nothing else does.
1893 16. A malformed maildirsize file could cause Exim to calculate negative values
1894 for the mailbox size or file count. Odd effects could occur as a result.
1895 The maildirsize information is now recalculated if the size or filecount
1898 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1899 support for a long time. Removed HAVE_SYS_VFS_H.
1901 18. Updated exipick to current release from John Jetmore.
1903 19. Allow an empty sender to be matched against a lookup in an address list.
1904 Previously the only cases considered were a regular expression, or an
1907 20. Exim went into a mad DNS lookup loop when doing a callout where the
1908 host was specified on the transport, if the DNS lookup yielded more than
1911 21. The RFC2047 encoding function was originally intended for short strings
1912 such as real names; it was not keeping to the 75-character limit for
1913 encoded words that the RFC imposes. It now respects the limit, and
1914 generates multiple encoded words if necessary. To be on the safe side, I
1915 have increased the buffer size for the ${rfc2047: expansion operator from
1918 22. Failure to deliver a bounce message always caused it to be frozen, even if
1919 there was an errors_to setting on the router. The errors_to setting is now
1922 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1923 canonical form (fully expanded) before being placed in
1924 $sender_host_address.
1926 24. Updated eximstats to version 1.33
1928 25. Include certificate and key file names in error message when GnuTLS fails
1929 to set them up, because the GnuTLS error message doesn't include the name
1930 of the failing file when there is a problem reading it.
1932 26. Expand error message when OpenSSL has problems setting up cert/key files.
1935 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1936 (this can affect the format of dates).
1938 28. exim_tidydb, when checking for the continued existence of a message for
1939 which it has found a message-specific retry record, was not finding
1940 messages that were in split spool directories. Consequently, it was
1941 deleting retry records that should have stayed in existence.
1943 29. eximstats updated to version 1.35
1944 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1945 1.35 - bugfix such that pie charts by volume are generated correctly
1947 30. The SPA authentication driver was not abandoning authentication and moving
1948 on to the next authenticator when an expansion was forced to fail,
1949 contradicting the general specification for all authenticators. Instead it
1950 was generating a temporary error. It now behaves as specified.
1952 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1953 (the order specifies the preference for clients). The order is now AES256,
1954 AES128, 3DES, ARCFOUR128.
1956 31. Small patch to Sieve code - explicitly set From: when generating an
1959 32. Exim crashed if a remote delivery caused a very long error message to be
1960 recorded - for instance if somebody sent an entire SpamAssassin report back
1961 as a large number of 550 error lines. This bug was coincidentally fixed by
1962 increasing the size of one of Exim's internal buffers (big_buffer) that
1963 happened as part of the Exiscan merge. However, to be on the safe side, I
1964 have made the code more robust (and fixed the comments that describe what
1967 33. Some experimental protocols are using DNS PTR records for new purposes. The
1968 keys for these records are domain names, not reversed IP addresses. The
1969 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1970 leaves it alone. Component reversal etc. now happens only for IP addresses.
1973 34. The host_aton() function is supposed to be passed a string that is known
1974 to be a valid IP address. However, in the case of IPv6 addresses, it was
1975 not checking this. This is a hostage to fortune. Exim now panics and dies
1976 if the condition is not met. A case was found where this could be provoked
1977 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1978 components; fortuitously, this particular loophole had already been fixed
1979 by change 4.50/55 or 4.44/33 above.
1981 If there are any other similar loopholes, the new check in host_aton()
1982 itself should stop them being exploited. The report I received stated that
1983 data on the command line could provoke the exploit when Exim was running as
1984 exim, but did not say which command line option was involved. All I could
1985 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1986 running as the user.
1989 35. There was a buffer overflow vulnerability in the SPA authentication code
1990 (which came originally from the Samba project). I have added a test to the
1991 spa_base64_to_bits() function which I hope fixes it.
1994 36. The daemon start-up calls getloadavg() while still root for those OS that
1995 need the first call to be done as root, but it missed one case: when
1996 deliver_queue_load_max is set with deliver_drop_privilege. This is
1997 necessary for the benefit of the queue runner, because there is no re-exec
1998 when deliver_drop_privilege is set.
2000 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2001 was in use, was not putting the data itself into the right store pool;
2002 consequently, it could be overwritten for a subsequent message in the same
2003 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2006 38. Sometimes the final signoff response after QUIT could fail to get
2007 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2008 < 0 before doing a fflush(). This bug looks as though it goes back to the
2009 introduction of TLS in release 3.20, but "sometimes" must have been rare
2010 because the tests only now provoked it.
2016 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2017 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2018 Now it returns an integer. A number of calls were still expecting a T/F
2019 return. Fortuitously, in all cases, the tests worked in OK situations,
2020 which is the norm. However, things would have gone wrong on any write
2021 failures on the smtp file descriptor. This function is used when sending
2022 messages over SMTP and also when doing verify callouts.
2024 2. When Exim is called to do synchronous delivery of a locally submitted
2025 message (the -odf or -odi options), it no longer closes stderr before doing
2028 3. Implemented the mua_wrapper option.
2030 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2032 5. Implemented the functions header_remove(), header_testname(),
2033 header_add_at_position(), and receive_remove_recipient(), and exported them
2036 6. If an ACL "warn" statement specified the addition of headers, Exim already
2037 inserted X-ACL-Warn: at the start if there was no header name. However, it
2038 was not making this test for the second and subsequent header lines if
2039 there were newlines in the string. This meant that an invalid header could
2040 be inserted if Exim was badly configured.
2042 7. Allow an ACL "warn" statement to add header lines at the start or after all
2043 the Received: headers, as well as at the end.
2045 8. Added the rcpt_4xx retry error code.
2047 9. Added postmaster_mailfrom=xxx to callout verification option.
2049 10. Added mailfrom=xxxx to the callout verification option, for verify=
2052 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2053 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2055 12. Inserted some casts to stop certain compilers warning when using pointer
2056 differences as field lengths or precisions in printf-type calls (mostly
2057 affecting debugging statements).
2059 13. Added optional readline() support for -be (dynamically loaded).
2061 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2062 same clock tick as a message's arrival, so that its received time was the
2063 same as the "first fail" time on the retry record, and that message
2064 remained on the queue past the ultimate address timeout, every queue runner
2065 would try a delivery (because it was past the ultimate address timeout) but
2066 after another failure, the ultimate address timeout, which should have then
2067 bounced the address, did not kick in. This was a "< instead of <=" error;
2068 in most cases the first failure would have been in the next clock tick
2069 after the received time, and all would be well.
2071 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2072 being recognized when the domain list was tested by the match_domain
2073 condition in an expansion string.
2075 16. Added the ${str2b64: operator.
2077 17. Exim was always calling setrlimit() to set a large limit for the number of
2078 processes, without checking whether the existing limit was already
2079 adequate. (It did check for the limit on file descriptors.) Furthermore,
2080 errors from getrlimit() and setrlimit() were being ignored. Now they are
2081 logged to the main and panic logs, but Exim does carry on, to try to do its
2082 job under whatever limits there are.
2084 18. Imported PCRE 5.0.
2086 19. Trivial typo in log message " temporarily refused connection" (the leading
2089 20. If the log selector return_path_on_delivery was set and an address was
2090 redirected to /dev/null, the delivery process crashed because it assumed
2091 that a return path would always be set for a "successful" delivery. In this
2092 case, the whole delivery is bypassed as an optimization, and therefore no
2095 21. Internal re-arrangement: the function for sending a challenge and reading
2096 a response while authentication was assuming a zero-terminated challenge
2097 string. It's now changed to take a pointer and a length, to allow for
2098 binary data in such strings.
2100 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2102 23. Exim was not respecting finduser_retries when seeking the login of the
2103 uid under which it was called; it was always trying 10 times. (The default
2104 setting of finduser_retries is zero.) Also, it was sleeping after the final
2105 failure, which is pointless.
2107 24. Implemented tls_on_connect_ports.
2109 25. Implemented acl_smtp_predata.
2111 26. If the domain in control=submission is set empty, Exim assumes that the
2112 authenticated id is a complete email address when it generates From: or
2113 Sender: header lines.
2115 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2116 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2117 chown and chgrp in /bin and hostname in /usr/bin.
2119 28. Exim was keeping the "process log" file open after each use, just as it
2120 does for the main log. This opens the possibility of it remaining open for
2121 long periods when the USR1 signal hits a daemon. Occasional processlog
2122 errors were reported, that could have been caused by this. Anyway, it seems
2123 much more sensible not to leave this file open at all, so that is what now
2126 29. The long-running daemon process does not normally write to the log once it
2127 has entered its main loop, and it closes the log before doing so. This is
2128 so that log files can straightforwardly be renamed and moved. However,
2129 there are a couple of unusual error situations where the daemon does write
2130 log entries, and I had neglected to close the log afterwards.
2132 30. The text of an SMTP error response that was received during a remote
2133 delivery was being truncated at 512 bytes. This is too short for some of
2134 the long messages that one sometimes sees. I've increased the limit to
2137 31. It is now possible to make retry rules that apply only when a message has a
2138 specific sender, in particular, an empty sender.
2140 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2141 it possible to be selective about when SMTP synchronization is enforced.
2143 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2145 32. Implemented hosts_connection_nolog.
2147 33. Added an ACL for QUIT.
2149 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2152 35. Added mailbox_size and mailbox_filecount to appendfile.
2154 36. Added control = no_multiline_responses to ACLs.
2156 37. There was a bug in the logic of the code that waits for the clock to tick
2157 in the case where the clock went backwards by a substantial amount such
2158 that the microsecond fraction of "now" was more than the microsecond
2159 fraction of "then" (but the whole seconds number was less).
2161 38. Added support for the libradius Radius client library this is found on
2162 FreeBSD (previously only the radiusclient library was supported).
2168 1. When certain lookups returned multiple values in the form name=value, the
2169 quoting of the values was not always being done properly. Specifically:
2170 (a) If the value started with a double quote, but contained no whitespace,
2172 (b) If the value contained whitespace other than a space character (i.e.
2173 tabs or newlines or carriage returns) it was not quoted.
2174 This fix has been applied to the mysql and pgsql lookups by writing a
2175 separate quoting function and calling it from the lookup code. The fix
2176 should probably also be applied to nisplus, ibase and oracle lookups, but
2177 since I cannot test any of those, I have not disturbed their existing code.
2179 2. A hit in the callout cache for a specific address caused a log line with no
2180 reason for rejecting RCPT. Now it says "Previous (cached) callout
2181 verification failure".
2183 3. There was an off-by-one bug in the queryprogram router. An over-long
2184 return line was truncated at 256 instead of 255 characters, thereby
2185 overflowing its buffer with the terminating zero. As well as fixing this, I
2186 have increased the buffer size to 1024 (and made a note to document this).
2188 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2189 when Exim is waiting for an SMTP response from a remote server, Exim
2190 restarts its select() call on the socket, thereby resetting its timeout.
2191 This is not a problem when such interrupts are rare. Somebody set up a cron
2192 job to run exiwhat every 2 minutes, which is less than the normal select()
2193 timeout (5 or 10 minutes). This meant that the select() timeout never
2194 kicked in because it was always reset. I have fixed this by comparing the
2195 time when an interrupt arrives with the time at the start of the first call
2196 to select(). If more time than the timeout has elapsed, the interrupt is
2197 treated as a timeout.
2199 5. Some internal re-factoring in preparation for the addition of Sieve
2200 extensions (by MH). In particular, the "personal" test is moved to a
2201 separate function, and given an option for scanning Cc: and Bcc: (which is
2202 not set for Exim filters).
2204 6. When Exim created an email address using the login of the caller as the
2205 local part (e.g. when creating a From: or Sender: header line), it was not
2206 quoting the local part when it contained special characters such as @.
2208 7. Installed new OpenBSD configuration files.
2210 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2211 try to make them clearer.
2213 9. Callout options, other than the timeout value, were being ignored when
2214 verifying sender addresses in header lines. For example, when using
2216 verify = header_sender/callout=no_cache
2218 the cache was (incorrectly) being used.
2220 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2221 only the Cygwin environment.
2223 11. When return_path_on_delivery was set as a log selector, if different remote
2224 addresses in the same message used different return paths and parallel
2225 remote delivery occurred, the wrong values would sometimes be logged.
2226 (Whenever a remote delivery process finished, the return path value from
2227 the most recently started remote delivery process was logged.)
2229 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2230 lines when AUTH and/or TLS are in use. This is the "received protocol"
2231 field. Exim used to use "asmtp" for authenticated SMTP, without any
2232 indication (in the protocol name) for TLS use. Now it follows the RFC and
2233 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2234 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2235 names appear in log lines as well as in Received: header lines.
2237 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2238 extensions, and comparison tests, and to fix some bugs.
2240 14. Changes to the "personal" filter test:
2242 (1) The test was buggy in that it was just doing the equivalent of
2243 "contains" tests on header lines. For example, if a user's address was
2244 anne@some.where, the "personal" test would incorrectly be true for
2246 To: susanne@some.where
2248 This test is now done by extracting each address from the header in turn,
2249 and checking the entire address. Other tests that are part of "personal"
2250 are now done using regular expressions (for example, to check local parts
2251 of addresses in From: header lines).
2253 (2) The list of non-personal local parts in From: addresses has been
2254 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2255 taken from the Sieve specification recommendations.
2257 (3) If the message contains any header line starting with "List-" it is
2258 treated as non-personal.
2260 (4) The test for "circular" in the Subject: header line has been removed
2261 because it now seems ill-conceived.
2263 15. Minor typos in src/EDITME comments corrected.
2265 16. Installed latest exipick from John Jetmore.
2267 17. If headers_add on a router specified a text string that was too long for
2268 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2269 of string_sprintf() is now avoided.
2271 18. $message_body_size was not set (it was always zero) when running the DATA
2272 ACL and the local_scan() function.
2274 19. For the "mail" command in an Exim filter, no default was being set for
2275 the once_repeat time, causing a random time value to be used if "once" was
2276 specified. (If the value happened to be <= 0, no repeat happened.) The
2277 default is now 0s, meaning "never repeat". The "vacation" command was OK
2278 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2279 (I found it when inspecting the code).
2281 20. There is now an overall timeout for performing a callout verification. It
2282 defaults to 4 times the callout timeout, which applies to individual SMTP
2283 commands during the callout. The overall timeout applies when there is more
2284 than one host that can be tried. The timeout is checked before trying the
2285 next host. This prevents very long delays if there are a large number of
2286 hosts and all are timing out (e.g. when the network connections are timing
2287 out). The value of the overall timeout can be changed by specifying an
2288 additional sub-option for "callout", called "maxwait". For example:
2290 verify = sender/callout=5s,maxwait=20s
2292 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2293 to the end before writing, but this should make it even safer).
2295 22. Exim was forgetting that it had advertised PIPELINING for the second and
2296 subsequent messages on an SMTP connection. It was also not resetting its
2297 memory on STARTTLS and an internal HELO.
2299 23. When Exim logs an SMTP synchronization error within a session, it now
2300 records whether PIPELINING has been advertised or not.
2302 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2303 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2304 rather than long int.
2306 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2308 26. Added the never_mail option to autoreply.
2314 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2315 crash if the getsockname() call failed; this can happen if a connection is
2316 closed very soon after it is established. The problem was simply in the
2317 order in which certain operations were done, causing Exim to try to write
2318 to the SMTP stream before it had set up the file descriptor. The bug has
2319 been fixed by making things happen in the correct order.
2325 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2326 before the connection was closed, thus losing the rejection response.
2328 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2329 some early Solaris releases, but causes trouble in current releases where
2330 socklen_t is defined.
2332 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2335 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2338 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2339 that releasing the top of it at the end releases what was used for sub-
2340 expansions (unless the block got too big). However, discard this block if
2341 the first thing is a variable or header, so that we can use its block when
2342 it is dynamic (useful for very large $message_headers, for example).
2344 6. Lookups now cache *every* query, not just the most recent. A new, separate
2345 store pool is used for this. It can be recovered when all lookup caches are
2346 flushed. Lookups now release memory at the end of their result strings.
2347 This has involved some general refactoring of the lookup sources.
2349 7. Some code has been added to the store_xxx() functions to reduce the amount
2350 of flapping under certain conditions.
2352 8. log_incoming_interface used to affect only the <= reception log lines. Now
2353 it causes the local interface and port to be added to several more SMTP log
2354 lines, for example "SMTP connection from", and rejection lines.
2356 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2358 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2360 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2361 could be overwritten at the end of the current message (or the start of a
2362 new message if it was set in a HELO ACL). The value is now preserved for
2363 the duration of the SMTP connection.
2365 12. If a transport had a headers_rewrite setting, and a matching header line
2366 contained an unqualified address, that address was qualified, even if it
2367 did not match any rewriting rules. The underlying bug was that the values
2368 of the flags that permit the existence of unqualified sender and recipient
2369 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2370 non-local messages, and by -bnq for local messages) were not being
2371 preserved with the message after it was received.
2373 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2374 "next input=" as part of the text comprising the host identity instead of
2375 the correct text. The code was using the same buffer for two different
2376 strings. However, depending on which order the printing function evaluated
2377 its arguments, the bug did not always show up. Under Linux, for example, my
2378 test suite worked just fine.
2380 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2381 doesn't work with some older versions of Perl. It has been changed to "my",
2382 which in any case is probably the better facility to use.
2384 15. A really picky compiler found some instances of statements for creating
2385 error messages that either had too many or two few arguments for the format
2388 16. The size of the buffer for calls to the DNS resolver has been increased
2389 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2390 for addresses that have a lot of PTR records. This alleviates a problem; it
2391 does not fully solve it.
2393 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2394 buffer now truncates the list and logs the incident, which is the same
2395 action as happens when Exim is looking up a host name and its aliases.
2396 Previously in this situation something unpredictable would happen;
2397 sometimes it was "internal error: store_reset failed".
2399 18. If a server dropped the connection unexpectedly when an Exim client was
2400 using GnuTLS and trying to read a response, the client delivery process
2401 crashed while trying to generate an error log message.
2403 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2404 string, for example:
2406 warn message = H1: something\nH2: something
2408 the text was added as a single header line from Exim's point of view
2409 though it ended up OK in the delivered message. However, searching for the
2410 second and subsequent header lines using $h_h2: did not work. This has been
2411 fixed. Similarly, if a system filter added multiple headers in this way,
2412 the routers could not see them.
2414 20. Expanded the error message when iplsearch is called with an invalid key to
2415 suggest using net-iplsearch in a host list.
2417 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2418 ACLs are no longer actually imposed (and a message to that effect is
2421 22. If a "gecos" field in a passwd entry contained escaped characters, in
2422 particular, if it contained a \" sequence, Exim got it wrong when building
2423 a From: or a Sender: header from that name. A second bug also caused
2424 incorrect handling when an unquoted " was present following a character
2425 that needed quoting.
2427 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2428 was not being matched caselessly.
2430 24. Arranged for all hyphens in the exim.8 source to be escaped with
2433 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2434 properly. Recipient callout cache records were still being keyed to include
2435 the sender, even when use_sender was set false. This led to far more
2436 callouts that were necessary. The sender is no longer included in the key
2437 when use_sender is false.
2439 26. Added "control = submission" modifier to ACLs.
2441 27. Added the ${base62d: operator to decode base 62 numbers.
2443 28. dnsdb lookups can now access SRV records.
2445 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2446 the configuration file.
2448 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2449 (-v) mode. This makes the output for a verbose queue run more intelligible.
2451 31. Added a use_postmaster feature to recipient callouts.
2453 32. Added the $body_zerocount variable, containing the number of binary zero
2454 bytes in the message body.
2456 33. The time of last modification of the "new" subdirectory is now used as the
2457 "mailbox time last read" when there is a quota error for a maildir
2460 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2462 35. Added +ignore_unknown as a special item in host lists.
2464 36. Code for decoding IPv6 addresses in host lists is now included, even if
2465 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2466 address was recognized as an IP address, but was then not correctly decoded
2467 into binary, causing unexpected and incorrect effects when compared with
2474 1. Very minor rewording of debugging text in manualroute to say "list of
2475 hosts" instead of "hostlist".
2477 2. If verify=header_syntax was set, and a header line with an unqualified
2478 address (no domain) and a large number of spaces between the end of the
2479 name and the colon was received, the reception process suffered a buffer
2480 overflow, and (when I tested it) crashed. This was caused by some obsolete
2481 code that should have been removed. The fix is to remove it!
2483 3. When running in the test harness, delay a bit after writing a bounce
2484 message to get a bit more predictability in the log output.
2486 4. Added a call to search_tidyup() just before forking a reception process. In
2487 theory, someone could use a lookup in the expansion of smtp_accept_max_
2488 per_host which, without the tidyup, could leave open a database connection.
2490 5. Added the variables $recipient_data and $sender_data which get set from a
2491 lookup success in an ACL "recipients" or "senders" condition, or a router
2492 "senders" option, similar to $domain_data and $local_part_data.
2494 6. Moved the writing of debug_print from before to after the "senders" test
2497 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2498 problems for message scanning, either using a data ACL, or using
2499 local_scan() because the Received: header was not generated till after they
2500 were called (in order to set the time as the time of reception completion).
2501 I have revised the way this works. The header is now generated after the
2502 body is received, but before the ACL or local_scan() are called. After they
2503 are run, the timestamp in the header is updated.
2509 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2510 before starting a queue runner without re-exec. This happened only when
2511 deliver_drop_privilege was set or when the Exim user was set to root. The
2512 effect of the bug was that timeouts during subsequent deliveries caused
2513 crashes instead of being properly handled. The handler is now left at its
2514 default (and expected) setting.
2516 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2517 message, again when deliver_drop_privilege is set or Exim is run as root.
2518 The bug described in (1) was not present in this case, but the tidying up
2519 of the other signals was missing. I have made the two cases consistent.
2521 3. The ignore_target_hosts setting on a manualroute router was being ignored
2522 for hosts that were looked up using the /MX notation.
2524 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2527 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2528 operated on the sender address. After changing the $sender_address to <>
2529 for the sender address verify, Exim was re-instated it as the original
2530 (before rewriting) address, but remembering that it had rewritten it, so it
2531 wasn't rewriting it again. This bug also had the effect of breaking the
2532 sender address verification caching when the sender address was rewritten.
2534 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2535 This has been changed so that if the ip literal address matches
2536 ignore_target_hosts, the router declines.
2538 7. Added expansion conditions match_domain, match_address, and match_local_
2539 part (NOT match_host).
2541 8. The placeholder for the Received: header didn't have a length field set.
2543 9. Added code to Exim itself and to exim_lock to test for a specific race
2544 condition that could lead to file corruption when using MBX delivery. The
2545 issue is with the lockfile that is created in /tmp. If this file is removed
2546 after a process has opened it but before that process has acquired a lock,
2547 there is the potential for a second process to recreate the file and also
2548 acquire a lock. This could lead to two Exim processes writing to the file
2549 at the same time. The added code performs the same test as UW imapd; it
2550 checks after acquiring the lock that its file descriptor still refers to
2551 the same named file.
2553 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2554 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2557 11. Added the smtp_active_hostname option. If used, this will typically be made
2558 to depend on the incoming interface address. Because $interface_address is
2559 not set up until the daemon has forked a reception process, error responses
2560 that can happen earlier (such as "too many connections") no longer contain
2563 12. If an expansion in a condition on a "warn" statement fails because a lookup
2564 defers, the "warn" statement is abandoned, and the next ACL statement is
2565 processed. Previously this caused the whole ACL to be aborted.
2567 13. Added the iplsearch lookup type.
2569 14. Added ident_timeout as a log selector.
2571 15. Added tls_certificate_verified as a log selector.
2573 16. Added a global option tls_require_ciphers (compare the smtp transport
2574 option of the same name). This controls incoming TLS connections.
2576 17. I finally figured out how to make tls_require_ciphers do a similar thing
2577 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2578 before starting the TLS session.
2580 18. Tabs are now shown as \t in -bP output.
2582 19. If the log selector return_path_on_delivery was set, Exim crashed when
2583 bouncing a message because it had too many Received: header lines.
2585 20. If two routers both had headers_remove settings, and the first one included
2586 a superfluous trailing colon, the final name in the first list and the
2587 first name in the second list were incorrectly joined into one item (with a
2588 colon in the middle).
2594 1. Added -C and -D options to the exinext utility, mainly to make it easier
2595 to include in the automated testing, but these could be helpful when
2596 multiple configurations are in use.
2598 2. The exinext utility was not formatting the output nicely when there was
2599 an alternate port involved in the retry record key, nor when there was a
2600 message id as well (for retries that were specific to a specific message
2601 and a specific host). It was also confused by IPv6 addresses, because of
2602 the additional colons they contain. I have fixed the IPv4 problem, and
2603 patched it up to do a reasonable job for IPv6.
2605 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2606 delivery, the log line now contains "pipelined" if PIPELINING was used.
2608 4. An SMTP transport process used to panic and die if the bind() call to set
2609 an explicit outgoing interface failed. This has been changed; it is now
2610 treated in the same way as a connect() failure.
2612 5. A reference to $sender_host_name in the part of a conditional expansion
2613 that was being skipped was still causing a DNS lookup. This no longer
2616 6. The def: expansion condition was not recognizing references to header lines
2617 that used bh_ and bheader_.
2619 7. Added the _cache feature to named lists.
2621 8. The code for checking quota_filecount in the appendfile transport was
2622 allowing one more file than it should have been.
2624 9. For compatibility with Sendmail, the command line option
2632 and sets the incoming protocol and host name (for trusted callers). The
2633 host name and its colon can be omitted when only the protocol is to be set.
2634 Note the Exim already has two private options, -pd and -ps, that refer to
2635 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2636 "s", but I don't think that's a major issue.
2638 10. A number of refactoring changes to the code, none of which should affect
2641 (a) The number of logging options was getting close to filling up the
2642 32-bit word that was used as a bit map. I have split them into two classes:
2643 those that are passed in the argument to log_write(), and those that are
2644 only ever tested independently outside of that function. These are now in
2645 separate 32-bit words, so there is plenty of room for expansion again.
2646 There is no change in the user interface or the logging behaviour.
2648 (b) When building, for example, log lines, the code previously used a
2649 macro that called string_cat() twice, in order to add two strings. This is
2650 not really sufficiently general. Furthermore, there was one instance where
2651 it was actually wrong because one of the argument was used twice, and in
2652 one call a function was used. (As it happened, calling the function twice
2653 did not affect the overall behaviour.) The macro has been replaced by a
2654 function that can join an arbitrary number of extra strings onto a growing
2657 (c) The code for expansion conditions now uses a table and a binary chop
2658 instead of a serial search (which was left over from when there were very
2659 few conditions). Also, it now recognizes conditions like "pam" even when
2660 the relevant support is not compiled in: a suitably worded error message is
2661 given if an attempt is made to use such a condition.
2663 11. Added ${time_interval:xxxxx}.
2665 12. A bug was causing one of the ddress fields not to be passed back correctly
2666 from remote delivery subprocesses. The field in question was not being
2667 subsequently used, so this caused to problems in practice.
2669 13. Added new log selectors queue_time and deliver_time.
2671 14. Might have fixed a bug in maildirsizefile handling that threw up
2672 "unexpected character" debug warnings, and recalculated the data
2673 unnecessarily. In any case, I expanded the warning message to give more
2676 15. Added the message "Restricted characters in address" to the statements in
2677 the default ACL that block characters like @ and % in local parts.
2679 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2680 Three changes have been made:
2682 (a) There was a serious bug; a negative response to MAIL caused the whole
2683 recipient domain to be cached as invalid, thereby blocking all messages
2684 to all local parts at the same domain, from all senders. This bug has
2685 been fixed. The domain is no longer cached after a negative response to
2686 MAIL if the sender used is not empty.
2688 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2691 (c) A new callout option, "use_sender" has been added for people who want
2692 the modified behaviour.
2698 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2701 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2702 indeed breaks things for older releases.
2704 3. Added additional logging to the case where there is a problem reading data
2705 from a filter that is running in a subprocess using a pipe, in order to
2706 try to track down a specific problem.
2708 4. Testing facility fudge: when running in the test harness and attempting
2709 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2710 getting "No route to host". Convert this to a timeout.
2712 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2715 6. Some OS don't have socklen_t but use size_t instead. This affects the
2716 fifth argument of getsockopt() amongst other things. This is now
2717 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2718 can be set for individual OS. I have set it for SunOS5, OSF1, and
2719 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2720 some earlier ones do not.
2722 7. Change 4.30/15 was not doing the test caselessly.
2724 8. The standard form for an IPv6 address literal was being rejected by address
2725 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2726 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2727 this, as well as the form without the "IPv6" on the front (but only when
2728 address literals are enabled, of course).
2730 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2732 10. Exim crashed if a message with an empty sender address specified by -f
2733 encountered a router with an errors_to setting. This could be provoked only
2734 by a command such as
2738 where an empty string was supplied; "<>" did not hit this bug.
2740 11. Installed PCRE release 4.5.
2742 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2743 remained set. It is now erased.
2745 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2746 times from message ids (which are base 36 rather than the normal 62).
2748 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2749 were being counted as actual protocol errors, and logged if the log
2750 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2751 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2752 and DATA following a set of rejected RCPTs do not count as protocol errors.
2753 In other words, Exim assumes they were pipelined, though this may not
2754 actually be the case. Of course, in all cases the client gets an
2755 appropriate error code.
2757 15. If a lookup fails in an ACL condition, a message about the failure may
2758 be available; it is used if testing the ACL cannot continue, because most
2759 such messages specify what the cause of the deferral is. However, some
2760 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2761 that caused an old message to be retained and used if a later statement
2762 caused a defer, replacing the real cause of the deferral.
2764 16. If an IP address had so many PTR records that the DNS lookup buffer
2765 was not large enough to hold them, Exim could crash while trying to process
2766 the truncated data. It now detects and logs this case.
2768 17. Further to 4.21/58, another change has been made: if (and only if) the
2769 first line of a message (the first header line) ends with CRLF, a bare LF
2770 in a subsequent header line has a space inserted after it, so as not to
2771 terminate the header.
2773 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2774 unnecessarily, used atoi() instead of strtol(), and didn't check the
2775 termination when getting file sizes from file names by regex.
2777 19. Completely re-implemented the support for maildirsize files, in the light
2778 of a number of problems with the previous contributed implementation
2779 (4.30/29). In particular:
2781 . If the quota is zero, the maildirsize file is maintained, but no quota is
2784 . If the maildir directory does not exist, it is created before any attempt
2785 to write a maildirsize file.
2787 . The quota value in the file is just a cache; if the quota is changed in
2788 the transport, the new value overrides.
2790 . A regular expression is available for excluding directories from the
2793 20. The autoreply transport checks the characters in options that define the
2794 message's headers; it allows continued headers, but it was checking with
2795 isspace() after an embedded newline instead of explicitly looking for a
2798 21. If all the "regular" hosts to which an address was routed had passed their
2799 expiry times, and had not reached their retry times, the address was
2800 bounced, even if fallback hosts were defined. Now Exim should go on to try
2803 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2804 equivalent code in the SMTP transport. Some hosts send humungous responses
2805 to HELO/EHLO, more than 1024 it seems.
2807 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2808 gives compiler warnings in some environments. I've now done it "properly",
2811 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2812 (because of problems with the built-in one) was declared to return uschar *
2813 instead of char *, causing compiler failure.
2815 25. Fixed a file descriptor leak when processing alias/forward files.
2817 26. Fixed a minor format string issue in dbfn.c.
2819 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2821 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2822 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2825 29. When a sender address is verified, it is cached, to save repeating the test
2826 when there is more than one recipient in a message. However, when the
2827 verification involves a callout, it is possible for different callout
2828 options to be set for different recipients. It is too complicated to keep
2829 track of this in the cache, so now Exim always runs a verification when a
2830 callout is required, relying on the callout cache for the optimization.
2831 The overhead is duplication of the address routing, but this should not be
2834 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2835 to be verified with callout=postmaster, and the main callout worked but the
2836 postmaster check failed, the verification correctly failed. However, if a
2837 subsequent RCPT command asked for sender verification *without* the
2838 postmaster check, incorrect caching caused this verification also to fail,
2841 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2842 it was not caching the DNS options (qualify_single, search_parents) that
2843 were used when the lookup failed. A subsequent lookup with different
2844 options therefore always gave the same answer, though there were cases
2845 where it should not have. (Example: a "domains = !$mx_any" option on a
2846 dnslookup router: the "domains" option is always processed without any
2847 widening, but the router might have qualify_single set.) Now Exim uses the
2848 cached value only when the same options are set.
2850 32. Added John Jetmore's "exipick" utility to the distribution.
2852 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2853 than a timeout (e.g. a certificate is required, and is not provided), an
2854 Exim server now closes the connection immediately. Previously it waited for
2855 the client to close - but if the client is SSL, it seems that they each
2856 wait for each other, leading to a delay before one of them times out.
2858 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2859 maintained 0.8.x compatibility because I don't think many are using it, and
2860 it is clearly obsolete.
2862 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2865 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2866 client certificate was expired. A simple patch fixes this, though I don't
2867 understand the full logic of why the verify callback is called multiple
2870 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2871 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2872 which causes problems with some clients (such as the Certicom SSL Plus
2873 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2874 disables the coutermeasure allowing Eudora to connect."
2876 38. Exim was not checking that a write() to a log file succeeded. This could
2877 lead to Bad Things if a log got too big, in particular if it hit a file
2878 size limit. Exim now panics and dies if it cannot write to a log file, just
2879 as it does if it cannot open a log file.
2881 39. Modified OS/Makefile-Linux so that it now contains
2883 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2885 The two -D definitions ensure that Exim is compiled with large file
2886 support, which makes it possible to handle log files that are bigger than
2889 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2890 instance) a domain was checked against a named list that involved a lookup,
2891 causing $domain_data to be set, then another domain was checked against the
2892 same list, then the first domain was re-checked, the value of $domain_data
2893 after the final check could be wrong. In particular, if the second check
2894 failed, it could be set empty. This bug probably also applied to
2897 41. The strip_trailing_dot option was not being applied to the address given
2898 with the -f command-line option.
2900 42. The code for reading a message's header from the spool was incrementing
2901 $received_count, but never initializing it. This meant that the value was
2902 incorrect (doubled) while delivering a message in the same process in which
2903 it was received. In the most common configuration of Exim, this never
2904 happens - a fresh exec is done - but it can happen when
2905 deliver_drop_privilege is set.
2907 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2908 it now includes up to 150 characters of the unexpected data in the log
2911 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2912 and building data strings. The size of both of these buffers was 10 000
2913 bytes - far larger than anybody would *ever* want, thought I. Needless to
2914 say, somebody hit the limit. I have increased the maximum line length to
2915 20 000 and the maximum data length of concatenated lines to 100 000. I have
2916 also fixed two bugs, because there was no checking on these buffers. Tsk,
2917 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2918 buffer is too small.
2920 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2921 lsearch lookups. Now it does.
2923 46. When parsing a route_list item in a manualroute router, a fixed-length
2924 buffer was used for the list of hosts. I made this 1024 bytes long,
2925 thinking that nobody would ever have a list of hosts that long. Wrong.
2926 Somebody had a whole pile of complicated expansion conditions, and the
2927 string was silently truncated, leading to an expansion error. It turns out
2928 that it is easier to change to an unlimited length (owing to other changes
2929 that have happened since this code was originally written) than to build
2930 structure for giving a limitation error. The length of the item that
2931 expands into the list of hosts is now unlimited.
2933 47. The lsearch lookup could not handle data where the length of text line was
2934 more than 4095 characters. Such lines were truncated, leading to shortened
2935 data being returned. It should now handle lines of any length.
2937 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2938 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2941 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2942 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2943 no longer happens in comment lines. A list of replacements is now placed
2944 at the head of all of the source files, except those whose only change is
2945 to replace PERL_COMMAND in the very first #! line.
2947 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2948 messages on the queue, with a bottom-up merge sort, using code contributed
2949 by Michael Haardt. This should make operations like -bp somewhat faster on
2950 large queues. It won't affect queue runners, except when queue_run_in_order
2953 51. Installed eximstats 1.31 in the distribution.
2955 52. Added support for SRV lookups to the dnslookup router.
2957 53. If an ACL referred to $message_body or $message_body_end, the value was not
2958 reset for any messages that followed in the same SMTP session.
2960 54. The store-handling optimization for building very long strings was not
2961 differentiating between the different store pools. I don't think this
2962 actually made any difference in practice, but I've tidied it.
2964 55. While running the routers to verify a sender address, $sender_address
2965 was still set to the sender address. This is wrong, because when routing to
2966 send a bounce to the sender, it would be empty. Therefore, I have changed
2967 it so that, while verifying a sender address, $sender_address is set to <>.
2968 (There is no change to what happens when verifying a recipient address.)
2970 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2971 target A or AAAA records (if not already returned) without resetting the
2972 qualify_single or search_parents options of the DNS resolver. These are
2973 inappropriate in this case because the targets of MX and SRV records must
2974 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2975 target that, when qualified, matched something in the local domain. These
2976 two options are now turned off when doing these lookups.
2978 57. It seems that at least some releases of Reiserfs (which does not have the
2979 concept of a fixed number of inodes) returns zero and not -1 for the
2980 number of available inodes. This interacted badly with check_spool_inodes,
2981 which assumed that -1 was the "no such thing" setting. What I have done is
2982 to check that the total number of inodes is greater than zero before doing
2983 the test of how many are available.
2985 58. When a "warn" ACL statement has a log_message modifier, the message is
2986 remembered, and not repeated. This is to avoid a lot of repetition when a
2987 message has many recipients that cause the same warning to be written.
2988 Howewer, Exim was preserving the list of already written lines for an
2989 entire SMTP session, which doesn't seem right. The memory is now reset if a
2990 new message is started.
2992 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2993 debugging output unless log_rewrite was also set.
2995 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2996 of (int)(handle) when we know that handle contains (void *)(-1).
2998 61. The Exim daemon panic-logs an error return when it closes the incoming
2999 connection. However "connection reset by peer" seems to be common, and
3000 isn't really an error worthy of noting specially, so that particular error
3003 62. When Exim is trying to find all the local interfaces, it used to panic and
3004 die if the ioctl to get the interface flags failed. However, it seems that
3005 on at least one OS (Solaris 9) it is possible to have an interface that is
3006 included in the list of interfaces, but for which you get a failure error
3007 for this call. This happens when the interface is not "plumbed" into a
3008 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3009 failure of the "get flags" call assumes that the interface is down.
3011 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3012 makes life easier for people who are doing arithmetic on fields extracted
3013 from dates, where you often get leading zeros that should not be
3014 interpreted as octal.
3016 64. Added qualify_domain to the redirect router, to override the global
3019 65. If a pathologically long header line contained very many addresses (the
3020 report of this problem mentioned 10 000) and each of them was rewritten,
3021 Exim could use up a very large amount of memory. (It kept on making new
3022 copies of the header line as it rewrote, and never released the old ones.)
3023 At the expense of a bit more processing, the header rewriting function has
3024 been changed so that it no longer eats memory in this way.
3026 66. The generation of the Received: header has been moved from the time that a
3027 message starts to be received, to the time that it finishes. The timestamp
3028 in the Received: header should now be very close to that of the <= log
3029 line. There are two side-effects of this change:
3031 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3032 logged header lines no longer include the local Received: line, because
3033 it has not yet been created. The same applies to a copy of the message
3034 that is returned to a non-SMTP sender when a message is rejected.
3036 (b) When a filter file is tested using -bf, no additional Received: header
3037 is added to the test message. After some thought, I decided that this
3040 This change does not affect the value of $received_for. It is still set
3041 after address rewriting, but before local_scan() is called.
3043 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3045 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3046 gave an unhelpful panic error message, and a defer error. I have managed to
3047 change this behaviour so that it now rejects any supplied certificate,
3048 which seems right, as the list of acceptable certificates is empty.
3050 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3051 gave an unhelpful defer error. I have not managed to make this reject any
3052 supplied certificates, but the error message it gives is "no certificate
3053 supplied", which is not helpful.
3055 70. exigrep's output now also includes lines that are not associated with any
3056 message, but which match the given pattern. Implemented by a patch from
3057 Martin Sluka, which also tidied up the Perl a bit.
3059 71. Recipient callout verification, like sender verification, was using <> in
3060 the MAIL FROM command. This isn't really the right thing, since the actual
3061 sender may affect whether the remote host accepts the recipient or not. I
3062 have changed it to use the actual sender in the callout; this means that
3063 the cache record is now keyed on a recipient/sender pair, not just the
3064 recipient address. There doesn't seem to be a real danger of callout loops,
3065 since a callout by the remote host to check the sender would use <>.
3066 [SEE ABOVE: changed after hitting problems.]
3068 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3069 temporary errors. However, in the case of such a code being given after
3070 the end of a data transmission (i.e. after ".") Exim was failing to write
3071 a retry record for the message. (Yes, there was some broken host that was
3072 actually sending 8xx at this point.)
3074 73. An unknown lookup type in a host list could cause Exim to panic-die when
3075 the list was checked. (An example that provoked this was putting <; in the
3076 middle of a list instead of at the start.) If this happened during a DATA
3077 ACL check, a -D file could be left lying around. This kind of configuration
3078 error no longer causes Exim to die; instead it causes a defer errror. The
3079 incident is still logged to the main and panic logs.
3081 74. Buglet left over from Exim 3 conversion. The message "too many messages
3082 in one connection" was written to the rejectlog but not the mainlog, except
3083 when address rewriting (yes!) was being logged.
3085 75. Added write_rejectlog option.
3087 76. When a system filter was run not as root (that is, when system_filter_user
3088 was set), the values of the $n variables were not being returned to the
3089 main process; thus, they were not subsequently available in the $sn
3092 77. Added +return_path_on_delivery log selector.
3094 78. A connection timeout was being treated differently from recipients deferred
3095 when testing hosts_max_try with a message that was older than the host's
3096 retry timeout. (The host should not be counted, thus allowing all hosts to
3097 be tried at least once before bouncing.) This may have been the cause of an
3098 occasionally reported bug whereby a message would remain on the queue
3099 longer than the retry timeout, but would be bounced if a delivery was
3100 forced. I say "may" because I never totally pinned down the problem;
3101 setting up timeout/retry tests is difficult. See also the next item.
3103 79. The ultimate address timeout was not being applied to errors that involved
3104 a combination of host plus message (for example, a timeout on a MAIL
3105 command). When an address resolved to a number of possible hosts, and they
3106 were not all tried for each delivery (e.g. because of hosts_max_try), a
3107 message could remain on the queue longer than the retry timeout.
3109 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3112 81. Fixed an obscure SMTP outgoing bug which required at least the following
3113 conditions: (a) there was another message waiting for the same server;
3114 (b) the server returned 5xx to all RCPT commands in the first message so
3115 that the message was not completed; (c) the server dropped the connection
3116 or gave a negative response to the RSET that Exim sends to abort the
3117 transaction. The observed case was a dropped connection after DATA that had
3118 been sent in pipelining mode. That is, the server had advertised PIPELINING
3119 but was not implementing it correctly. The effect of the bug was incorrect
3120 behaviour, such as trying another host, and this could lead to a crash.
3126 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3127 and daemon.c were passed as pointers to ints; they should have been
3128 pointers to socklen_t variables (which are typically unsigned ints).
3130 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3133 3. Fixed a really odd bug that affected only the testing scheme; patching a
3134 certain fixed string in the binary changed the value of another string that
3135 happened to be identical to the end of the original first string.
3137 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3138 name", it returns that address as the IP address. On some operating
3139 systems (e.g. Solaris), it also passes back the IP address string as the
3140 "host name". However, on others (e.g. Linux), it passes back an empty
3141 string. Exim wasn't checking for this, and was changing the host name to an
3142 empty string, assuming it had been canonicized.
3144 5. Although rare, it is permitted to have more than one PTR record for a given
3145 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3146 all the names associated with an address, because they do in Solaris.
3147 However, it seems that they do not in Linux for data that comes from the
3148 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3149 I found this out when I moved to a new Linux workstation and tried to run
3150 the Exim test suite.
3152 To get round this problem I have changed the code so that it now does its
3153 own call to the DNS to look up PTR records when searching for a host name.
3154 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3155 addresses that are only in /etc/hosts are still found.
3157 This behaviour is, however, controlled by an option called host_lookup_
3158 order, which defaults to "bydns:byaddr". If people want to use the other
3159 order, or indeed, just use one or the other means of lookup, they can
3160 specify it in this variable.
3162 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3163 some operating systems, this comes back from gethostbyaddr() as an empty
3164 string, and this is what Exim used to test for. However, it seems that in
3165 other systems, "." is yielded. Exim now tests for this case too.
3167 7. The values of check_spool_space and check_log_space are now held internally
3168 as a number of kilobytes instead of an absolute number of bytes. If a
3169 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3170 kilobyte. This means that much larger values can be stored.
3172 8. Exim monitor: an attempt to get the action menu when not actually pointing
3173 at a message produces an empty menu entitled "No message selected". This
3174 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3175 no entries in it ("Shell widget menu has zero width and/or height"). So I
3176 have added a single, blank menu entry in this case.
3178 9. Added ${quote_local_part.
3180 10. MIME decoding is now applied to the contents of Subject: header lines when
3183 11. Now that a reference to $sender_host_address automatically causes a reverse
3184 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3185 host lookup before query-style lookups in lists that might use this
3186 variable. This has therefore been abolished, and the "net-" prefix is no
3187 longer necessary for query-style lookups.
3189 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3190 have been a typo for LFLAGS, so it has been changed.
3192 13. The install script calls Exim with "-C /dev/null" in order to find the
3193 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3194 to be output. Howeve, since Exim outputs its version number before the
3195 error, it didn't break the script. It just looked ugly. I fixed this by
3196 always allowing "-C /dev/null" if the caller is root.
3198 14. Ignore overlarge ACL variable number when reading spool file - insurance
3199 against a later release with more variables having written the file.
3201 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3202 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3203 this, as well as the form without the "IPv6" on the front.
3205 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3206 OS/Makefile-Darwin file.
3208 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3209 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3211 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3212 variable. However, some OS use ints for inodes. I've added cast to long int
3213 to get rid of the compiler warning.
3215 19. I had forgotten to lock out "/../" in configuration file names when
3216 ALT_CONFIG_PREFIX was set.
3218 20. Routers used for verification do not need to specify transports. However,
3219 if such a router generated a host list, and callout was configured, Exim
3220 crashed, because it could not find a port number from the (non-existent)
3221 transport. It now assumes port 25 in this circumstance.
3223 21. Added the -t option to exigrep.
3225 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3226 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3227 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3228 them is set, LOOKUP_LSEARCH is forced.
3230 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3232 24. Added sender and host information to the "rejected by local_scan()" log
3233 line; previously there was no indication of these.
3235 25. Added .include_if_exists.
3237 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3238 Linux. It turns out that not all file systems support this. Apparently some
3239 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3240 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3241 not supported on the file descriptor, is now ignored when Exim is trying to
3242 sync a directory. This applies only to Linux.
3244 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3246 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3247 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3250 29. Added support for maildirsize files from supplied patch (modified a bit).
3252 30. The use of :fail: followed by an empty string could lead Exim to respond to
3253 sender verification failures with (e.g.):
3255 550 Verification failed for <xxx>
3256 550 Sender verify failed
3258 where the first response line was missing the '-' that indicates it is not
3259 the final line of the response.
3261 31. The loop for finding the name of the user that called Exim had a hardwired
3262 limit of 10; it now uses the value of finduser_retries, which is used for
3263 all other user lookups.
3265 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3268 33. Exim was neglecting to zero errno before one call of strtol() when
3269 expanding a string and expecting an integer value. On some systems this
3270 resulted in spurious "integer overflow" errors. Also, it was casting the
3271 result into an int without checking.
3273 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3274 did not work. The code looks as if it has *never* worked, though it appears
3275 to have been documented since at least releast 1.62. I have made it work.
3277 35. The "timeout_DNS" error in retry rules, also documented since at least
3278 1.62, also never worked. As it isn't clear exactly what this means, and
3279 clearly it isn't a major issue, I have abolished the feature by treating it
3280 as "timeout", and writing a warning to the main and panic logs.
3282 36. The display of retry rules for -brt wasn't always showing the error code
3285 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3286 timeout_connect_A, timeout_connect_MX.
3288 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3289 to the empty sender.
3291 39. The daemon was not analysing the content of -oX till after it had closed
3292 stderr and disconnected from the controlling terminal. This meant that any
3293 syntax errors were only noted on the panic log, and the return code from
3294 the command was 0. By re-arranging the code a little, I've made the
3295 decoding happen first, so such errors now appear on stderr, and the return
3296 code is 1. However, the actual setting up of the sockets still happens in
3297 the disconnected process, so errors there are still only recorded on the
3300 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3301 connections (as happens on some IP stacks) was logged at start up time as
3302 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3303 it from "IPv6 and IPv4", which means that two separate sockets are being
3306 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3307 says whether AF_INET or AF_INET6 was passed as an argument.
3309 42. Exiwhat output was messed up when time zones were included in log
3312 43. Exiwhat now gives more information about the daemon's listening ports,
3313 and whether -tls-on-connect was used.
3315 44. The "port" option of the smtp transport is now expanded.
3317 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3318 silently ignored. Now an error message is written to the main and panic
3321 46. There's a new ACL modifier called "logwrite" which writes to a log file
3322 as soon as it is encountered.
3324 47. Added $local_user_uid and $local_user_gid at routing time.
3326 48. Exim crashed when trying to verify a sender address that was being
3329 49. Exim was recognizing only a space character after ".include". It now also
3330 recognizes a tab character.
3332 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3333 extracting the relevant information from the specification. The man page no
3334 longer contains scrambled data for the -d option, and I've added a section
3335 at the front about calling Exim under different names.
3337 51. Added "extra_headers" argument to the "mail" command in filter files.
3339 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3342 53. Installed eximstats 1.29.
3344 54. Added transport_filter_timeout as a generic transport option.
3346 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3347 Cc: header lines. This was required by RFC 822, but it not required by RFC
3350 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3351 incoming messages that did not have them. Now it does so only if the
3352 message originates locally, that is, if there is no associated remote host
3353 address. When Resent- header lines are present, this applies to the Resent-
3354 lines rather than the non-Resent- lines.
3356 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3357 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3359 58. Messages for configuration errors now include the name of the main
3360 configuration file - useful now that there may be more than one file in a
3361 list (.included file names were always shown).
3363 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3364 for those rare installations that do not start the daemon as root or run it
3365 setuid root. I've cut out the call to initgroups() if the daemon is not
3368 60. The Exim user and group can now be bound into the binary as text strings
3369 that are looked up at the start of Exim's processing.
3371 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3373 62. Added $mailstore_basename variable.
3375 63. Installed patch to sieve.c from Michael Haardt.
3377 64. When Exim failed to open the panic log after failing to open the main log,
3378 the original message it was trying to log was written to stderr and debug
3379 output, but if they were not available (the usual case in production), it
3380 was lost. Now it is written to syslog before the two lines that record the
3381 failures to open the logs.
3383 65. Users' Exim filters run in subprocesses under the user's uid. It is
3384 possible for a "deliver" command or an alias in a "personal" command to
3385 provoke an address rewrite. If logging of address rewriting is configured,
3386 this fails because the process is not running as root or exim. There may be
3387 a better way of dealing with this, but for the moment (because 4.30 needs
3388 to be released), I have disabled address rewrite logging when running a
3389 filter in a non-root, non-exim process.
3395 1. The buildconfig auxiliary program wasn't quoting the value set for
3396 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3397 not defined. This bug was masked in 4.22 by the effect that was fixed in
3400 2. Some messages that were rejected after a message id was allocated were
3401 shown as "incomplete" by exigrep. It no longer does this for messages that
3402 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3404 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3405 have allow_domain_literals set, the ID did not get logged in the <= line.
3406 Domain literals are now always recognized in Message-ID: header lines.
3408 4. The first argument for a ${extract expansion item is the key name or field
3409 number. Leading and trailing spaces in this item were not being ignored,
3410 causing some misleading effects.
3412 5. When deliver_drop_privilege was set, single queue runner processes started
3413 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3414 same command in the process it spins off) were not dropping privilege.
3416 6. When the daemon running as "exim" started a queue runner, it always
3417 re-executed Exim in the spun-off process. This is a waste of effort when
3418 deliver_drop_privilege is set. The new process now just calls the
3419 queue-runner function directly.
3425 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3428 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3429 ignored. Though the use of -oP was forcing the writing of a pid file, it
3430 was always written to the default place.
3432 3. If the message "no IP address found for host xxxx" is generated during
3433 incoming verification, it is now followed by identification of the incoming
3434 connection (so you can more easily find what provoked it).
3436 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3438 5. Added some features to "harden" Exim a bit more against certain attacks:
3440 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3441 be put in Local/Makefile. This is like the never_users runtime option,
3442 but it cannot be overridden. The default setting is "root".
3444 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3445 prefix string with which any file named in a -C command line option
3448 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3449 is retained for -C and -D only if the caller of Exim is root. Without
3450 it, the exim user may also use -C and -D and retain privilege.
3452 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3453 command line option is disabled.
3455 6. Macro names set by the -D option must start with an upper case letter, just
3456 like macro names defined in the configuration file.
3458 7. Added "dereference=" facility to LDAP.
3460 8. Two instances of the typo "uknown" in the source files are fixed.
3462 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3463 the Configure-Makefile script screwed up while processing it.
3465 10. Incorporated PCRE 4.4.
3467 11. The SMTP synchronization check was not operating right at the start of an
3468 SMTP session. For example, it could not catch a HELO sent before the client
3469 waited for the greeting. There is now a check for outstanding input at the
3470 point when the greeting is written. Because of the duplex, asynchronous
3471 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3472 way, but not yet received, when the check is performed.
3474 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3475 on TCP/IP sockets, because this apparently causes some broken clients to
3478 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3479 unchanged) from the Cygwin maintainer.
3481 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3482 was supported instead of testing for mbx. Effectively a typo.
3484 15. The spa authenticator server code was not checking that the input it
3485 received was valid base64.
3487 16. The debug output line for the "set" modifier in ACLs was not showing the
3488 name of the variable that was being set.
3490 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3492 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3493 lookup on its own. Something else was needed to trigger the lookup. For
3494 example, a match in host_lookup or the need for a host name in a host list.
3495 Now, if $sender_host_name is referenced and the host name has not yet been
3496 looked up, a lookup is performed. If the lookup fails, the variable remains
3497 empty, and $host_lookup_failed is set to "1".
3499 19. Added "eqi" as a case-independent comparison operator.
3501 20. The saslauthd authentication condition could segfault if neither service
3502 nor realm was specified.
3504 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3505 error message that was logged was misleading, and incoming SMTP
3506 connections were dropped. The message is now more accurate, and temporary
3507 errors are given to SMTP connections.
3509 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3510 (except RSET) with a 421 error, until QUIT is received. However, it was
3511 failing to send a response to QUIT.
3513 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3514 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3515 ACL is now run after the helo_try_verify_hosts code.
3517 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3518 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3519 case-independent, so other case variants are also recognized). Apparently
3520 some systems use these upper case variants.
3522 25. If more than two messages were waiting for the same host, and a transport
3523 filter was specified for the transport, Exim sent two messages over the
3524 same TCP/IP connection, and then failed with "socket operation on non-
3525 socket" when it tried to send the third.
3527 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3529 27. The extern definition of crypt16() in expand.c was not being excluded when
3530 the OS had its own crypt16() function.
3532 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3533 as a preferred synonym for return_size_limit, both as an option and as an
3536 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3538 30. Changed the default configuration ACL to relax the local part checking rule
3539 for addresses that are not in any local domains. For these addresses,
3540 slashes and pipe symbols are allowed within local parts, but the sequence
3541 /../ is explicitly forbidden.
3543 31. SPA server authentication was not clearing the challenge buffer before
3546 32. log_message in a "warn" ACL statement was writing to the reject log as
3547 well as to the main log, which contradicts the documentation and doesn't
3548 seem right (because no rejection is happening). So I have stopped it.
3550 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3551 However, I am unable to do any testing of this.
3553 34. Fixed an infelicity in the appendfile transport. When checking directories
3554 for a mailbox, to see if any needed to be created, it was accidentally
3555 using path names with one or more superfluous leading slashes; tracing
3556 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3558 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3559 contents are added to the log line that is written for every discarded
3560 recipient. (Previously a log_message setting was ignored.)
3562 36. The ${quote: operator now quotes the string if it is empty.
3564 37. The install script runs exim in order to find its version number. If for
3565 some reason other than non-existence or emptiness, which it checks, it
3566 could not run './exim', it was installing it with an empty version number,
3567 i.e. as "exim-". This error state is now caught, and the installation is
3570 38. An argument was missing from the function that creates an error message
3571 when Exim fails to connect to the socket for saslauthd authentication.
3572 This could cause Exim to crash, or give a corrupted message.
3574 39. Added isip, isip4, and isip6 to ${if conditions.
3576 40. The ACL variables $acl_xx are now saved with the message, and can be
3577 accessed later in routers, transports, and filters.
3579 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3580 strings in the file are not string-expanded.
3582 42. If a MAIL command specified a SIZE value that was too large to fit into an
3583 int variable, the check against message_size_limit failed. Such values are
3584 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3585 day this will have to be increased, but I don't think I want to be around
3586 when emails are that large.
3593 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3594 iconv() is not standard in FreeBSD.
3596 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3597 IPv6 enabled. The observed symptom was a segmentation fault on return from
3598 the function os_common_find_running_interfaces() in src/os.c.
3600 3. In the check_special_case() function in daemon.c I had used "errno" as an
3601 argument name, which causes warnings on some systems. This was basically a
3602 typo, since it was named "eno" in the comments!
3604 4. The code that waits for the clock to tick (at a resolution of some fraction
3605 of a second) so as to ensure message-id uniqueness was always waiting for
3606 at least one whole tick, when it could have waited for less. [This is
3607 almost certainly not relevant at current processor speeds, where it is
3608 unlikely to ever wait at all. But we try to future-proof.]
3610 5. The function that sleeps for a time interval that includes fractions of a
3611 second contained a race. It did not block SIGALRM between setting the
3612 timer, and suspending (a couple of lines later). If the interval was short
3613 and the sigsuspend() was delayed until after it had expired, the suspension
3614 never ended. On busy systems this could lead to processes getting stuck for
3617 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3618 process, before it forks any delivery processes. The open lookup caching
3619 mechanism meant that the open file or database connection was passed into
3620 the delivery process. The problem was that delivery processes always tidy
3621 up cached lookup data. This could cause a problem for the next delivery
3622 process started by the queue runner, because the external queue runner
3623 process does not know about the closure. So the next delivery process
3624 still has data in the lookup cache. In the case of a file lookup, there was
3625 no problem because closing a file descriptor in a subprocess doesn't affect
3626 the parent. However, if the lookup was caching a connection to a database,
3627 the connection was closed, and the second delivery process was likely to
3628 see errors such as "PGSQL: query failed: server closed the connection
3629 unexpectedly". The problem has been fixed by closing all cached lookups
3630 in a queue runner before running a delivery process.
3632 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3633 seem to have the "const" qualifier which it has on other OS. I've
3636 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3637 *of the same type* (client or server) with the same public name that an
3638 error should be diagnosed.
3640 9. When Exim looked up a host name for an IP address, but failed to find the
3641 original IP address when looking up the host name (a safety check), it
3642 output the message "<ip address> does not match any IP for NULL", which was
3643 confusing, to say the least. The bug was that the host name should have
3644 appeared instead of "NULL".
3646 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3647 user that is built into the binary, and the -C or -D options is used, root
3648 privilege is dropped before the configuration file is read. In addition,
3649 logging is switched to stderr instead of the normal log files. If the
3650 configuration then re-defines the Exim user, the unprivileged environment
3651 is probably not what is expected, so Exim logs a panic warning message (but
3654 However, if deliver_drop_privilege is set, the unprivileged state may well
3655 be exactly what is intended, so the warning has been cut out in that case,
3656 and Exim is allowed to try to write to its normal log files.
3662 1. smtp_return_error_details was not giving details for temporary sender
3663 or receiver verification errors.
3665 2. Diagnose a configuration error if two authenticators have the same public
3668 3. Exim used not to create the message log file for a message until the first
3669 delivery attempt. This could be confusing when incoming messages were held
3670 for policy or load reasons. The message log file is now created at the time
3671 the message is received, and an initial "Received" line is written to it.
3673 4. The automatically generated man page for command line options had a minor
3674 bug that caused no ill effects; however, a more serious problem was that
3675 the procedure for building the man page automatically didn't always
3676 operate. Consequently, release 4.20 contains an out-of-date version. This
3677 shouldn't happen again.
3679 5. When building Exim with embedded Perl support, the script that builds the
3680 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3681 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3683 6. The freeze_tell option was not being used for messages that were frozen on
3684 arrival, either by an ACL or by local_scan().
3686 7. Added the smtp_incomplete_transaction log selector.
3688 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3689 was accepting AUTH without a new EHLO.
3691 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3692 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3695 10. Logging of TCP/IP connections (when configured) now happens in the main
3696 daemon process instead of the child process, so that the TCP/IP connection
3697 count is more accurate (but it can never be perfect).
3699 11. The use of "drop" in a nested ACL was not being handled correctly in the
3700 outer ACL. Now, if condition failure induced by the nested "drop" causes
3701 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3702 or "require"), the connection is dropped.
3704 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3705 that yield "discard" can now be used with an "accept" or a "discard" verb,
3706 but an error is generated for any others (because I can't see a useful way
3707 to define what should happen).
3709 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3710 are now processed in the same way as lines in the Exim configuration file.
3711 In particular, continuation lines are supported.
3713 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3715 15. Added -ti meaning -t -i.
3717 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3718 domains, and warn by logging if others are found.
3720 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3721 returned from GIFCONF when Exim is trying to find the list of interfaces on
3722 a host. The code in os.c has been modified to copy each ifreq to an aligned
3723 structure in all cases.
3725 Also, in some cases, the returned ifreq's were being copied to a 'struct
3726 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3727 means the last couple of bytes of an IPv6 address could be chopped if the
3728 ifreq contained only a normal sockaddr (14 bytes storage).
3730 18. Named domain lists were not supported in the hosts_treat_as_local option.
3731 An entry such as +xxxx was not recognized, and was treated as a literal
3734 19. Ensure that header lines added by a DATA ACL are included in the reject log
3735 if the ACL subsequently rejects the message.
3737 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3738 MD5 (which is deprecated).
3740 21. When testing a filter file using -bf, Exim was writing a message when it
3741 took the sender from a "From " line in the message, but it was not doing so
3742 when it took $return_path from a Return-Path: header line. It now does.
3744 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3745 with a valid header line field name (a series of printing characters
3746 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3748 23. Changed "disc" in the source to "disk" to conform to the documentation and
3749 the book and for uniformity.
3751 24. Ignore Sendmail's -Ooption=value command line item.
3753 25. When execve() failed while trying to run a command in a pipe transport,
3754 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3755 could be confused with a return value of 69 from the command itself. This
3756 has been changed to 127, the value the shell returns if it is asked to run
3757 a non-existent command. The wording for the related log line suggests a
3758 non-existent command as the problem.
3760 26. If received_header_text expands to an empty string, do not add a Received:
3761 header line to the message. (Well, it adds a token one on the spool, but
3762 marks it "old" so that it doesn't get used or transmitted.)
3764 27. Installed eximstats 1.28 (addition of -nt option).
3766 28. There was no check for failure on the call to getsockname() in the daemon
3767 code. This can fail if there is a shortage of resources on the system, with
3768 ENOMEM, for example. A temporary error is now given on failure.
3770 29. Contrary to the C standard, it seems that in some environments, the
3771 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3772 program. Exim now does this explicitly; it affects the formatting of
3773 timestamps using strftime().
3775 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3776 complaints. I've now initialized all the variables, to avoid this.
3778 32. Header lines added by a system filter were not being "seen" during
3779 transport-time rewrites.
3781 33. The info_callback() function passed to OpenSSL is set up with type void
3782 (*)(SSL *, int, int), as described somewhere. However, when calling the
3783 function (actually a macro) that sets it up, the type void(*)() is
3784 expected. I've put in a cast to prevent warnings from picky compilers.
3786 34. If a DNS black list lookup found a CNAME record, but there were no A
3787 records associated with the domain it pointed at, Exim crashed.
3789 35. If a DNS black list lookup returned more than one A record, Exim ignored
3790 all but the first. It now scans all returned addresses if a particular IP
3791 value is being sought. In this situation, the contents of the
3792 $dnslist_value variable are a list of all the addresses, separated by a
3795 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3796 to accept a host name and all its aliases if the forward lookup for any of
3797 them yielded the IP address of the incoming connection. Now it accepts only
3798 those names whose forward lookup yields the correct IP address. Any other
3799 names are discarded. This closes a loophole whereby a rogue DNS
3800 administrator could create reverse DNS records to break through a
3801 wildcarded host restriction in an ACL.
3803 37. If a user filter or a system filter that ran in a subprocess used any of
3804 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3805 the wrong values were passed to the pipe command ($thisaddress had the
3806 value of $0, $0 had the value of $1, etc). This bug was introduced by
3807 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3809 38. Improved the line breaking for long SMTP error messages from ACLs.
3810 Previously, if there was no break point between 40 and 75 characters, Exim
3811 left the rest of the message alone. Two changes have been made: (a) I've
3812 reduced the minimum length to 35 characters; (b) if it can't find a break
3813 point between 35 and 75 characters, it looks ahead and uses the first one
3814 that it finds. This may give the occasional overlong line, but at least the
3815 remaining text gets split now.
3817 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3818 file descriptors might be low, and that setting 1000 would always raise it.
3819 It turns out that in some environments, the limit is already over 1000 and
3820 that lowering it causes trouble. So now Exim takes care not to decrease it.
3822 40. When delivering a message, the value of $return_path is set to $sender_
3823 address at the start of routing (routers may change the value). By an
3824 oversight, this default was not being set up when an address was tested by
3825 -bt or -bv, which affected the outcome if any router or filter referred to
3828 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3829 log only when "message" or "log_message" are set. However, if one of the
3830 conditions was an address verification, or a call to a nested ACL, the
3831 messages generated by the underlying test were being passed through. This
3832 no longer happens. The underlying message is available in $acl_verify_
3833 message for both "message" and "log_message" expansions, so it can be
3834 passed through if needed.
3836 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3837 new expansion $bh_ to give the encoded byte string without charset
3838 translation. Translation happens only if iconv() is available; HAVE_ICONV
3839 indicates this at build time. HEADERS_CHARSET gives the charset to
3840 translate to; headers_charset can change it in the configuration, and
3841 "headers charset" can change it in an individual filter file.
3843 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3844 that creates RFC 2047 encoded "words" labels them as that charset instead
3845 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3846 expansion operator; (ii) when Exim creates a From: line for a local
3847 message; (iii) when a header line is rewritten to include a "phrase" part.
3849 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3850 buggy, causing it to skip the first lines of messages whose message ID
3851 ended in 'D'. This would not have bitten before Exim release 4.14, because
3852 message IDs were unlikely to end in 'D' before then. The effect was to have
3853 incorrect size information for certain domains.
3855 45. #include "config.h" was missing at the start of the crypt16.c module. This
3856 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3859 46. If there was a timeout during a "random" callout check, Exim treated it as
3860 a failure of the random address, and carried on sending RSET and the real
3861 address. If the delay was just some slowness somewhere, the response to the
3862 original RCPT would be taken as a response to RSET and so on, causing
3863 mayhem of various kinds.
3865 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3866 when I implemented it. It didn't allow for the fact that some option values
3867 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3868 the right test for positive values.
3870 48. Domain names in DNS records are case-independent. Exim always looks them up
3871 in lower case. Some resolvers return domain names in exactly the case they
3872 appear in the zone file, that is, they may contain uppercase letters. Not
3873 all resolvers do this - some return always lower case. Exim was treating a
3874 change of case by a resolver as a change of domain, similar to a widening
3875 of a domain abbreviation. This triggered its re-routing code and so it was
3876 trying to route what was effectively the same domain again. This normally
3877 caused routing to fail (because the router wouldn't handle the domain
3878 twice). Now Exim checks for this case specially, and just changes the
3879 casing of the domain that it ultimately uses when it transmits the message
3882 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3885 50. If a filter generated a file delivery with a non-absolute name (possible if
3886 no home directory exists for the router), the forbid_file option was not
3889 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3890 the existing equality matching.
3892 52. Exim was using ints instead of ino_t variables in some places where it was
3893 dealing with inode numbers.
3895 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3896 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3897 called TMPDIR, and if it finds it is different, it changes its value.
3899 54. The smtp_printf() function is now made available to local_scan() so
3900 additional output lines can be written before returning. There is also an
3901 smtp_fflush() function to enable the detection of a dropped connection.
3902 The variables smtp_input and smtp_batched_input are exported to
3905 55. Changed the default runtime configuration: the message "Unknown user"
3906 has been removed from the ACL, and instead placed on the localuser router,
3907 using the cannot_route_message feature. This means that any verification
3908 failures that generate their own messages won't get overridden. Similarly,
3909 the "Unrouteable address" message that was in the ACL for unverifiable
3910 relay addresses has also been removed.
3912 56. Added hosts_avoid_esmtp to the smtp transport.
3914 57. The exicyclog script was not checking for the esoteric option
3915 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3916 will work only if exicyclog is run under the appropriate euid.
3918 58. Following a discussion on the list, the rules by which Exim recognises line
3919 endings on incoming messages have been changed. The -dropcr and drop_cr
3920 options are now no-ops, retained only for backwards compatibility. The
3921 following line terminators are recognized: LF CRLF CR. However, special
3922 processing applies to CR:
3924 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3925 nor a local message in the state where . is a terminator.
3927 (ii) If a bare CR is encountered in a header line, an extra space is added
3928 after the line terminator so as not to end the header. The reasoning
3929 behind this is that bare CRs in header lines are most likely either
3930 to be mistakes, or people trying to play silly games.
3932 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3933 was being incorrectly given as 18 bytes larger than it should have been.
3934 This is a VOB (very old bug).
3936 60. This may never have affected anything current, but just in case it has:
3937 When the local host is found other than at the start of a list of hosts,
3938 the local host, those with the same MX, and any that follow, are discarded.
3939 When the list in question was part of a longer list of hosts, the following
3940 hosts (not currently being processed) were also being discarded. This no
3941 longer happens. I'm not sure if this situation could ever has previously
3944 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3947 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3948 header. This is something that is recommended in a new Internet Draft, and
3949 is something that is documented as being done by Sendmail. There are two
3950 possible values. For messages generated by the autoreply transport, Exim
3953 Auto-Submitted: auto-replied
3955 whereas for all other generated messages (e.g. bounces) it adds
3957 Auto-Submitted: auto-generated
3959 63. The "personal" condition in filters now includes a test for the
3960 Auto-Submitted: header. If it contains the string "auto-" the message it
3961 not considered personal.
3963 64. Added rcpt_include_affixes as a generic transport option.
3965 65. Added queue_only_override (default true).
3967 66. Added the syslog_duplication option.
3969 67. If what should have been the first header line of a message consisted of
3970 a space followed by a colon, Exim was mis-interpreting it as a header line.
3971 It isn't of course - it is syntactically invalid and should therefore be
3972 treated as the start of the message body. The misbehaviour could have
3973 caused a number of strange effects, including loss of data in subsequent
3974 header lines, and spool format errors.
3976 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3977 client host had authenticated. This control can now be exercised by an ACL
3978 for more flexibility.
3980 69. By default, callouts do not happen when testing with -bh. There is now a
3981 variant, -bhc, which does actually run the callout code, including
3982 consulting and updating the callout cache.
3984 70. Added support for saslauthd authentication, courtesy of Alexander
3987 71. If statvfs() failed on the spool or log directories while checking their
3988 size for availability, Exim confusingly gave the error "space shortage".
3989 Furthermore, in debugging mode it crashed with a floating point exception.
3990 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3991 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3992 serious problem, Exim now writes to the main and panic logs when this
3993 happens, with details of the failure. It then refuses to accept the
3994 incoming message, giving the message "spool directory problem" or "log
3995 directory problem" with a 421 code for SMTP messages.
3997 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3998 0, 1, and 2 exist, because some OS complain for execs without them (see
3999 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4000 descriptors. However, the code omitted to check that the open succeeded,
4001 causing mysterious errors if for some reason the permissions on /dev/null
4002 got screwed. Now Exim writes a message to the main and panic logs, and
4003 bombs out if it can't open /dev/null.
4005 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4006 interact so that it is all more flexible. It is supposed to remain
4007 backwards compatible. Also added extra_local_interfaces.
4009 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4010 to bomb out with an assertion failure - to the client this appears as a
4011 connection drop. This problem occurs in the part of the code that was taken
4012 from the Samba project. Fortunately, the assertion is in a very simple
4013 function, so I have fixed this by reproducing the function inline in the
4014 one place where it is called, and arranging for authentication to fail
4015 instead of killing the process with assert().
4017 75. The SPA client code was not working when the server requested OEM rather
4018 than Unicode encoding.
4020 76. Added code to make require_files with a specific uid setting more usable in
4021 the case where statting the file as root fails - usually a non-root-mounted
4022 NFS file system. When this happens and the failure is EACCES, Exim now
4023 forks a subprocess and does the per-uid checking as the relevant uid.
4025 77. Added process_log_path.
4027 78. If log_file_path was not explicitly set, a setting of check_log_space or
4028 check_log_inodes was ignored.
4030 79. If a space check for the spool or log partitions fails, the incident is now
4031 logged. Of course, in the latter case the data may get lost...
4033 80. Added the %p formatting code to string_format() so that it can be used to
4034 print addresses in debug_print(). Adjusted all the address printing in the
4035 debugging in store.c to use %p rather than %d.
4037 81. There was a concern that a line of code in smtp_in.c could overflow a
4038 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4039 initially expressed, the concern was not well-founded, because trailing
4040 spaces are removed early. However, if the trailing spaces were followed by
4041 a NULL, they did not get removed, so the overflow was possible. Two fixes
4044 (a) I re-wrote the offending code in a cleaner fashion.
4045 (b) If an incoming SMTP command contains a NULL character, it is rejected
4048 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4049 runs initgroups(), so that if the Exim user is in any additional groups,
4050 they will be used during message reception.
4056 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob