1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.371 2006/07/07 13:54:32 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
81 PH/13 If write_rejectlog was set false when logging was sent to syslog with
82 syslog_duplication set false, log lines that would normally be written
83 both the the main log and to the reject log were not written to syslog at
86 PH/14 In the default configuration, change the use of "message" in ACL warn
87 statements to "add_header".
93 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
94 other effects) broke the use of negated acl sub-conditions.
96 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
99 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
100 "Deny" causes Exim to reject the incoming connection with a 554 error.
101 Unfortunately, if there is a major crisis, such as a disk failure,
102 tcp-wrappers gives "deny", whereas what one would like would be some
103 kind of temporary error. A kludge has been added to help with this.
104 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
105 554 error is used if errno is still zero or contains ENOENT (which occurs
106 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
109 PH/03 Add -lutil to the default FreeBSD LIBS setting.
111 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
112 errors. Otherwise a message that provokes a temporary error (when other
113 messages do not) can cause a whole host to time out.
115 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
116 addresses were routed directly to files or pipes from a redirect router.
117 File deliveries just didn't batch; pipe deliveries might have suffered
120 PH/06 A failure to get a lock for a hints database would erroneously always say
121 "Failed to get write lock", even when it was really a read lock.
123 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
124 of 0600. This has been changed to use the value of the lockfile_mode
125 option (which defaults to 0600).
127 PH/08 Applied small patch from the Sieve maintainer.
129 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
130 folder from quota calculations, a direct delivery into this folder messed
131 up the contents of the maildirsize file. This was because the regex was
132 used only to exclude .Trash (or whatever) when the size of the mailbox
133 was calculated. There was no check that a delivery was happening into an
134 excluded directory. This bug has been fixed by ignoring all quota
135 processing for deliveries into excluded directories.
137 PH/10 Added the maildirfolder_create_regex option to appendfile.
143 PH/01 The code for finding all the local interface addresses on a FreeBSD
144 system running IPv6 was broken. This may well have applied to all BSD
145 systems, as well as to others that have similar system calls. The broken
146 code found IPv4 interfaces correctly, but gave incorrect values for the
147 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
148 that it would not match correctly against @[] and not recognize the IPv6
151 PH/02 The ipliteral router was not recognizing addresses of the form user@
152 [ipv6:....] because it didn't know about the "ipv6:" prefix.
154 PH/03 Added disable_ipv6.
156 PH/04 Changed $reply_address to use the raw form of the headers instead of the
157 decoded form, because it is most often used to construct To: headers
158 lines in autoreplies, and the decoded form may well be syntactically
159 invalid. However, $reply_address has leading white space removed, and all
160 newlines turned into spaces so that the autoreply transport does not
163 PH/05 If group was specified without a user on a router, and no group or user
164 was specified on a transport, the group from the router was ignored.
166 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
167 for visible compile-time settings that can be used to change these
168 numbers, for those that want even more. Backwards compatibility with old
169 spool files has been maintained. However, going back to a previous Exim
170 release will lost any variables that are in spool files.
172 PH/07 Two small changes when running in the test harness: increase delay when
173 passing a TCP/IP connection to a new process, in case the original
174 process has to generate a bounce, and remove special handling of
175 127.0.0.2 (sic), which is no longer necessary.
177 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
178 be the same on different OS.
180 PH/09 Moved a debug statement in filter processing to avoid a race problem when
183 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
184 whether --show-vars was specified or not
186 JJ/02 exipick: Added support for new ACL variable spool format introduced
189 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
190 syntactically invalid From: or Reply-to: line, and a filter used this to
191 generate an autoreply, and therefore failed to obtain an address for the
192 autoreply, Exim could try to deliver to a non-existent relative file
193 name, causing unrelated and misleading errors. What now happens is that
194 it logs this as a hard delivery error, but does not attempt to create a
197 PH/11 The exinext utility has a -C option for testing purposes, but although
198 the given file was scanned by exinext itself; it wasn't being passed on
199 when Exim was called.
201 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
202 an end-of-file indication when reading a command response.
204 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
205 compiled. In many other places in Exim, IPv6 addresses are always
206 recognized, so I have changed this. It also means that IPv4 domain
207 literals of the form [IPV4:n.n.n.n] are now always recognized.
209 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
210 used if the router is not running as root, for example, when verifying at
211 ACL time, or when using -bh. The debugging output from this situation was
212 non-existent - all you got was a failure to exec. I have made two
215 (a) Failures to set uid/gid, the current directory, or a process leader
216 in a subprocess such as that created by queryprogram now generate
217 suitable debugging ouput when -d is set.
219 (b) The queryprogram router detects when it is not running as root,
220 outputs suitable debugging information if -d is set, and then runs
221 the subprocess without attempting to change uid/gid.
223 PH/15 Minor change to Makefile for building test_host (undocumented testing
226 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
227 additional section of a DNS packet that returns MX or SRV records.
228 Instead, it always explicitly searches for A/AAAA records. This avoids
229 major problems that occur when a DNS server includes only records of one
230 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
231 fixed another bug: if SRV records were looked up and the corresponding
232 address records were *not* found in the additional section, the port
233 values from the SRV records were lost.
235 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
236 using the correct key (the original address) when searching the retry
237 rules in order to find which one to use for generating the retry hint.
239 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
240 adding the default one. Similarly, if it contains a Reply-To: header, the
241 errors_reply_to option, if set, is not used.
243 PH/19 When calculating a retry time, Exim used to measure the "time since
244 failure" by looking at the "first failed" field in the retry record. Now
245 it does not use this if it is later than than the arrival time of the
246 message. Instead it uses the arrival time. This makes for better
247 behaviour in cases where some deliveries succeed, thus re-setting the
248 "first failed" field. An example is a quota failure for a huge message
249 when small messages continue to be delivered. Without this change, the
250 "time since failure" will always be short, possible causing more frequent
251 delivery attempts for the huge message than are intended.
252 [Note: This change was subsequently modified - see PH/04 for 4.62.]
254 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
255 $1, $2, $3) because the numerical variables can be reset during some
256 expansion items (e.g. "match"), thereby losing the authentication data.
258 PH/21 Make -bV show the size of off_t variables so that the test suite can
259 decide whether to run tests for quotas > 2G.
261 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
262 mailbox_size, and mailbox_filecount in the appendfile transport. If a
263 filecount value is greater than 2G or if a quota value is greater than 2G
264 on a system where the size of off_t is not greater than 4, a panic error
267 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
268 never match. The debug and -bh output now contains an explicit error
269 message indicating a malformed IPv4 address or mask.
271 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
272 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
275 PH/25 Do not write to syslog when running in the test harness. The only
276 occasion when this arises is a failure to open the main or panic logs
277 (for which there is an explicit test).
279 PH/26 Added the /no_tell option to "control=freeze".
281 PH/27 If a host name lookup failed very early in a connection, for example, if
282 the IP address matched host_lookup and the reverse lookup yielded a name
283 that did not have a forward lookup, an error message of the form "no IP
284 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
285 could be logged. Now it outputs the IP address instead of "NULL".
287 PH/28 An enabling patch from MH: add new function child_open_exim2() which
288 allows the sender and the authenticated sender to be set when
289 submitting a message from within Exim. Since child_open_exim() is
290 documented for local_scan(), the new function should be too.
292 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
293 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
294 results in an empty string is now treated as unset.
296 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
298 PH/31 Added sender_verify_fail logging option.
300 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
301 needed by RFC 822 but not by RFC 2822 was commented out. I have now
302 tidied the source and removed it altogether.
304 PH/33 When a queue run was abandoned because the load average was too high, a
305 log line was always written; now it is written only if the queue_run log
306 selector is set. In addition, the log line for abandonment now contains
307 information about the queue run such as the pid. This is always present
308 in "start" and "stop" lines but was omitted from the "abandon" line.
310 PH/34 Omit spaces between a header name and the colon in the error message that
311 is given when verify = headers_syntax fails (if there are lots of them,
312 the message gets confusing).
314 PH/35 Change the default for dns_check_names_pattern to allow slashes within
315 names, as there are now some PTR records that contain slashes. This check
316 is only to protect against broken name servers that fall over on strange
317 characters, so the fact that it applies to all lookups doesn't matter.
319 PH/36 Now that the new test suite is complete, we can remove some of the
320 special code in Exim that was needed for the old test suite. For example,
321 sorting DNS records because real resolvers return them in an arbitrary
322 order. The new test suite's fake resolver always returns records in the
325 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
326 bounces) except when queue_only is set, to avoid logging races between
327 the different processes.
329 PH/38 Panic-die if .include specifies a non-absolute path.
331 PH/39 A tweak to the "H" retry rule from its user.
333 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
334 a label. They prevented compilation on older perls.
336 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
337 a warning to be raised on newish perls.
339 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
340 on queue. Changes to match documented behaviour of showing count of
341 messages matching specified criteria.
343 PH/40 Changed the default ident timeout from 30s to 5s.
345 PH/41 Added support for the use of login_cap features, on those BSD systems
346 that have them, for controlling the resources used by pipe deliveries.
348 PH/42 The content-scanning code uses fopen() to create files in which to put
349 message data. Previously it was not paying any attention to the mode of
350 the files. Exim runs with umask(0) because the rest of the code creates
351 files with open(), and sets the required mode explicitly. Thus, these
352 files were ending up world-writeable. This was not a big issue, because,
353 being within the spool directory, they were not world-accessible. I have
354 created a function called modefopen, which takes an additional mode
355 argument. It sets umask(777), creates the file, chmods it to the required
356 mode, then resets the umask. All the relevant calls to fopen() in the
357 content scanning code have been changed to use this function.
359 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
360 to 24 hours. This avoids potential overflow problems when processing G
361 and H retry rules. I suspect nobody ever tinkers with this value.
363 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
365 PH/45 When the plaintext authenticator is running as a client, the server's
366 challenges are checked to ensure they are valid base64 strings. By
367 default, the authentication attempt is cancelled if an invalid string is
368 received. Setting client_ignore_invalid_base64 true ignores these errors.
369 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
370 they are received. Thus, the responses can be made to depend on the
371 challenges. If an invalid string is ignored, an empty string is placed in
374 PH/46 Messages that are created by the autoreply transport now contains a
375 References: header, in accordance with RFCs 2822 and 3834.
377 PH/47 Added authenticated_sender_force to the smtp transport.
379 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
381 PH/49 Installed latest patch from the Sieve maintainer.
383 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
384 was also set, the appendfile transport was unnecessarily scanning a
385 directory of message files (e.g. for maildir delivery) to find the count
386 of files (along with the size), even though it did not need this
387 information. It now does the scan only if it needs to find either the
388 size of the count of files.
390 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
392 PH/52 Two bugs concerned with error handling when the smtp transport is
395 (i) Exim was not creating retry information for temporary errors given
396 for individual recipients after the DATA command when the smtp transport
397 was used in LMTP mode. This meant that they could be retried too
398 frequently, and not timed out correctly.
400 (ii) Exim was setting the flag that allows error details to be returned
401 for LMTP errors on RCPT commands, but not for LMTP errors for individual
402 recipients that were returned after the DATA command.
404 PH/53 This is related to PH/52, but is more general: for any failing address,
405 when detailed error information was permitted to be returned to the
406 sender, but the error was temporary, then after the final timeout, only
407 "retry timeout exceeded" was returned. Now it returns the full error as
408 well as "retry timeout exceeded".
410 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
411 do this, and (what is worse) MTAs that accept it.
413 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
414 will now be deprecated.
416 PH/56 New os.c-cygwin from the Cygwin maintainer.
418 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
419 formats (previously only available in exim formats via -bpr, -bpru,
420 and -bpra. Now also available in native and exiqgrep formats)
422 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
423 with very large, slow to parse queues
425 JJ/08 exipick: added ! as generic prefix to negate any criteria format
427 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
429 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
430 responses to authentication challenges, though it was showing the
431 challenges; (ii) I've removed the CR characters from the debug output for
434 PH/58 Allow for the insertion of a newline as well as a space when a string
435 is turned into more than one encoded-word during RFC 2047 encoding. The
436 Sieve code now uses this.
438 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
439 data_4xx, lost_connection, tls_required.
441 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
442 message was being sent as an SMTP response.
444 PH/61 Add -l and -k options to exicyclog.
446 PH/62 When verifying, if an address was redirected to one new address, so that
447 verification continued, and the new address failed or deferred after
448 having set something in $address_data, the value of $address_data was not
449 passed back to the ACL. This was different to the case when no
450 redirection occurred. The value is now passed back in both cases.
452 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
453 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
454 use login_cap.h, so on its own it isn't the distinguishing feature. The
455 new name refers directly to the setclassresources() function.
457 PH/65 Added configuration files for NetBSD3.
459 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
461 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
462 is preferred over IPv4.
464 PH/68 The bounce_return_message and bounce_return_body options were not being
465 honoured for bounces generated during the reception of non-SMTP messages.
466 In particular, this applied to messages rejected by the ACL. This bug has
467 been fixed. However, if bounce_return_message is true and bounce_return_
468 body is false, the headers that are returned for a non-SMTP message
469 include only those that have been read before the error was detected.
470 (In the case of an ACL rejection, they have all been read.)
472 PH/69 The HTML version of the specification is now built in a directory called
473 spec_html instead of spec.html, because the latter looks like a path with
474 a MIME-type, and this confuses some software.
476 PH/70 Catch two compiler warnings in sieve.c.
478 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
479 function verify_get_ident() calls ip_connect() to connect a socket, but
480 if the "connect()" function timed out, ip_connect() used to close the
481 socket. However, verify_get_ident() also closes the socket later, and in
482 between Exim writes to the log, which may get opened at this point. When
483 the socket was closed in ip_connect(), the log could get the same file
484 descriptor number as the socket. This naturally causes chaos. The fix is
485 not to close the socket in ip_connect(); the socket should be closed by
486 the function that creates it. There was only one place in the code where
487 this was missing, in the iplookup router, which I don't think anybody now
488 uses, but I've fixed it anyway.
490 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
491 well as to direct DNS lookups. Otherwise the handling of names in host
492 lists is inconsistent and therefore confusing.
498 PH/01 Two changes to the default runtime configuration:
500 (1) Move the checks for relay_from_hosts and authenticated clients from
501 after to before the (commented out) DNS black list checks.
503 (2) Add control=submission to the relay_from_hosts and authenticated
504 clients checks, on the grounds that messages accepted by these
505 statements are most likely to be submissions.
507 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
509 (1) Generate an error if the third argument for the ${prvs expansion is
512 (2) Treat a missing third argument of ${prvscheck as if it were an empty
515 (3) Reset the variables that are obtained from the first argument of
516 ${prvscheck and used in the second argument before leaving the code,
517 because their memory is reclaimed, so using them afterwards may do
520 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
521 one (it's much easier than Tom thought :-).
523 (5) Because of (4), we can now allow for the use of $prvscheck_result
524 inside the third argument.
526 PH/03 For some reason, the default setting of PATH when running a command from
527 a pipe transport was just "/usr/bin". I have changed it to
530 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
531 anything to be listed in the output from -bV.
533 PH/05 When a filter generated an autoreply, the entire To: header line was
534 quoted in the delivery log line, like this:
536 => >A.N.Other <ano@some.domain> <original@ddress> ...
538 This has been changed so that it extracts the operative address. There
539 may be more than one such address. If so, they are comma-separated, like
542 => >ano@some.domain,ona@other.domain <original@ddress> ...
544 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
545 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
546 not being looked up in the rDNS to get a host name, Exim was showing the
547 IP address twice in Received: lines, even though the IP addresses were
548 identical. For example:
550 Received: from [1.2.3.4] (helo=[1.2.3.4])
552 However, if the real host name was known, it was omitting the HELO data
553 if it matched the actual IP address. This has been tidied up so that it
554 doesn't show the same IP address twice.
556 PH/07 When both +timestamp and +memory debugging was on, the value given by
557 $tod_xxx expansions could be wrong, because the tod_stamp() function was
558 called by the debug printing, thereby overwriting the timestamp buffer.
559 Debugging no longer uses the tod_stamp() function when +timestamp is set.
561 PH/08 When the original message was included in an autoreply transport, it
562 always said "this is a copy of the message, including all the headers",
563 even if body_only or headers_only was set. It now gives an appropriate
566 PH/09 Applied a patch from the Sieve maintainer which:
568 o fixes some comments
569 o adds the (disabled) notify extension core
570 o adds some debug output for the result of if/elsif tests
571 o points to the current vacation draft in the documentation
572 and documents the missing references header update
576 o fixes a bug in processing the envelope test (when testing
577 multiple envelope elements, the last element determinted the
580 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
581 Electronic Mail") by including:
583 Auto-submitted: auto-generated
585 in the messages that it generates (bounce messages and others, such as
586 warnings). In the case of bounce messages for non-SMTP mesages, there was
587 also a typo: it was using "Auto_submitted" (underscore instead of
588 hyphen). Since every message generated by Exim is necessarily in response
589 to another message, thes have all been changed to:
591 Auto-Submitted: auto-replied
593 in accordance with these statements in the RFC:
595 The auto-replied keyword:
597 - SHOULD be used on messages sent in direct response to another
598 message by an automatic process,
600 - MUST NOT be used on manually-generated messages,
602 - MAY be used on Delivery Status Notifications (DSNs) and Message
603 Disposition Notifications (MDNs),
605 - MUST NOT be used on messages generated by automatic or periodic
606 processes, except for messages which are automatic responses to
609 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
610 to the default Received: header definition.
612 PH/12 Added log selector acl_warn_skipped (default on).
614 PH/13 After a successful wildlsearch lookup, discard the values of numeric
615 variables because (a) they are in the wrong storage pool and (b) even if
616 they were copied, it wouldn't work properly because of the caching.
618 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
619 checking when decoding. Apparently there are clients that generate
620 overlong encoded strings. Why am I not surprised?
622 PH/15 If the first argument of "${if match_address" was not empty, but did not
623 contain an "@" character, Exim crashed. Now it writes a panic log message
624 and treats the condition as false.
626 PH/16 In autoreply, treat an empty string for "once" the same as unset.
628 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
629 extension "envelope-auth". The code is finished and in agreement with
630 other implementations, but there is no documentation so far and in fact,
631 nobody wrote the draft yet. This extension is currently #undef'ed, thus
632 not changing the active code.
634 Print executed "if" and "elsif" statements when debugging is used. This
635 helps a great deal to understand what a filter does.
637 Document more things not specified clearly in RFC3028. I had all this
638 sorted out, when out of a sudden new issues came to my mind. Oops."
640 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
643 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
644 canonical form (as documented). However, after a host name lookup from
645 the IP address, check_host() was doing a simple string comparison with
646 addresses acquired from the DNS when checking that the found name did
647 have the original IP as one of its addresses. Since any found IPv6
648 addresses are likely to be in abbreviated form, the comparison could
649 fail. Luckily, there already exists a function for doing the comparison
650 by converting both addresses to binary, so now that is used instead of
653 PH/20 There was another similar case to PH/19, when a complete host name was
654 given in a host list; looking up its IP address could give an abbreviated
655 form, whereas the current host's name might or might not be abbreviated.
656 The same fix has been applied.
662 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
663 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
666 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
667 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
669 PH/03 Typo: missing ".o" in src/pcre/Makefile.
671 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
672 header line, restrict the check to what is listed in RFCs 2369 and 2929.
673 Also, for "Auto-Submitted", treat anything other than "no" as
674 non-personal, in accordance with RFC 3834. (Previously it treated
675 anything starting "auto-" as non-personal.)
677 TF/01 The control=submission/name=... option had a problem with syntax
678 errors if the name included a slash character. The /name= option
679 now slurps the rest of the string, so it can include any characters
680 but it must come last in the list of options (after /sender_retain
683 PH/05 Some modifications to the interface to the fake nameserver for the new
691 TK/01 Added the "success_on_redirect" address verification option. See
692 NewStuff for rationale and an example.
694 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
696 PH/02 Patch to exigrep to allow it to work on syslog lines.
698 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
699 fread() to skip over the body file's header line, because in Cygwin the
700 header line is locked and is inaccessible.
702 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
703 co-exist for some time) to make it clear that it is the Exim ID that is
704 referenced, not the Message-ID: header line.
706 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
707 string_format() function, because snprintf() does not exist on all
710 PH/06 The use of forbid_filter_existstest now also locks out the use of the
711 ${stat: expansion item.
713 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
714 protocol synchronization error", to keep the pedants happy.
716 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
717 well as for IRIX systems, when gcc is being used. See the host.c source
720 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
722 PH/10 Named domain lists were not working if used in a queue_smtp_domains
725 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
726 transport and to the smtp transport in LMTP mode.
728 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
730 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
731 run a filter in a subprocess. This could lead to confusion in subsequent
732 lookups in the parent process. There should also be a search_tidyup() at
733 the end of the subprocess.
735 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
736 only if the host matched helo_try_verify_hosts, which caused the
737 verification to occur when the EHLO/HELO command was issued. The ACL just
738 tested the remembered result. Now, if a previous verification attempt has
739 not happened, "verify = helo" does it there and then.
741 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
743 TK/03 Fix log output including CR from clamd.
745 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
746 exist provoked a memory error which could cause a segfault.
748 PH/15 Installed PCRE 6.2
750 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
752 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
753 of the problem. Specifically, suggested +O2 rather than +O1 for the
756 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
758 PH/20 If a delivery was routed to a non-standard port by means of an SRV
759 record, the port was not correctly logged when the outgoing_port log
760 selector was set (it logged the transort's default port).
762 PH/21 Added support for host-specific ports to manualroute, queryprogram,
763 fallback_hosts, and "hosts" in the smtp transport.
765 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
766 host errors such as "Connection refused".
768 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
769 authentication with radiusclient 0.4.9:
771 - Error returned from rc_read_config was caught wrongly
772 - Username/password not passed on to radius server due to wrong length.
774 The presumption is that some radiusclient API changes for 4.51/PH/17
775 were not taken care of correctly. The code is still untested by me (my
776 Linux distribution still has 0.3.2 of radiusclient), but it was
777 contributed by a Radius user.
779 PH/24 When doing a callout, the value of $domain wasn't set correctly when
780 expanding the "port" option of the smtp transport.
782 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
783 while reading a MIME header. Thanks to Tom Hughes for a patch.
785 PH/24 Include config.h inside local_scan.h so that configuration settings are
788 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
789 that in an ACL for RCPT (for example), you can examine exactly what was
792 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
793 commands, but it was not correctly comparing the address with the actual
794 client host address. Thus, it would show the EHLO address in Received:
795 header lines when this was not necessary.
797 PH/27 Added the % operator to ${eval:}.
799 PH/28 Exim tries to create and chdir to its spool directory when it starts;
800 it should be ignoring failures (because with -C, for example, it has lost
801 privilege). It wasn't ignoring creation failures other than "already
804 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
807 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
808 because an input line was too long, either on its own, or by virtue of
809 too many continuations, the temporary file was not being removed, and the
810 return code was incorrect.
812 PH/31 Missing "BOOL" in function definition in filtertest.c.
814 PH/32 Applied Sieve patches from the maintainer.
816 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
818 PH/33 Added "verify = not_blind".
820 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
821 Local/Makefile (with some defaults set). These are used in built scripts
822 such as exicyclog, but they have never been used in the exim_install
823 script (though there are many overriding facilities there). I have
824 arranged that the exim_install script now takes note of these two
827 PH/35 Installed configuration files for Dragonfly.
829 PH/36 When a locally submitted message by a trusted user did not contain a
830 From: header, and the sender address was obtained from -f or from an SMTP
831 MAIL command, and the trusted user did not use -F to supply a sender
832 name, $originator_name was incorrectly used when constructing a From:
833 header. Furthermore, $originator_name was used for submission mode
834 messages from external hosts without From: headers in a similar way,
835 which is clearly wrong.
837 PH/37 Added control=suppress_local_fixups.
839 PH/38 When log_selector = +received_sender was set, and the addition of the
840 sender made the log line's construction buffer exactly full, or one byte
841 less than full, an overflow happened when the terminating "\n" was
844 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
845 when the result of a list match is failure because a DNS lookup failed.
847 PH/40 RM_COMMAND is now used in the building process.
849 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
850 the "build-* directories that it finds.
852 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
853 address in a domain literal was a prefix of an interface address.
855 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
856 when verifying a sender address, unless rewrite_headers is false.
858 PH/44 Wrote a long comment about why errors_to addresses are verified as
859 recipients, not senders.
861 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
862 the ratelimit ACL was added.
864 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
866 PH/46 Added extra information about PostgreSQL errors to the error string.
868 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
869 avoiding the need to install special zones in a real server. This is
870 backwards compatible; if it can't find the fake resolver, it drops back.
871 Thus, both old and new test suites can be run.
873 TF/02 Added util/ratelimit.pl
875 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
876 clock is set back in time.
878 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
879 Candler <B.Candler@pobox.com>.
881 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
882 OK for addresses that are the result of redirections.
884 PH/48 A number of further additions for the benefit of the new test suite,
885 including a fake gethostbyname() that interfaces to the fake DNS resolver
888 TF/06 The fix for widen_domains has also been applied to qualify_single and
889 search_parents which are the other dnslookup options that can cause
892 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
895 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
897 TF/07 Exim produced the error message "an SRV record indicated no SMTP
898 service" if it encountered an MX record with an empty target hostname.
899 The message is now "an MX or SRV record indicated no SMTP service".
901 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
902 if the DNS of the sending site is misconfigured. This is quite a
903 common situation. This change restores the behaviour of treating a
904 helo verification defer as a failure.
906 PH/51 If self=fail was set on a router, the bounce message did not include the
907 actual error message.
913 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
915 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
916 command itself ended in error, the underlying message about the transport
917 filter timeout was being overwritten with the pipe command error. Now the
918 underlying error message should be appended to the second error message.
920 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
922 PH/02 Reduce the amount of output that "make" produces by default. Full output
923 can still be requested.
925 PH/03 The warning log line about a condition test deferring for a "warn" verb
926 was being output only once per connection, rather than after each
927 occurrence (because it was using the same function as for successful
928 "warn" verbs). This seems wrong, so I have changed it.
930 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
931 it should not have, which might have caused a crash in the right
932 circumstances, but probably never did.
934 PH/04 Installed a modified version of Tony Finch's patch to make submission
935 mode fix the return path as well as the Sender: header line, and to
936 add a /name= option so that you can make the user's friendly name appear
939 TF/03 Added the control = fakedefer ACL modifier.
941 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
942 Mark Lowes for thorough testing.
944 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
946 TK/03 Merged latest SRS patch from Miles Wilton.
948 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
949 with the definition in sysexits.h (which is #included earlier).
950 Fortunately, Exim does not actually use EX_OK. The code used to try to
951 preserve the sysexits.h value, by assumimg that macro definitions were
952 scanned for macro replacements. I have been disabused of this notion,
953 so now the code just undefines EX_OK before #including unistd.h.
955 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
956 in the smtp transport. When a block could not be written in a single
957 write() function, the timeout was being re-applied to each part-write.
958 This seems wrong - if the receiver was accepting one byte at a time it
959 would take for ever. The timeout is now adjusted when this happens. It
960 doesn't have to be particularly precise.
962 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
963 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
965 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
966 without a domain if the check to <postmaster@domain> fails.
968 SC/01 Eximstats: added -xls and the ability to specify output files
969 (patch written by Frank Heydlauf).
971 SC/02 Eximstats: use FileHandles for outputing results.
973 SC/03 Eximstats: allow any combination of xls, txt, and html output.
975 SC/04 Eximstats: fixed display of large numbers with -nvr option
977 SC/05 Eximstats: fixed merging of reports with empty tables.
979 SC/06 Eximstats: added the -include_original_destination flag
981 SC/07 Eximstats: removed tabs and trailing whitespace.
983 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
985 TK/06 MBOX spool code: Add real "From " MBOX separator line
986 so the .eml file is really in mbox format (even though
987 most programs do not really care). Patch from Alex Miller.
989 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
990 The latter is generated from $received_to and is only set if the
991 message has one envelope recipient. SA can use these headers,
992 obviously out-of-the-box. Patch from Alex Miller.
994 PH/08 The ${def test on a variable was returning false if the variable's
995 value was "0", contrary to what the specification has always said!
996 The result should be true unless the variable is empty.
998 PH/09 The syntax error of a character other than { following "${if
999 def:variable_name" (after optional whitespace) was not being diagnosed.
1000 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1001 accidental colon was present, for example, could give incorrect results.
1003 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1004 result is used (not including appendfile, where other changes are about
1007 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1008 This involved changing a lot of size variables from int to off_t. It
1009 should work with maildirs and everything.
1011 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1012 spamd dying while we are connected to it.
1014 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1017 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1018 function for BDB 4.3.
1020 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1022 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1023 the sources for PCRE have been re-arranged and more files are now
1026 PH/15 The code I had for printing potentially long long variables in PH/11
1027 above was not the best (it lost precision). The length of off_t variables
1028 is now inspected at build time, and an appropriate printing format (%ld
1029 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1030 to be "long long int" or "long int". This is needed for the internal
1031 formatting function string_vformat().
1033 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1034 the configuration file to be ":syslog", then the script "guesses" where
1035 the logs files are, rather than using the compiled in default. In our
1036 case the guess is not the same as the compiled default, so the script
1037 suddenly stopped working when I started to use syslog. The patch checks
1038 to see if log_file_path is "". If so, it attempts to read it from exim
1039 with no configuration file to get the compiled in version, before it
1040 falls back to the previous guessing code."
1042 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1043 implementing BATV in an Exim configuration. See NewStuff for the gory
1046 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1047 Makefile that are specific to HP-UX.
1049 PH/18 If the "use_postmaster" option was set for a recipient callout together
1050 with the "random" option, the postmaster address was used as the MAIL
1051 FROM address for the random test, but not for the subsequent recipient
1052 test. It is now used for both.
1054 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1055 patch removes a few documentation additions to RFC 3028, because the
1056 latest draft now contains them. It adds the new en;ascii-case comparator
1057 and a new error check for 8bit text in MIME parts. Comparator and
1058 require names are now matched exactly. I enabled the subaddress
1059 extension, but it is not well tested yet (read: it works for me)."
1061 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1062 rework some of the code of TK/09 above to avoid the hardwired use of
1063 "%lld" and "long long". Replaced the call to snprintf() with a call to
1066 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1067 records point to non-existent hosts", "retry timeout exceeded", and
1068 "retry time not reached for any host after a long failure period".
1070 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1071 experimental DomainKeys support:
1073 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1074 (2) On an error such as an illegally used "control", the wrong name for
1075 the control was given.
1077 These problems did NOT occur unless DomainKeys support was compiled.
1079 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1081 PH/24 Added ${if match_ip condition.
1083 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1084 databases so that it will be absolutely obvious if a crash occurs in the
1085 DB library. This is a regular occurrence (often caused by mis-matched
1088 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1089 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1090 picked up on a user's system that detects such things. There doesn't seem
1091 to be a gcc warning option for this - only an attribute that has to be
1092 put on the function's prototype. It seems that in Fedora Core 4 they have
1093 set this on a number of new functions. No doubt there will be more in due
1096 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1097 specify a transport. However, if an address that was verified by such a
1098 router was the subject of a callout, Exim crashed because it tried to
1099 read the rcpt_include_affixes from the non-existent transport. Now it
1100 just assumes that the setting of that option is false. This bug was
1101 introduced by 4.51/PH/31.
1103 PH/28 Changed -d+all to exclude +memory, because that information is very
1104 rarely of interest, but it makes the output a lot bigger. People tend to
1105 do -d+all out of habit.
1107 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1108 code in os-type was giving problems when libc.so lives in lib64, like on
1111 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1112 aren't the modern standard, and it seems that some systems' include files
1113 don't always have them. Exim was already checking for some of the newer
1114 ones like T_AAAA, and defining it itself. I've added checks for all the
1115 record types that Exim uses.
1117 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1118 not automatically generating a new one, as it is supposed to. This
1119 prevented TLS from working. If the file did exist, but contained invalid
1120 data, a new version was generated, as expected. It was only the case of a
1121 non-existent file that was broken.
1123 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1124 with a change in libdomainkeys > 0.64.
1126 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1127 from DNS. If the selector record carries the flag, it now has
1128 precedence over the domain-wide flag.
1130 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1132 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1133 the use of a port name that isn't defined in /etc/services) occurred, the
1134 message was deferred as in a normal delivery, and thus remained on the
1135 spool, instead of being failed because of the mua_wrapper setting. This
1136 is now fixed, and I tidied up some of the mua_wrapper messages at the
1139 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1140 the messages in a hash of arrays rather than using individual hashes.
1141 This is a bit cleaner and results in dramatic memory savings, albeit
1142 at a slight CPU cost.
1144 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1145 as requested by Marc Sherman.
1147 SC/10 Eximstats: added histograms for user specified patterns as requested
1150 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1152 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1153 fopen() in the content-scanning modules that did not already have it.
1159 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1160 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1162 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1164 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1166 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1167 to test Sieve filters that use "vacation".
1169 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1170 that changes the way the GnuTLS parameters are stored in the cache file.
1171 The new format can be generated externally. For backward compatibility,
1172 if the data in the cache doesn't make sense, Exim assumes it has read an
1173 old-format file, and it generates new data and writes a new file. This
1174 means that you can't go back to an older release without removing the
1177 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1178 work if there are any delivery delays because "one_time" forces the
1179 parent to be marked "delivered", so its unseen clone is never tried
1180 again. For this reason, Exim now forbids the simultaneous setting of
1183 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1184 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1185 entry if you want to know the details. The fix, however, neglected to
1186 consider the case when local delivery batching is involved. The test for
1187 "previously delivered" was not happening when checking to see if an
1188 address could be batched with a previous (undelivered) one; under
1189 certain circumstances this could lead to multiple deliveries to the same
1192 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1193 in its include files, and this causes problems building Exim.
1195 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1196 header_syntax) but Exim was just ignoring anything given after a slash.
1197 In particular, this caused confusion with an attempt to use "verify =
1198 reverse_host_lookup/defer_ok". An error is now given when options are
1199 supplied for verify items that do not have them. (Maybe reverse_host_
1200 lookup should have a defer_ok option, but that's a different point.)
1202 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1203 defined by RFC 821) to 2048, because there were problems with some AUTH
1204 commands, and RFC 1869 says the size should be increased for extended
1205 SMTP commands that take arguments.
1207 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1210 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1211 "unknown" error; now it says that the functionality isn't in the binary.
1213 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1214 an address' error message when a string expansion fails (syntax or
1215 whatever). Otherwise the password may appear in the log. Following change
1216 PH/42 below, there is no longer a chance of it appearing in a bounce
1219 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1221 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1222 of its addresses was ever tried. (Bugzilla bug #2.)
1224 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1225 the result incorrectly in the debug output. (It correctly added a newline
1226 to what was transported.)
1228 TF/01 Added $received_time.
1230 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1231 commented out examples of how to interface to a virus scanner and to
1232 SpamAssassin. Also added commented examples of av_scanner and
1233 spamd_address settings.
1235 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1236 and controls are allowed in which ACLs. There were a couple of minor
1237 errors. Some of the entries in the conditions table (which is a table of
1238 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1239 negation of where the condition IS allowed.
1241 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1243 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1244 header file does not have a version number, so I've had to invent a new
1245 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1246 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1247 radiusclient), but it was contributed by a Radius user.
1249 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1250 files or directories, for OpenSSL.
1252 PH/19 When an Exim process that is running as root has to create an Exim log
1253 file, it does so in a subprocess that runs as exim:exim so as to get the
1254 ownership right at creation (otherwise, other Exim processes might see
1255 the file with the wrong ownership). There was no test for failure of this
1256 fork() call, which would lead to the process getting stuck as it waited
1257 for a non-existent subprocess. Forks do occasionally fail when resources
1258 run out. I reviewed all the other calls to fork(); they all seem to check
1261 PH/20 When checking for unexpected SMTP input at connect time (before writing
1262 the banner), Exim was not dealing correctly with a non-positive return
1263 from the read() function. If the client had disconnected by this time,
1264 the result was a log entry for a synchronization error with an empty
1265 string after "input=" when read() returned zero. If read() returned -1
1266 (an event I could not check), uninitialized data bytes were printed.
1267 There were reports of junk text (parts of files, etc) appearing after
1270 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1272 PH/22 Added support for macro redefinition, and (re)definition in between
1273 driver and ACL definitions.
1275 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1276 forgetting to use the resulting value; it was using the unexpanded value.
1278 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1279 hadn't been configured. The fix is from Juergen Kreileder, who
1280 understands it better than I do:
1282 "Here's what I see happening with three configured cyrus_sasl
1283 authenticators configured (plain, login, cram-md5):
1285 On startup auth_cyrus_sasl_init() gets called for each of these.
1286 This means three calls to sasl_listmech() without a specified mech_list.
1287 => SASL tests which mechs of all available mechs actually work
1288 => three warnings about OTP not working
1289 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1291 With the patch, sasl_listmech() also gets called three times. But now
1292 SASL's mech_list option is set to the server_mech specified in the the
1293 authenticator. Or in other words, the answer from sasl_listmech()
1294 gets limited to just the mech you're testing for (which is different
1296 => the return list contains just 'plain' or 'login', 'cram-md5' or
1297 nothing depending on the value of ob->server_mech.
1299 I've just tested the patch: Authentication still works fine,
1300 unavailable mechs specified in the exim configuration are still
1301 caught, and the auth.log warnings about OTP are gone."
1303 PH/25 When debugging is enabled, the contents of the command line are added
1304 to the debugging output, even when log_selector=+arguments is not
1307 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1308 answer is "GNU", and only if the return is "GNU/something" is the answer
1311 PH/27 $acl_verify_message is now set immediately after the failure of a
1312 verification in an ACL, and so is available in subsequent modifiers. In
1313 particular, the message can be preserved by coding like this:
1315 warn !verify = sender
1316 set acl_m0 = $acl_verify_message
1318 Previously, $acl_verify_message was set only while expanding "message"
1319 and "log_message" when a very denied access.
1321 PH/28 Modified OS/os.c-Linux with
1323 -#ifndef OS_LOAD_AVERAGE
1324 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1326 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1327 nomenclature these days.)
1329 PH/29 Installed patch from the Sieve maintainer that adds the options
1330 sieve_useraddress and sieve_subaddress to the redirect router.
1332 PH/30 In these circumstances:
1333 . Two addresses routed to the same list of hosts;
1334 . First host does not offer TLS;
1335 . First host accepts first address;
1336 . First host gives temporary error to second address;
1337 . Second host offers TLS and a TLS session is established;
1338 . Second host accepts second address.
1339 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1340 and peerdn, if requested) that were in fact used only for the second
1343 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1344 attention to any local part prefix or suffix that was matched by the
1345 router that accepted the address. It now behaves in the same way as it
1346 does for delivery: the affixes are removed from the local part unless
1347 rcpt_include_affixes is set on the transport.
1349 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1350 timeout during the DATA phase of an incoming message.
1352 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1353 applied a patch sanctioned by the Sieve maintainer.
1355 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1356 the uid or gid is negative. A case of a negative gid caused this to be
1357 noticed. The fix allows for either to be negative.
1359 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1360 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1363 PH/36 The change PH/12 above was broken. Fixed it.
1365 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1366 the grounds that routing the same address twice would always produce the
1367 same answer. This might have been true once, but it is certainly no
1368 longer true now. Routing a child address may depend on the previous
1369 routing that produced that child. Some complicated redirection strategies
1370 went wrong when messages had multiple recipients, and made Exim's
1371 behaviour dependent on the order in which the addresses were given.
1373 I have moved the duplicate checking until after the routing is complete.
1374 Exim scans the addresses that are assigned to local and remote
1375 transports, and removes any duplicates. This means that more work will be
1376 done, as duplicates will always all be routed, but duplicates are
1377 presumably rare, so I don't expect this is of any significance.
1379 For deliveries to pipes, files, and autoreplies, the duplicate checking
1380 still happens during the routing process, since they are not going to be
1383 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1384 It corrects a timeout issue with spamd. This is Ian's comment: "The
1385 background is that sometimes spamd either never reads data from a
1386 connection it has accepted, or it never writes response data. The exiscan
1387 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1388 blindly assumes that writes won't block so it may never time out."
1390 PH/39 Allow G after quota size as well as K and M.
1392 PH/40 The value set for $authenticated_id in an authenticator may not contain
1393 binary zeroes or newlines because the value is written to log lines and
1394 to spool files. There was no check on this. Now the value is run through
1395 the string_printing() function so that such characters are converted to
1396 printable escape sequences.
1398 PH/41 $message_linecount is a new variable that contains the total number of
1399 lines in the message. Compare $body_linecount, which is the count for the
1402 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1403 bounce and delay warning messages, except in certain special cases, which
1406 (a) An SMTP error message from a remote host;
1407 (b) A message specified in a :fail: redirection;
1408 (c) A message specified in a "fail" command in a system filter;
1409 (d) A message specified in a FAIL return from the queryprogram router;
1410 (e) A message specified by the cannot_route_message router option.
1412 In these cases only, Exim does include the error details in bounce and
1413 warning messages. There are also a few cases where bland messages such
1414 as "unrouteable address" or "local delivery error" are given.
1416 PH/43 $value is now also set for the "else" part of a ${run expansion.
1418 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1419 being worked on, but at least Exim now implements the latest version to
1422 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1423 process to complete was treated as a delivery failure, a timeout while
1424 writing the message to the pipe was logged, but erroneously treated as a
1425 successful delivery. Such timeouts include transport filter timeouts. For
1426 consistency with the overall process timeout, these timeouts are now
1427 treated as errors, giving rise to delivery failures by default. However,
1428 there is now a new Boolean option for the pipe transport called
1429 timeout_defer, which, if set TRUE, converts the failures into defers for
1430 both kinds of timeout. A transport filter timeout is now identified in
1433 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1434 systems where "make" and "gmake" are different, calling "gmake" at top
1435 level broke things. I've arranged for the value of $(MAKE) to be passed
1436 from the Makefile to this script so that it can call the same version of
1440 A note about Exim versions 4.44 and 4.50
1441 ----------------------------------------
1443 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1444 changes of various kinds. As a consequence, a big documentation update was
1445 needed. This delayed the release for rather longer than seemed good, especially
1446 in the light of a couple of (minor) security issues. Therefore, the changes
1447 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1448 release. So 4.44 and 4.50 are in effect two different branches that both start
1451 I have left the 4.50 change log unchanged; it contains all the changes since
1452 4.43. The change log for 4.44 is below; many of its items are identical to
1453 those for 4.50. This seems to be the most sensible way to preserve the
1454 historical information.
1460 1. Minor wording change to the doc/README.SIEVE file.
1462 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1463 computation of the current number of files was incorrect.
1465 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1466 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1469 4. Give more explanation in the error message when the command for a transport
1470 filter fails to execute.
1472 5. There are several places where Exim runs a non-Exim command in a
1473 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1474 was being done only for the command run by the queryprogram router. It is
1475 now done for all such subprocesses. The other cases are: ${run, transport
1476 filters, and the commands run by the lmtp and pipe transports.
1478 6. Added CONFIGURE_GROUP build-time option.
1480 7. Some older OS have a limit of 256 on the maximum number of file
1481 descriptors. Exim was using setrlimit() to set 1000 as a large value
1482 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1483 systems. I've change it so that if it can't get 1000, it tries for 256.
1485 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1486 was an oversight, and furthermore, ever since the addition of extra
1487 controls (e.g. 4.43/32), the checks on when to allow different forms of
1488 "control" were broken. There should now be diagnostics for all cases when a
1489 control that does not make sense is encountered.
1491 9. Added the /retain_sender option to "control=submission".
1493 10. $recipients is now available in the predata ACL (oversight).
1495 11. Tidy the search cache before the fork to do a delivery from a message
1496 received from the command line. Otherwise the child will trigger a lookup
1497 failure and thereby defer the delivery if it tries to use (for example) a
1498 cached ldap connection that the parent has called unbind on.
1500 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1501 of $address_data from the recipient verification was clobbered by the
1502 sender verification.
1504 13. The value of address_data from a sender verification is now available in
1505 $sender_address_data in subsequent conditions in the ACL statement.
1507 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1509 15. Added a new option "connect=<time>" to callout options, to set a different
1512 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1513 was its contents. (It was OK if the option was not defined at all.)
1515 17. A "Completed" log line is now written for messages that are removed from
1516 the spool by the -Mrm option.
1518 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1519 information about exactly what failed.
1521 19. Added -dd to debug only the daemon process.
1523 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1524 handles timeouts, both on the server side and network timeouts. Renamed the
1525 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1527 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1528 It is now set to "smtps".
1530 22. $host_address is now set to the target address during the checking of
1531 ignore_target_hosts.
1533 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1534 being passed; this would have caused $sender_host_name to have been used if
1535 matching the list had actually called for a host name (not very likely,
1536 since this list is usually IP addresses). A host name is now passed as
1539 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1540 code that specifies a non-restarting handler (typically sigaction() in
1541 modern systems) in an attempt to fix a rare and obscure crash bug.
1543 25. Narrowed the window for a race in the daemon that could cause it to ignore
1544 SIGCHLD signals. This is not a major problem, because they are used only to
1545 wake it up if nothing else does.
1547 26. A malformed maildirsize file could cause Exim to calculate negative values
1548 for the mailbox size or file count. Odd effects could occur as a result.
1549 The maildirsize information is now recalculated if the size or filecount
1552 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1553 support for a long time. Removed HAVE_SYS_VFS_H.
1555 28. Installed the latest version of exipick from John Jetmore.
1557 29. In an address list, if the pattern was not a regular expression, an empty
1558 subject address (from a bounce message) matched only if the pattern was an
1559 empty string. Non-empty patterns were not even tested. This was the wrong
1560 because it is perfectly reasonable to use an empty address as part of a
1561 database query. An empty address is now tested by patterns that are
1562 lookups. However, all the other forms of pattern expect the subject to
1563 contain a local part and a domain, and therefore, for them, an empty
1564 address still always fails if the pattern is not itself empty.
1566 30. Exim went into a mad DNS loop when attempting to do a callout where the
1567 host was specified on an smtp transport, and looking it up yielded more
1568 than one IP address.
1570 31. Re-factored the code for checking spool and log partition space into a
1571 function that finds that data and another that does the check. The former
1572 is then used to implement four new variables: $spool_space, $log_space,
1573 $spool_inodes, and $log_inodes.
1575 32. The RFC2047 encoding function was originally intended for short strings
1576 such as real names; it was not keeping to the 75-character limit for
1577 encoded words that the RFC imposes. It now respects the limit, and
1578 generates multiple encoded words if necessary. To be on the safe side, I
1579 have increased the buffer size for the ${rfc2047: expansion operator from
1582 33. It is now permitted to omit both strings after an "if" condition; if the
1583 condition is true, the result is "true". As before, when the second string
1584 is omitted, a false condition yields an empty string. This makes it less
1585 cumbersome to write custom ACL and router conditions.
1587 34. Failure to deliver a bounce message always caused it to be frozen, even if
1588 there was an errors_to setting on the router. The errors_to setting is now
1591 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1592 canonical form (fully expanded) before being placed in
1593 $sender_host_address.
1595 36. The table in the code that translates DNS record types into text (T_A to
1596 "A" for instance) was missing entries for NS and CNAME. It is just possible
1597 that this could have caused confusion if both these types were looked up
1598 for the same domain, because the text type is used as part of Exim's
1599 per-process caching. But the chance of anyone hitting this buglet seems
1602 37. The dnsdb lookup has been extended in a number of ways.
1604 (1) There is a new type, "zns", which walks up the domain tree until it
1605 finds some nameserver records. It should be used with care.
1607 (2) There is a new type, "mxh", which is like "mx" except that it returns
1608 just the host names, not the priorities.
1610 (3) It is now possible to give a list of domains (or IP addresses) to be
1611 looked up. The behaviour when one of the lookups defers can be
1612 controlled by a keyword.
1614 (4) It is now possible to specify the separator character for use when
1615 multiple records are returned.
1617 38. The dnslists ACL condition has been extended: it is now possible to supply
1618 a list of IP addresses and/or domains to be looked up in a particular DNS
1621 39. Added log_selector=+queue_time_overall.
1623 40. When running the queue in the test harness, wait just a tad after forking a
1624 delivery process, to get repeatability of debugging output.
1626 41. Include certificate and key file names in error message when GnuTLS fails
1627 to set them up, because the GnuTLS error message doesn't include the name
1628 of the failing file when there is a problem reading it.
1630 42. Allow both -bf and -bF in the same test run.
1632 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1634 44. The "Exiscan patch" is now merged into the mainline Exim source.
1636 45. Sometimes the final signoff response after QUIT could fail to get
1637 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1638 < 0 before doing a fflush(). This bug looks as though it goes back to the
1639 introduction of TLS in release 3.20, but "sometimes" must have been rare
1640 because the tests only now provoked it.
1642 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1643 (this can affect the format of dates).
1645 47. exim_tidydb, when checking for the continued existence of a message for
1646 which it has found a message-specific retry record, was not finding
1647 messages that were in split spool directories. Consequently, it was
1648 deleting retry records that should have stayed in existence.
1650 48. Steve fixed some bugs in eximstats.
1652 49. The SPA authentication driver was not abandoning authentication and moving
1653 on to the next authenticator when an expansion was forced to fail,
1654 contradicting the general specification for all authenticators. Instead it
1655 was generating a temporary error. It now behaves as specified.
1657 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1658 (the order specifies the preference for clients). The order is now AES256,
1659 AES128, 3DES, ARCFOUR128.
1661 51. Small patch to Sieve code - explicitly set From: when generating an
1664 52. Exim crashed if a remote delivery caused a very long error message to be
1665 recorded - for instance if somebody sent an entire SpamAssassin report back
1666 as a large number of 550 error lines. This bug was coincidentally fixed by
1667 increasing the size of one of Exim's internal buffers (big_buffer) that
1668 happened as part of the Exiscan merge. However, to be on the safe side, I
1669 have made the code more robust (and fixed the comments that describe what
1672 53. Now that there can be additional text after "Completed" in log lines (if
1673 the queue_time_overall log selector is set), a one-byte patch to exigrep
1674 was needed to allow it to recognize "Completed" as not the last thing in
1677 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1678 patch that reportedly fixes this has been added. I am not expert enough to
1679 create a test for it. This is what the patch creator wrote:
1681 "I found a little strange behaviour of ldap code when working with
1682 Windows 2003 AD Domain, where users was placed in more than one
1683 Organization Units. When I tried to give exim partial DN, the exit code
1684 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1685 But simultaneously result of request was absolutely normal ldap result,
1686 so I produce this patch..."
1688 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1689 REFERENCE, so I have modified the code to exclude the patch when that macro
1692 55. Some experimental protocols are using DNS PTR records for new purposes. The
1693 keys for these records are domain names, not reversed IP addresses. The
1694 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1695 leaves it alone. Component reversal etc. now happens only for IP addresses.
1698 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1700 57. Double the size of the debug message buffer (to 2048) so that more of very
1701 long debug lines gets shown.
1703 58. The exicyclog utility now does better if the number of log files to keep
1704 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1706 59. Two changes related to the smtp_active_hostname option:
1708 (1) $smtp_active_hostname is now available as a variable.
1709 (2) The default for smtp_banner uses $smtp_active_hostname instead
1710 of $primary_hostname.
1712 60. The host_aton() function is supposed to be passed a string that is known
1713 to be a valid IP address. However, in the case of IPv6 addresses, it was
1714 not checking this. This is a hostage to fortune. Exim now panics and dies
1715 if the condition is not met. A case was found where this could be provoked
1716 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1717 components; fortuitously, this particular loophole had already been fixed
1718 by change 4.50/55 above.
1720 If there are any other similar loopholes, the new check in host_aton()
1721 itself should stop them being exploited. The report I received stated that
1722 data on the command line could provoke the exploit when Exim was running as
1723 exim, but did not say which command line option was involved. All I could
1724 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1725 running as the user.
1728 61. There was a buffer overflow vulnerability in the SPA authentication code
1729 (which came originally from the Samba project). I have added a test to the
1730 spa_base64_to_bits() function which I hope fixes it.
1733 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1734 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1736 63. The daemon start-up calls getloadavg() while still root for those OS that
1737 need the first call to be done as root, but it missed one case: when
1738 deliver_queue_load_max is set with deliver_drop_privilege. This is
1739 necessary for the benefit of the queue runner, because there is no re-exec
1740 when deliver_drop_privilege is set.
1742 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1743 This has been fixed.
1745 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1746 was in use, was not putting the data itself into the right store pool;
1747 consequently, it could be overwritten for a subsequent message in the same
1748 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1751 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1753 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1754 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1755 the function were treating the return as a boolean value, which happened to
1756 work because 0=false and not-0=true, but is not correct code.
1758 68. The host_aton() function was not handling scoped IPv6 addresses (those
1759 with, for example, "%eth0" on the end) correctly.
1761 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1762 negated items (that is, ~something) in unsigned ints. Some compilers
1763 apparently mutter when there is no cast.
1765 70. If an address verification called from an ACL failed, and did not produce a
1766 user-specific message (i.e. there was only a "system" message), nothing was
1767 put in $acl_verify_message. In this situation, it now puts the system
1770 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1771 session; change 4.31/43 added the unwanted input to the log line - except
1772 that it did not do this in the start of session case. It now does.
1774 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1775 This is wrong and can cause the other end to generate a synchronization
1776 error if it is another Exim or anything else that does the synchronization
1777 check. A QUIT command is no longer sent after a timeout.
1779 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1780 during host lookups.
1782 74. The defer_ok option of callout verification was not working if it was used
1783 when verifying addresses in header lines, that is, for this case:
1785 verify = header_sender/callout=defer_ok
1787 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1788 those file descriptors could be used for SMTP connections. If anything
1789 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1790 could be sent to the SMTP client, causing chaos. The daemon now opens
1791 stdin, stdout, and stderr to /dev/null when it puts itself into the
1794 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1795 log by default. The user can override this with suitable Perl magic.
1797 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1798 the log message when discard triggers, was not working for the DATA ACL or
1799 for the non-SMTP ACL.
1801 78. Error message wording change in sieve.c.
1803 79. If smtp_accept_max_per_host was set, the number of connections could be
1804 restricted to fewer than expected, because the daemon was trying to set up
1805 a new connection before checking whether the processes handling previous
1806 connections had finished. The check for completed processes is now done
1807 earlier. On busy systems, this bug wouldn't be noticed because something
1808 else would have woken the daemon, and it would have reaped the completed
1811 80. If a message was submitted locally by a user whose login name contained one
1812 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1813 It caused a spool format error. I have fixed the spool reading code. A
1814 related problem was that the "from" clause in the Received: line became
1815 illegal because of the space(s). It is now covered by ${quote_local_part.
1817 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1820 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1822 83. Patch from Sieve maintainer to fix unterminated string problem in
1823 "vacation" handling.
1825 84. Some minor changes to the Linux configuration files to help with other
1826 OS variants using glibc.
1828 85. One more patch for Sieve to update vacation handling to latest spec.
1831 ----------------------------------------------------
1832 See the note above about the 4.44 and 4.50 releases.
1833 ----------------------------------------------------
1839 1. Change 4.43/35 introduced a bug that caused file counts to be
1840 incorrectly computed when quota_filecount was set in an appendfile
1843 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1844 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1847 3. Give more explanation in the error message when the command for a transport
1848 filter fails to execute.
1850 4. There are several places where Exim runs a non-Exim command in a
1851 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1852 was being done only for the command run by the queryprogram router. It is
1853 now done for all such subprocesses. The other cases are: ${run, transport
1854 filters, and the commands run by the lmtp and pipe transports.
1856 5. Some older OS have a limit of 256 on the maximum number of file
1857 descriptors. Exim was using setrlimit() to set 1000 as a large value
1858 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1859 systems. I've change it so that if it can't get 1000, it tries for 256.
1861 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1862 was an oversight, and furthermore, ever since the addition of extra
1863 controls (e.g. 4.43/32), the checks on when to allow different forms of
1864 "control" were broken. There should now be diagnostics for all cases when a
1865 control that does not make sense is encountered.
1867 7. $recipients is now available in the predata ACL (oversight).
1869 8. Tidy the search cache before the fork to do a delivery from a message
1870 received from the command line. Otherwise the child will trigger a lookup
1871 failure and thereby defer the delivery if it tries to use (for example) a
1872 cached ldap connection that the parent has called unbind on.
1874 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1875 of $address_data from the recipient verification was clobbered by the
1876 sender verification.
1878 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1879 was its contents. (It was OK if the option was not defined at all.)
1881 11. A "Completed" log line is now written for messages that are removed from
1882 the spool by the -Mrm option.
1884 12. $host_address is now set to the target address during the checking of
1885 ignore_target_hosts.
1887 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1888 being passed; this would have caused $sender_host_name to have been used if
1889 matching the list had actually called for a host name (not very likely,
1890 since this list is usually IP addresses). A host name is now passed as
1893 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1894 code that specifies a non-restarting handler (typically sigaction() in
1895 modern systems) in an attempt to fix a rare and obscure crash bug.
1897 15. Narrowed the window for a race in the daemon that could cause it to ignore
1898 SIGCHLD signals. This is not a major problem, because they are used only to
1899 wake it up if nothing else does.
1901 16. A malformed maildirsize file could cause Exim to calculate negative values
1902 for the mailbox size or file count. Odd effects could occur as a result.
1903 The maildirsize information is now recalculated if the size or filecount
1906 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1907 support for a long time. Removed HAVE_SYS_VFS_H.
1909 18. Updated exipick to current release from John Jetmore.
1911 19. Allow an empty sender to be matched against a lookup in an address list.
1912 Previously the only cases considered were a regular expression, or an
1915 20. Exim went into a mad DNS lookup loop when doing a callout where the
1916 host was specified on the transport, if the DNS lookup yielded more than
1919 21. The RFC2047 encoding function was originally intended for short strings
1920 such as real names; it was not keeping to the 75-character limit for
1921 encoded words that the RFC imposes. It now respects the limit, and
1922 generates multiple encoded words if necessary. To be on the safe side, I
1923 have increased the buffer size for the ${rfc2047: expansion operator from
1926 22. Failure to deliver a bounce message always caused it to be frozen, even if
1927 there was an errors_to setting on the router. The errors_to setting is now
1930 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1931 canonical form (fully expanded) before being placed in
1932 $sender_host_address.
1934 24. Updated eximstats to version 1.33
1936 25. Include certificate and key file names in error message when GnuTLS fails
1937 to set them up, because the GnuTLS error message doesn't include the name
1938 of the failing file when there is a problem reading it.
1940 26. Expand error message when OpenSSL has problems setting up cert/key files.
1943 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1944 (this can affect the format of dates).
1946 28. exim_tidydb, when checking for the continued existence of a message for
1947 which it has found a message-specific retry record, was not finding
1948 messages that were in split spool directories. Consequently, it was
1949 deleting retry records that should have stayed in existence.
1951 29. eximstats updated to version 1.35
1952 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1953 1.35 - bugfix such that pie charts by volume are generated correctly
1955 30. The SPA authentication driver was not abandoning authentication and moving
1956 on to the next authenticator when an expansion was forced to fail,
1957 contradicting the general specification for all authenticators. Instead it
1958 was generating a temporary error. It now behaves as specified.
1960 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1961 (the order specifies the preference for clients). The order is now AES256,
1962 AES128, 3DES, ARCFOUR128.
1964 31. Small patch to Sieve code - explicitly set From: when generating an
1967 32. Exim crashed if a remote delivery caused a very long error message to be
1968 recorded - for instance if somebody sent an entire SpamAssassin report back
1969 as a large number of 550 error lines. This bug was coincidentally fixed by
1970 increasing the size of one of Exim's internal buffers (big_buffer) that
1971 happened as part of the Exiscan merge. However, to be on the safe side, I
1972 have made the code more robust (and fixed the comments that describe what
1975 33. Some experimental protocols are using DNS PTR records for new purposes. The
1976 keys for these records are domain names, not reversed IP addresses. The
1977 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1978 leaves it alone. Component reversal etc. now happens only for IP addresses.
1981 34. The host_aton() function is supposed to be passed a string that is known
1982 to be a valid IP address. However, in the case of IPv6 addresses, it was
1983 not checking this. This is a hostage to fortune. Exim now panics and dies
1984 if the condition is not met. A case was found where this could be provoked
1985 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1986 components; fortuitously, this particular loophole had already been fixed
1987 by change 4.50/55 or 4.44/33 above.
1989 If there are any other similar loopholes, the new check in host_aton()
1990 itself should stop them being exploited. The report I received stated that
1991 data on the command line could provoke the exploit when Exim was running as
1992 exim, but did not say which command line option was involved. All I could
1993 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1994 running as the user.
1997 35. There was a buffer overflow vulnerability in the SPA authentication code
1998 (which came originally from the Samba project). I have added a test to the
1999 spa_base64_to_bits() function which I hope fixes it.
2002 36. The daemon start-up calls getloadavg() while still root for those OS that
2003 need the first call to be done as root, but it missed one case: when
2004 deliver_queue_load_max is set with deliver_drop_privilege. This is
2005 necessary for the benefit of the queue runner, because there is no re-exec
2006 when deliver_drop_privilege is set.
2008 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2009 was in use, was not putting the data itself into the right store pool;
2010 consequently, it could be overwritten for a subsequent message in the same
2011 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2014 38. Sometimes the final signoff response after QUIT could fail to get
2015 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2016 < 0 before doing a fflush(). This bug looks as though it goes back to the
2017 introduction of TLS in release 3.20, but "sometimes" must have been rare
2018 because the tests only now provoked it.
2024 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2025 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2026 Now it returns an integer. A number of calls were still expecting a T/F
2027 return. Fortuitously, in all cases, the tests worked in OK situations,
2028 which is the norm. However, things would have gone wrong on any write
2029 failures on the smtp file descriptor. This function is used when sending
2030 messages over SMTP and also when doing verify callouts.
2032 2. When Exim is called to do synchronous delivery of a locally submitted
2033 message (the -odf or -odi options), it no longer closes stderr before doing
2036 3. Implemented the mua_wrapper option.
2038 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2040 5. Implemented the functions header_remove(), header_testname(),
2041 header_add_at_position(), and receive_remove_recipient(), and exported them
2044 6. If an ACL "warn" statement specified the addition of headers, Exim already
2045 inserted X-ACL-Warn: at the start if there was no header name. However, it
2046 was not making this test for the second and subsequent header lines if
2047 there were newlines in the string. This meant that an invalid header could
2048 be inserted if Exim was badly configured.
2050 7. Allow an ACL "warn" statement to add header lines at the start or after all
2051 the Received: headers, as well as at the end.
2053 8. Added the rcpt_4xx retry error code.
2055 9. Added postmaster_mailfrom=xxx to callout verification option.
2057 10. Added mailfrom=xxxx to the callout verification option, for verify=
2060 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2061 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2063 12. Inserted some casts to stop certain compilers warning when using pointer
2064 differences as field lengths or precisions in printf-type calls (mostly
2065 affecting debugging statements).
2067 13. Added optional readline() support for -be (dynamically loaded).
2069 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2070 same clock tick as a message's arrival, so that its received time was the
2071 same as the "first fail" time on the retry record, and that message
2072 remained on the queue past the ultimate address timeout, every queue runner
2073 would try a delivery (because it was past the ultimate address timeout) but
2074 after another failure, the ultimate address timeout, which should have then
2075 bounced the address, did not kick in. This was a "< instead of <=" error;
2076 in most cases the first failure would have been in the next clock tick
2077 after the received time, and all would be well.
2079 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2080 being recognized when the domain list was tested by the match_domain
2081 condition in an expansion string.
2083 16. Added the ${str2b64: operator.
2085 17. Exim was always calling setrlimit() to set a large limit for the number of
2086 processes, without checking whether the existing limit was already
2087 adequate. (It did check for the limit on file descriptors.) Furthermore,
2088 errors from getrlimit() and setrlimit() were being ignored. Now they are
2089 logged to the main and panic logs, but Exim does carry on, to try to do its
2090 job under whatever limits there are.
2092 18. Imported PCRE 5.0.
2094 19. Trivial typo in log message " temporarily refused connection" (the leading
2097 20. If the log selector return_path_on_delivery was set and an address was
2098 redirected to /dev/null, the delivery process crashed because it assumed
2099 that a return path would always be set for a "successful" delivery. In this
2100 case, the whole delivery is bypassed as an optimization, and therefore no
2103 21. Internal re-arrangement: the function for sending a challenge and reading
2104 a response while authentication was assuming a zero-terminated challenge
2105 string. It's now changed to take a pointer and a length, to allow for
2106 binary data in such strings.
2108 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2110 23. Exim was not respecting finduser_retries when seeking the login of the
2111 uid under which it was called; it was always trying 10 times. (The default
2112 setting of finduser_retries is zero.) Also, it was sleeping after the final
2113 failure, which is pointless.
2115 24. Implemented tls_on_connect_ports.
2117 25. Implemented acl_smtp_predata.
2119 26. If the domain in control=submission is set empty, Exim assumes that the
2120 authenticated id is a complete email address when it generates From: or
2121 Sender: header lines.
2123 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2124 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2125 chown and chgrp in /bin and hostname in /usr/bin.
2127 28. Exim was keeping the "process log" file open after each use, just as it
2128 does for the main log. This opens the possibility of it remaining open for
2129 long periods when the USR1 signal hits a daemon. Occasional processlog
2130 errors were reported, that could have been caused by this. Anyway, it seems
2131 much more sensible not to leave this file open at all, so that is what now
2134 29. The long-running daemon process does not normally write to the log once it
2135 has entered its main loop, and it closes the log before doing so. This is
2136 so that log files can straightforwardly be renamed and moved. However,
2137 there are a couple of unusual error situations where the daemon does write
2138 log entries, and I had neglected to close the log afterwards.
2140 30. The text of an SMTP error response that was received during a remote
2141 delivery was being truncated at 512 bytes. This is too short for some of
2142 the long messages that one sometimes sees. I've increased the limit to
2145 31. It is now possible to make retry rules that apply only when a message has a
2146 specific sender, in particular, an empty sender.
2148 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2149 it possible to be selective about when SMTP synchronization is enforced.
2151 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2153 32. Implemented hosts_connection_nolog.
2155 33. Added an ACL for QUIT.
2157 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2160 35. Added mailbox_size and mailbox_filecount to appendfile.
2162 36. Added control = no_multiline_responses to ACLs.
2164 37. There was a bug in the logic of the code that waits for the clock to tick
2165 in the case where the clock went backwards by a substantial amount such
2166 that the microsecond fraction of "now" was more than the microsecond
2167 fraction of "then" (but the whole seconds number was less).
2169 38. Added support for the libradius Radius client library this is found on
2170 FreeBSD (previously only the radiusclient library was supported).
2176 1. When certain lookups returned multiple values in the form name=value, the
2177 quoting of the values was not always being done properly. Specifically:
2178 (a) If the value started with a double quote, but contained no whitespace,
2180 (b) If the value contained whitespace other than a space character (i.e.
2181 tabs or newlines or carriage returns) it was not quoted.
2182 This fix has been applied to the mysql and pgsql lookups by writing a
2183 separate quoting function and calling it from the lookup code. The fix
2184 should probably also be applied to nisplus, ibase and oracle lookups, but
2185 since I cannot test any of those, I have not disturbed their existing code.
2187 2. A hit in the callout cache for a specific address caused a log line with no
2188 reason for rejecting RCPT. Now it says "Previous (cached) callout
2189 verification failure".
2191 3. There was an off-by-one bug in the queryprogram router. An over-long
2192 return line was truncated at 256 instead of 255 characters, thereby
2193 overflowing its buffer with the terminating zero. As well as fixing this, I
2194 have increased the buffer size to 1024 (and made a note to document this).
2196 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2197 when Exim is waiting for an SMTP response from a remote server, Exim
2198 restarts its select() call on the socket, thereby resetting its timeout.
2199 This is not a problem when such interrupts are rare. Somebody set up a cron
2200 job to run exiwhat every 2 minutes, which is less than the normal select()
2201 timeout (5 or 10 minutes). This meant that the select() timeout never
2202 kicked in because it was always reset. I have fixed this by comparing the
2203 time when an interrupt arrives with the time at the start of the first call
2204 to select(). If more time than the timeout has elapsed, the interrupt is
2205 treated as a timeout.
2207 5. Some internal re-factoring in preparation for the addition of Sieve
2208 extensions (by MH). In particular, the "personal" test is moved to a
2209 separate function, and given an option for scanning Cc: and Bcc: (which is
2210 not set for Exim filters).
2212 6. When Exim created an email address using the login of the caller as the
2213 local part (e.g. when creating a From: or Sender: header line), it was not
2214 quoting the local part when it contained special characters such as @.
2216 7. Installed new OpenBSD configuration files.
2218 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2219 try to make them clearer.
2221 9. Callout options, other than the timeout value, were being ignored when
2222 verifying sender addresses in header lines. For example, when using
2224 verify = header_sender/callout=no_cache
2226 the cache was (incorrectly) being used.
2228 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2229 only the Cygwin environment.
2231 11. When return_path_on_delivery was set as a log selector, if different remote
2232 addresses in the same message used different return paths and parallel
2233 remote delivery occurred, the wrong values would sometimes be logged.
2234 (Whenever a remote delivery process finished, the return path value from
2235 the most recently started remote delivery process was logged.)
2237 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2238 lines when AUTH and/or TLS are in use. This is the "received protocol"
2239 field. Exim used to use "asmtp" for authenticated SMTP, without any
2240 indication (in the protocol name) for TLS use. Now it follows the RFC and
2241 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2242 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2243 names appear in log lines as well as in Received: header lines.
2245 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2246 extensions, and comparison tests, and to fix some bugs.
2248 14. Changes to the "personal" filter test:
2250 (1) The test was buggy in that it was just doing the equivalent of
2251 "contains" tests on header lines. For example, if a user's address was
2252 anne@some.where, the "personal" test would incorrectly be true for
2254 To: susanne@some.where
2256 This test is now done by extracting each address from the header in turn,
2257 and checking the entire address. Other tests that are part of "personal"
2258 are now done using regular expressions (for example, to check local parts
2259 of addresses in From: header lines).
2261 (2) The list of non-personal local parts in From: addresses has been
2262 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2263 taken from the Sieve specification recommendations.
2265 (3) If the message contains any header line starting with "List-" it is
2266 treated as non-personal.
2268 (4) The test for "circular" in the Subject: header line has been removed
2269 because it now seems ill-conceived.
2271 15. Minor typos in src/EDITME comments corrected.
2273 16. Installed latest exipick from John Jetmore.
2275 17. If headers_add on a router specified a text string that was too long for
2276 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2277 of string_sprintf() is now avoided.
2279 18. $message_body_size was not set (it was always zero) when running the DATA
2280 ACL and the local_scan() function.
2282 19. For the "mail" command in an Exim filter, no default was being set for
2283 the once_repeat time, causing a random time value to be used if "once" was
2284 specified. (If the value happened to be <= 0, no repeat happened.) The
2285 default is now 0s, meaning "never repeat". The "vacation" command was OK
2286 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2287 (I found it when inspecting the code).
2289 20. There is now an overall timeout for performing a callout verification. It
2290 defaults to 4 times the callout timeout, which applies to individual SMTP
2291 commands during the callout. The overall timeout applies when there is more
2292 than one host that can be tried. The timeout is checked before trying the
2293 next host. This prevents very long delays if there are a large number of
2294 hosts and all are timing out (e.g. when the network connections are timing
2295 out). The value of the overall timeout can be changed by specifying an
2296 additional sub-option for "callout", called "maxwait". For example:
2298 verify = sender/callout=5s,maxwait=20s
2300 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2301 to the end before writing, but this should make it even safer).
2303 22. Exim was forgetting that it had advertised PIPELINING for the second and
2304 subsequent messages on an SMTP connection. It was also not resetting its
2305 memory on STARTTLS and an internal HELO.
2307 23. When Exim logs an SMTP synchronization error within a session, it now
2308 records whether PIPELINING has been advertised or not.
2310 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2311 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2312 rather than long int.
2314 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2316 26. Added the never_mail option to autoreply.
2322 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2323 crash if the getsockname() call failed; this can happen if a connection is
2324 closed very soon after it is established. The problem was simply in the
2325 order in which certain operations were done, causing Exim to try to write
2326 to the SMTP stream before it had set up the file descriptor. The bug has
2327 been fixed by making things happen in the correct order.
2333 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2334 before the connection was closed, thus losing the rejection response.
2336 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2337 some early Solaris releases, but causes trouble in current releases where
2338 socklen_t is defined.
2340 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2343 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2346 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2347 that releasing the top of it at the end releases what was used for sub-
2348 expansions (unless the block got too big). However, discard this block if
2349 the first thing is a variable or header, so that we can use its block when
2350 it is dynamic (useful for very large $message_headers, for example).
2352 6. Lookups now cache *every* query, not just the most recent. A new, separate
2353 store pool is used for this. It can be recovered when all lookup caches are
2354 flushed. Lookups now release memory at the end of their result strings.
2355 This has involved some general refactoring of the lookup sources.
2357 7. Some code has been added to the store_xxx() functions to reduce the amount
2358 of flapping under certain conditions.
2360 8. log_incoming_interface used to affect only the <= reception log lines. Now
2361 it causes the local interface and port to be added to several more SMTP log
2362 lines, for example "SMTP connection from", and rejection lines.
2364 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2366 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2368 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2369 could be overwritten at the end of the current message (or the start of a
2370 new message if it was set in a HELO ACL). The value is now preserved for
2371 the duration of the SMTP connection.
2373 12. If a transport had a headers_rewrite setting, and a matching header line
2374 contained an unqualified address, that address was qualified, even if it
2375 did not match any rewriting rules. The underlying bug was that the values
2376 of the flags that permit the existence of unqualified sender and recipient
2377 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2378 non-local messages, and by -bnq for local messages) were not being
2379 preserved with the message after it was received.
2381 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2382 "next input=" as part of the text comprising the host identity instead of
2383 the correct text. The code was using the same buffer for two different
2384 strings. However, depending on which order the printing function evaluated
2385 its arguments, the bug did not always show up. Under Linux, for example, my
2386 test suite worked just fine.
2388 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2389 doesn't work with some older versions of Perl. It has been changed to "my",
2390 which in any case is probably the better facility to use.
2392 15. A really picky compiler found some instances of statements for creating
2393 error messages that either had too many or two few arguments for the format
2396 16. The size of the buffer for calls to the DNS resolver has been increased
2397 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2398 for addresses that have a lot of PTR records. This alleviates a problem; it
2399 does not fully solve it.
2401 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2402 buffer now truncates the list and logs the incident, which is the same
2403 action as happens when Exim is looking up a host name and its aliases.
2404 Previously in this situation something unpredictable would happen;
2405 sometimes it was "internal error: store_reset failed".
2407 18. If a server dropped the connection unexpectedly when an Exim client was
2408 using GnuTLS and trying to read a response, the client delivery process
2409 crashed while trying to generate an error log message.
2411 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2412 string, for example:
2414 warn message = H1: something\nH2: something
2416 the text was added as a single header line from Exim's point of view
2417 though it ended up OK in the delivered message. However, searching for the
2418 second and subsequent header lines using $h_h2: did not work. This has been
2419 fixed. Similarly, if a system filter added multiple headers in this way,
2420 the routers could not see them.
2422 20. Expanded the error message when iplsearch is called with an invalid key to
2423 suggest using net-iplsearch in a host list.
2425 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2426 ACLs are no longer actually imposed (and a message to that effect is
2429 22. If a "gecos" field in a passwd entry contained escaped characters, in
2430 particular, if it contained a \" sequence, Exim got it wrong when building
2431 a From: or a Sender: header from that name. A second bug also caused
2432 incorrect handling when an unquoted " was present following a character
2433 that needed quoting.
2435 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2436 was not being matched caselessly.
2438 24. Arranged for all hyphens in the exim.8 source to be escaped with
2441 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2442 properly. Recipient callout cache records were still being keyed to include
2443 the sender, even when use_sender was set false. This led to far more
2444 callouts that were necessary. The sender is no longer included in the key
2445 when use_sender is false.
2447 26. Added "control = submission" modifier to ACLs.
2449 27. Added the ${base62d: operator to decode base 62 numbers.
2451 28. dnsdb lookups can now access SRV records.
2453 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2454 the configuration file.
2456 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2457 (-v) mode. This makes the output for a verbose queue run more intelligible.
2459 31. Added a use_postmaster feature to recipient callouts.
2461 32. Added the $body_zerocount variable, containing the number of binary zero
2462 bytes in the message body.
2464 33. The time of last modification of the "new" subdirectory is now used as the
2465 "mailbox time last read" when there is a quota error for a maildir
2468 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2470 35. Added +ignore_unknown as a special item in host lists.
2472 36. Code for decoding IPv6 addresses in host lists is now included, even if
2473 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2474 address was recognized as an IP address, but was then not correctly decoded
2475 into binary, causing unexpected and incorrect effects when compared with
2482 1. Very minor rewording of debugging text in manualroute to say "list of
2483 hosts" instead of "hostlist".
2485 2. If verify=header_syntax was set, and a header line with an unqualified
2486 address (no domain) and a large number of spaces between the end of the
2487 name and the colon was received, the reception process suffered a buffer
2488 overflow, and (when I tested it) crashed. This was caused by some obsolete
2489 code that should have been removed. The fix is to remove it!
2491 3. When running in the test harness, delay a bit after writing a bounce
2492 message to get a bit more predictability in the log output.
2494 4. Added a call to search_tidyup() just before forking a reception process. In
2495 theory, someone could use a lookup in the expansion of smtp_accept_max_
2496 per_host which, without the tidyup, could leave open a database connection.
2498 5. Added the variables $recipient_data and $sender_data which get set from a
2499 lookup success in an ACL "recipients" or "senders" condition, or a router
2500 "senders" option, similar to $domain_data and $local_part_data.
2502 6. Moved the writing of debug_print from before to after the "senders" test
2505 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2506 problems for message scanning, either using a data ACL, or using
2507 local_scan() because the Received: header was not generated till after they
2508 were called (in order to set the time as the time of reception completion).
2509 I have revised the way this works. The header is now generated after the
2510 body is received, but before the ACL or local_scan() are called. After they
2511 are run, the timestamp in the header is updated.
2517 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2518 before starting a queue runner without re-exec. This happened only when
2519 deliver_drop_privilege was set or when the Exim user was set to root. The
2520 effect of the bug was that timeouts during subsequent deliveries caused
2521 crashes instead of being properly handled. The handler is now left at its
2522 default (and expected) setting.
2524 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2525 message, again when deliver_drop_privilege is set or Exim is run as root.
2526 The bug described in (1) was not present in this case, but the tidying up
2527 of the other signals was missing. I have made the two cases consistent.
2529 3. The ignore_target_hosts setting on a manualroute router was being ignored
2530 for hosts that were looked up using the /MX notation.
2532 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2535 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2536 operated on the sender address. After changing the $sender_address to <>
2537 for the sender address verify, Exim was re-instated it as the original
2538 (before rewriting) address, but remembering that it had rewritten it, so it
2539 wasn't rewriting it again. This bug also had the effect of breaking the
2540 sender address verification caching when the sender address was rewritten.
2542 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2543 This has been changed so that if the ip literal address matches
2544 ignore_target_hosts, the router declines.
2546 7. Added expansion conditions match_domain, match_address, and match_local_
2547 part (NOT match_host).
2549 8. The placeholder for the Received: header didn't have a length field set.
2551 9. Added code to Exim itself and to exim_lock to test for a specific race
2552 condition that could lead to file corruption when using MBX delivery. The
2553 issue is with the lockfile that is created in /tmp. If this file is removed
2554 after a process has opened it but before that process has acquired a lock,
2555 there is the potential for a second process to recreate the file and also
2556 acquire a lock. This could lead to two Exim processes writing to the file
2557 at the same time. The added code performs the same test as UW imapd; it
2558 checks after acquiring the lock that its file descriptor still refers to
2559 the same named file.
2561 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2562 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2565 11. Added the smtp_active_hostname option. If used, this will typically be made
2566 to depend on the incoming interface address. Because $interface_address is
2567 not set up until the daemon has forked a reception process, error responses
2568 that can happen earlier (such as "too many connections") no longer contain
2571 12. If an expansion in a condition on a "warn" statement fails because a lookup
2572 defers, the "warn" statement is abandoned, and the next ACL statement is
2573 processed. Previously this caused the whole ACL to be aborted.
2575 13. Added the iplsearch lookup type.
2577 14. Added ident_timeout as a log selector.
2579 15. Added tls_certificate_verified as a log selector.
2581 16. Added a global option tls_require_ciphers (compare the smtp transport
2582 option of the same name). This controls incoming TLS connections.
2584 17. I finally figured out how to make tls_require_ciphers do a similar thing
2585 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2586 before starting the TLS session.
2588 18. Tabs are now shown as \t in -bP output.
2590 19. If the log selector return_path_on_delivery was set, Exim crashed when
2591 bouncing a message because it had too many Received: header lines.
2593 20. If two routers both had headers_remove settings, and the first one included
2594 a superfluous trailing colon, the final name in the first list and the
2595 first name in the second list were incorrectly joined into one item (with a
2596 colon in the middle).
2602 1. Added -C and -D options to the exinext utility, mainly to make it easier
2603 to include in the automated testing, but these could be helpful when
2604 multiple configurations are in use.
2606 2. The exinext utility was not formatting the output nicely when there was
2607 an alternate port involved in the retry record key, nor when there was a
2608 message id as well (for retries that were specific to a specific message
2609 and a specific host). It was also confused by IPv6 addresses, because of
2610 the additional colons they contain. I have fixed the IPv4 problem, and
2611 patched it up to do a reasonable job for IPv6.
2613 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2614 delivery, the log line now contains "pipelined" if PIPELINING was used.
2616 4. An SMTP transport process used to panic and die if the bind() call to set
2617 an explicit outgoing interface failed. This has been changed; it is now
2618 treated in the same way as a connect() failure.
2620 5. A reference to $sender_host_name in the part of a conditional expansion
2621 that was being skipped was still causing a DNS lookup. This no longer
2624 6. The def: expansion condition was not recognizing references to header lines
2625 that used bh_ and bheader_.
2627 7. Added the _cache feature to named lists.
2629 8. The code for checking quota_filecount in the appendfile transport was
2630 allowing one more file than it should have been.
2632 9. For compatibility with Sendmail, the command line option
2640 and sets the incoming protocol and host name (for trusted callers). The
2641 host name and its colon can be omitted when only the protocol is to be set.
2642 Note the Exim already has two private options, -pd and -ps, that refer to
2643 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2644 "s", but I don't think that's a major issue.
2646 10. A number of refactoring changes to the code, none of which should affect
2649 (a) The number of logging options was getting close to filling up the
2650 32-bit word that was used as a bit map. I have split them into two classes:
2651 those that are passed in the argument to log_write(), and those that are
2652 only ever tested independently outside of that function. These are now in
2653 separate 32-bit words, so there is plenty of room for expansion again.
2654 There is no change in the user interface or the logging behaviour.
2656 (b) When building, for example, log lines, the code previously used a
2657 macro that called string_cat() twice, in order to add two strings. This is
2658 not really sufficiently general. Furthermore, there was one instance where
2659 it was actually wrong because one of the argument was used twice, and in
2660 one call a function was used. (As it happened, calling the function twice
2661 did not affect the overall behaviour.) The macro has been replaced by a
2662 function that can join an arbitrary number of extra strings onto a growing
2665 (c) The code for expansion conditions now uses a table and a binary chop
2666 instead of a serial search (which was left over from when there were very
2667 few conditions). Also, it now recognizes conditions like "pam" even when
2668 the relevant support is not compiled in: a suitably worded error message is
2669 given if an attempt is made to use such a condition.
2671 11. Added ${time_interval:xxxxx}.
2673 12. A bug was causing one of the ddress fields not to be passed back correctly
2674 from remote delivery subprocesses. The field in question was not being
2675 subsequently used, so this caused to problems in practice.
2677 13. Added new log selectors queue_time and deliver_time.
2679 14. Might have fixed a bug in maildirsizefile handling that threw up
2680 "unexpected character" debug warnings, and recalculated the data
2681 unnecessarily. In any case, I expanded the warning message to give more
2684 15. Added the message "Restricted characters in address" to the statements in
2685 the default ACL that block characters like @ and % in local parts.
2687 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2688 Three changes have been made:
2690 (a) There was a serious bug; a negative response to MAIL caused the whole
2691 recipient domain to be cached as invalid, thereby blocking all messages
2692 to all local parts at the same domain, from all senders. This bug has
2693 been fixed. The domain is no longer cached after a negative response to
2694 MAIL if the sender used is not empty.
2696 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2699 (c) A new callout option, "use_sender" has been added for people who want
2700 the modified behaviour.
2706 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2709 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2710 indeed breaks things for older releases.
2712 3. Added additional logging to the case where there is a problem reading data
2713 from a filter that is running in a subprocess using a pipe, in order to
2714 try to track down a specific problem.
2716 4. Testing facility fudge: when running in the test harness and attempting
2717 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2718 getting "No route to host". Convert this to a timeout.
2720 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2723 6. Some OS don't have socklen_t but use size_t instead. This affects the
2724 fifth argument of getsockopt() amongst other things. This is now
2725 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2726 can be set for individual OS. I have set it for SunOS5, OSF1, and
2727 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2728 some earlier ones do not.
2730 7. Change 4.30/15 was not doing the test caselessly.
2732 8. The standard form for an IPv6 address literal was being rejected by address
2733 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2734 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2735 this, as well as the form without the "IPv6" on the front (but only when
2736 address literals are enabled, of course).
2738 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2740 10. Exim crashed if a message with an empty sender address specified by -f
2741 encountered a router with an errors_to setting. This could be provoked only
2742 by a command such as
2746 where an empty string was supplied; "<>" did not hit this bug.
2748 11. Installed PCRE release 4.5.
2750 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2751 remained set. It is now erased.
2753 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2754 times from message ids (which are base 36 rather than the normal 62).
2756 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2757 were being counted as actual protocol errors, and logged if the log
2758 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2759 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2760 and DATA following a set of rejected RCPTs do not count as protocol errors.
2761 In other words, Exim assumes they were pipelined, though this may not
2762 actually be the case. Of course, in all cases the client gets an
2763 appropriate error code.
2765 15. If a lookup fails in an ACL condition, a message about the failure may
2766 be available; it is used if testing the ACL cannot continue, because most
2767 such messages specify what the cause of the deferral is. However, some
2768 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2769 that caused an old message to be retained and used if a later statement
2770 caused a defer, replacing the real cause of the deferral.
2772 16. If an IP address had so many PTR records that the DNS lookup buffer
2773 was not large enough to hold them, Exim could crash while trying to process
2774 the truncated data. It now detects and logs this case.
2776 17. Further to 4.21/58, another change has been made: if (and only if) the
2777 first line of a message (the first header line) ends with CRLF, a bare LF
2778 in a subsequent header line has a space inserted after it, so as not to
2779 terminate the header.
2781 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2782 unnecessarily, used atoi() instead of strtol(), and didn't check the
2783 termination when getting file sizes from file names by regex.
2785 19. Completely re-implemented the support for maildirsize files, in the light
2786 of a number of problems with the previous contributed implementation
2787 (4.30/29). In particular:
2789 . If the quota is zero, the maildirsize file is maintained, but no quota is
2792 . If the maildir directory does not exist, it is created before any attempt
2793 to write a maildirsize file.
2795 . The quota value in the file is just a cache; if the quota is changed in
2796 the transport, the new value overrides.
2798 . A regular expression is available for excluding directories from the
2801 20. The autoreply transport checks the characters in options that define the
2802 message's headers; it allows continued headers, but it was checking with
2803 isspace() after an embedded newline instead of explicitly looking for a
2806 21. If all the "regular" hosts to which an address was routed had passed their
2807 expiry times, and had not reached their retry times, the address was
2808 bounced, even if fallback hosts were defined. Now Exim should go on to try
2811 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2812 equivalent code in the SMTP transport. Some hosts send humungous responses
2813 to HELO/EHLO, more than 1024 it seems.
2815 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2816 gives compiler warnings in some environments. I've now done it "properly",
2819 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2820 (because of problems with the built-in one) was declared to return uschar *
2821 instead of char *, causing compiler failure.
2823 25. Fixed a file descriptor leak when processing alias/forward files.
2825 26. Fixed a minor format string issue in dbfn.c.
2827 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2829 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2830 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2833 29. When a sender address is verified, it is cached, to save repeating the test
2834 when there is more than one recipient in a message. However, when the
2835 verification involves a callout, it is possible for different callout
2836 options to be set for different recipients. It is too complicated to keep
2837 track of this in the cache, so now Exim always runs a verification when a
2838 callout is required, relying on the callout cache for the optimization.
2839 The overhead is duplication of the address routing, but this should not be
2842 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2843 to be verified with callout=postmaster, and the main callout worked but the
2844 postmaster check failed, the verification correctly failed. However, if a
2845 subsequent RCPT command asked for sender verification *without* the
2846 postmaster check, incorrect caching caused this verification also to fail,
2849 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2850 it was not caching the DNS options (qualify_single, search_parents) that
2851 were used when the lookup failed. A subsequent lookup with different
2852 options therefore always gave the same answer, though there were cases
2853 where it should not have. (Example: a "domains = !$mx_any" option on a
2854 dnslookup router: the "domains" option is always processed without any
2855 widening, but the router might have qualify_single set.) Now Exim uses the
2856 cached value only when the same options are set.
2858 32. Added John Jetmore's "exipick" utility to the distribution.
2860 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2861 than a timeout (e.g. a certificate is required, and is not provided), an
2862 Exim server now closes the connection immediately. Previously it waited for
2863 the client to close - but if the client is SSL, it seems that they each
2864 wait for each other, leading to a delay before one of them times out.
2866 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2867 maintained 0.8.x compatibility because I don't think many are using it, and
2868 it is clearly obsolete.
2870 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2873 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2874 client certificate was expired. A simple patch fixes this, though I don't
2875 understand the full logic of why the verify callback is called multiple
2878 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2879 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2880 which causes problems with some clients (such as the Certicom SSL Plus
2881 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2882 disables the coutermeasure allowing Eudora to connect."
2884 38. Exim was not checking that a write() to a log file succeeded. This could
2885 lead to Bad Things if a log got too big, in particular if it hit a file
2886 size limit. Exim now panics and dies if it cannot write to a log file, just
2887 as it does if it cannot open a log file.
2889 39. Modified OS/Makefile-Linux so that it now contains
2891 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2893 The two -D definitions ensure that Exim is compiled with large file
2894 support, which makes it possible to handle log files that are bigger than
2897 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2898 instance) a domain was checked against a named list that involved a lookup,
2899 causing $domain_data to be set, then another domain was checked against the
2900 same list, then the first domain was re-checked, the value of $domain_data
2901 after the final check could be wrong. In particular, if the second check
2902 failed, it could be set empty. This bug probably also applied to
2905 41. The strip_trailing_dot option was not being applied to the address given
2906 with the -f command-line option.
2908 42. The code for reading a message's header from the spool was incrementing
2909 $received_count, but never initializing it. This meant that the value was
2910 incorrect (doubled) while delivering a message in the same process in which
2911 it was received. In the most common configuration of Exim, this never
2912 happens - a fresh exec is done - but it can happen when
2913 deliver_drop_privilege is set.
2915 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2916 it now includes up to 150 characters of the unexpected data in the log
2919 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2920 and building data strings. The size of both of these buffers was 10 000
2921 bytes - far larger than anybody would *ever* want, thought I. Needless to
2922 say, somebody hit the limit. I have increased the maximum line length to
2923 20 000 and the maximum data length of concatenated lines to 100 000. I have
2924 also fixed two bugs, because there was no checking on these buffers. Tsk,
2925 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2926 buffer is too small.
2928 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2929 lsearch lookups. Now it does.
2931 46. When parsing a route_list item in a manualroute router, a fixed-length
2932 buffer was used for the list of hosts. I made this 1024 bytes long,
2933 thinking that nobody would ever have a list of hosts that long. Wrong.
2934 Somebody had a whole pile of complicated expansion conditions, and the
2935 string was silently truncated, leading to an expansion error. It turns out
2936 that it is easier to change to an unlimited length (owing to other changes
2937 that have happened since this code was originally written) than to build
2938 structure for giving a limitation error. The length of the item that
2939 expands into the list of hosts is now unlimited.
2941 47. The lsearch lookup could not handle data where the length of text line was
2942 more than 4095 characters. Such lines were truncated, leading to shortened
2943 data being returned. It should now handle lines of any length.
2945 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2946 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2949 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2950 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2951 no longer happens in comment lines. A list of replacements is now placed
2952 at the head of all of the source files, except those whose only change is
2953 to replace PERL_COMMAND in the very first #! line.
2955 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2956 messages on the queue, with a bottom-up merge sort, using code contributed
2957 by Michael Haardt. This should make operations like -bp somewhat faster on
2958 large queues. It won't affect queue runners, except when queue_run_in_order
2961 51. Installed eximstats 1.31 in the distribution.
2963 52. Added support for SRV lookups to the dnslookup router.
2965 53. If an ACL referred to $message_body or $message_body_end, the value was not
2966 reset for any messages that followed in the same SMTP session.
2968 54. The store-handling optimization for building very long strings was not
2969 differentiating between the different store pools. I don't think this
2970 actually made any difference in practice, but I've tidied it.
2972 55. While running the routers to verify a sender address, $sender_address
2973 was still set to the sender address. This is wrong, because when routing to
2974 send a bounce to the sender, it would be empty. Therefore, I have changed
2975 it so that, while verifying a sender address, $sender_address is set to <>.
2976 (There is no change to what happens when verifying a recipient address.)
2978 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2979 target A or AAAA records (if not already returned) without resetting the
2980 qualify_single or search_parents options of the DNS resolver. These are
2981 inappropriate in this case because the targets of MX and SRV records must
2982 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2983 target that, when qualified, matched something in the local domain. These
2984 two options are now turned off when doing these lookups.
2986 57. It seems that at least some releases of Reiserfs (which does not have the
2987 concept of a fixed number of inodes) returns zero and not -1 for the
2988 number of available inodes. This interacted badly with check_spool_inodes,
2989 which assumed that -1 was the "no such thing" setting. What I have done is
2990 to check that the total number of inodes is greater than zero before doing
2991 the test of how many are available.
2993 58. When a "warn" ACL statement has a log_message modifier, the message is
2994 remembered, and not repeated. This is to avoid a lot of repetition when a
2995 message has many recipients that cause the same warning to be written.
2996 Howewer, Exim was preserving the list of already written lines for an
2997 entire SMTP session, which doesn't seem right. The memory is now reset if a
2998 new message is started.
3000 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3001 debugging output unless log_rewrite was also set.
3003 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3004 of (int)(handle) when we know that handle contains (void *)(-1).
3006 61. The Exim daemon panic-logs an error return when it closes the incoming
3007 connection. However "connection reset by peer" seems to be common, and
3008 isn't really an error worthy of noting specially, so that particular error
3011 62. When Exim is trying to find all the local interfaces, it used to panic and
3012 die if the ioctl to get the interface flags failed. However, it seems that
3013 on at least one OS (Solaris 9) it is possible to have an interface that is
3014 included in the list of interfaces, but for which you get a failure error
3015 for this call. This happens when the interface is not "plumbed" into a
3016 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3017 failure of the "get flags" call assumes that the interface is down.
3019 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3020 makes life easier for people who are doing arithmetic on fields extracted
3021 from dates, where you often get leading zeros that should not be
3022 interpreted as octal.
3024 64. Added qualify_domain to the redirect router, to override the global
3027 65. If a pathologically long header line contained very many addresses (the
3028 report of this problem mentioned 10 000) and each of them was rewritten,
3029 Exim could use up a very large amount of memory. (It kept on making new
3030 copies of the header line as it rewrote, and never released the old ones.)
3031 At the expense of a bit more processing, the header rewriting function has
3032 been changed so that it no longer eats memory in this way.
3034 66. The generation of the Received: header has been moved from the time that a
3035 message starts to be received, to the time that it finishes. The timestamp
3036 in the Received: header should now be very close to that of the <= log
3037 line. There are two side-effects of this change:
3039 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3040 logged header lines no longer include the local Received: line, because
3041 it has not yet been created. The same applies to a copy of the message
3042 that is returned to a non-SMTP sender when a message is rejected.
3044 (b) When a filter file is tested using -bf, no additional Received: header
3045 is added to the test message. After some thought, I decided that this
3048 This change does not affect the value of $received_for. It is still set
3049 after address rewriting, but before local_scan() is called.
3051 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3053 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3054 gave an unhelpful panic error message, and a defer error. I have managed to
3055 change this behaviour so that it now rejects any supplied certificate,
3056 which seems right, as the list of acceptable certificates is empty.
3058 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3059 gave an unhelpful defer error. I have not managed to make this reject any
3060 supplied certificates, but the error message it gives is "no certificate
3061 supplied", which is not helpful.
3063 70. exigrep's output now also includes lines that are not associated with any
3064 message, but which match the given pattern. Implemented by a patch from
3065 Martin Sluka, which also tidied up the Perl a bit.
3067 71. Recipient callout verification, like sender verification, was using <> in
3068 the MAIL FROM command. This isn't really the right thing, since the actual
3069 sender may affect whether the remote host accepts the recipient or not. I
3070 have changed it to use the actual sender in the callout; this means that
3071 the cache record is now keyed on a recipient/sender pair, not just the
3072 recipient address. There doesn't seem to be a real danger of callout loops,
3073 since a callout by the remote host to check the sender would use <>.
3074 [SEE ABOVE: changed after hitting problems.]
3076 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3077 temporary errors. However, in the case of such a code being given after
3078 the end of a data transmission (i.e. after ".") Exim was failing to write
3079 a retry record for the message. (Yes, there was some broken host that was
3080 actually sending 8xx at this point.)
3082 73. An unknown lookup type in a host list could cause Exim to panic-die when
3083 the list was checked. (An example that provoked this was putting <; in the
3084 middle of a list instead of at the start.) If this happened during a DATA
3085 ACL check, a -D file could be left lying around. This kind of configuration
3086 error no longer causes Exim to die; instead it causes a defer errror. The
3087 incident is still logged to the main and panic logs.
3089 74. Buglet left over from Exim 3 conversion. The message "too many messages
3090 in one connection" was written to the rejectlog but not the mainlog, except
3091 when address rewriting (yes!) was being logged.
3093 75. Added write_rejectlog option.
3095 76. When a system filter was run not as root (that is, when system_filter_user
3096 was set), the values of the $n variables were not being returned to the
3097 main process; thus, they were not subsequently available in the $sn
3100 77. Added +return_path_on_delivery log selector.
3102 78. A connection timeout was being treated differently from recipients deferred
3103 when testing hosts_max_try with a message that was older than the host's
3104 retry timeout. (The host should not be counted, thus allowing all hosts to
3105 be tried at least once before bouncing.) This may have been the cause of an
3106 occasionally reported bug whereby a message would remain on the queue
3107 longer than the retry timeout, but would be bounced if a delivery was
3108 forced. I say "may" because I never totally pinned down the problem;
3109 setting up timeout/retry tests is difficult. See also the next item.
3111 79. The ultimate address timeout was not being applied to errors that involved
3112 a combination of host plus message (for example, a timeout on a MAIL
3113 command). When an address resolved to a number of possible hosts, and they
3114 were not all tried for each delivery (e.g. because of hosts_max_try), a
3115 message could remain on the queue longer than the retry timeout.
3117 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3120 81. Fixed an obscure SMTP outgoing bug which required at least the following
3121 conditions: (a) there was another message waiting for the same server;
3122 (b) the server returned 5xx to all RCPT commands in the first message so
3123 that the message was not completed; (c) the server dropped the connection
3124 or gave a negative response to the RSET that Exim sends to abort the
3125 transaction. The observed case was a dropped connection after DATA that had
3126 been sent in pipelining mode. That is, the server had advertised PIPELINING
3127 but was not implementing it correctly. The effect of the bug was incorrect
3128 behaviour, such as trying another host, and this could lead to a crash.
3134 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3135 and daemon.c were passed as pointers to ints; they should have been
3136 pointers to socklen_t variables (which are typically unsigned ints).
3138 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3141 3. Fixed a really odd bug that affected only the testing scheme; patching a
3142 certain fixed string in the binary changed the value of another string that
3143 happened to be identical to the end of the original first string.
3145 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3146 name", it returns that address as the IP address. On some operating
3147 systems (e.g. Solaris), it also passes back the IP address string as the
3148 "host name". However, on others (e.g. Linux), it passes back an empty
3149 string. Exim wasn't checking for this, and was changing the host name to an
3150 empty string, assuming it had been canonicized.
3152 5. Although rare, it is permitted to have more than one PTR record for a given
3153 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3154 all the names associated with an address, because they do in Solaris.
3155 However, it seems that they do not in Linux for data that comes from the
3156 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3157 I found this out when I moved to a new Linux workstation and tried to run
3158 the Exim test suite.
3160 To get round this problem I have changed the code so that it now does its
3161 own call to the DNS to look up PTR records when searching for a host name.
3162 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3163 addresses that are only in /etc/hosts are still found.
3165 This behaviour is, however, controlled by an option called host_lookup_
3166 order, which defaults to "bydns:byaddr". If people want to use the other
3167 order, or indeed, just use one or the other means of lookup, they can
3168 specify it in this variable.
3170 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3171 some operating systems, this comes back from gethostbyaddr() as an empty
3172 string, and this is what Exim used to test for. However, it seems that in
3173 other systems, "." is yielded. Exim now tests for this case too.
3175 7. The values of check_spool_space and check_log_space are now held internally
3176 as a number of kilobytes instead of an absolute number of bytes. If a
3177 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3178 kilobyte. This means that much larger values can be stored.
3180 8. Exim monitor: an attempt to get the action menu when not actually pointing
3181 at a message produces an empty menu entitled "No message selected". This
3182 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3183 no entries in it ("Shell widget menu has zero width and/or height"). So I
3184 have added a single, blank menu entry in this case.
3186 9. Added ${quote_local_part.
3188 10. MIME decoding is now applied to the contents of Subject: header lines when
3191 11. Now that a reference to $sender_host_address automatically causes a reverse
3192 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3193 host lookup before query-style lookups in lists that might use this
3194 variable. This has therefore been abolished, and the "net-" prefix is no
3195 longer necessary for query-style lookups.
3197 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3198 have been a typo for LFLAGS, so it has been changed.
3200 13. The install script calls Exim with "-C /dev/null" in order to find the
3201 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3202 to be output. Howeve, since Exim outputs its version number before the
3203 error, it didn't break the script. It just looked ugly. I fixed this by
3204 always allowing "-C /dev/null" if the caller is root.
3206 14. Ignore overlarge ACL variable number when reading spool file - insurance
3207 against a later release with more variables having written the file.
3209 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3210 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3211 this, as well as the form without the "IPv6" on the front.
3213 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3214 OS/Makefile-Darwin file.
3216 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3217 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3219 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3220 variable. However, some OS use ints for inodes. I've added cast to long int
3221 to get rid of the compiler warning.
3223 19. I had forgotten to lock out "/../" in configuration file names when
3224 ALT_CONFIG_PREFIX was set.
3226 20. Routers used for verification do not need to specify transports. However,
3227 if such a router generated a host list, and callout was configured, Exim
3228 crashed, because it could not find a port number from the (non-existent)
3229 transport. It now assumes port 25 in this circumstance.
3231 21. Added the -t option to exigrep.
3233 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3234 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3235 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3236 them is set, LOOKUP_LSEARCH is forced.
3238 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3240 24. Added sender and host information to the "rejected by local_scan()" log
3241 line; previously there was no indication of these.
3243 25. Added .include_if_exists.
3245 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3246 Linux. It turns out that not all file systems support this. Apparently some
3247 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3248 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3249 not supported on the file descriptor, is now ignored when Exim is trying to
3250 sync a directory. This applies only to Linux.
3252 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3254 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3255 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3258 29. Added support for maildirsize files from supplied patch (modified a bit).
3260 30. The use of :fail: followed by an empty string could lead Exim to respond to
3261 sender verification failures with (e.g.):
3263 550 Verification failed for <xxx>
3264 550 Sender verify failed
3266 where the first response line was missing the '-' that indicates it is not
3267 the final line of the response.
3269 31. The loop for finding the name of the user that called Exim had a hardwired
3270 limit of 10; it now uses the value of finduser_retries, which is used for
3271 all other user lookups.
3273 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3276 33. Exim was neglecting to zero errno before one call of strtol() when
3277 expanding a string and expecting an integer value. On some systems this
3278 resulted in spurious "integer overflow" errors. Also, it was casting the
3279 result into an int without checking.
3281 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3282 did not work. The code looks as if it has *never* worked, though it appears
3283 to have been documented since at least releast 1.62. I have made it work.
3285 35. The "timeout_DNS" error in retry rules, also documented since at least
3286 1.62, also never worked. As it isn't clear exactly what this means, and
3287 clearly it isn't a major issue, I have abolished the feature by treating it
3288 as "timeout", and writing a warning to the main and panic logs.
3290 36. The display of retry rules for -brt wasn't always showing the error code
3293 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3294 timeout_connect_A, timeout_connect_MX.
3296 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3297 to the empty sender.
3299 39. The daemon was not analysing the content of -oX till after it had closed
3300 stderr and disconnected from the controlling terminal. This meant that any
3301 syntax errors were only noted on the panic log, and the return code from
3302 the command was 0. By re-arranging the code a little, I've made the
3303 decoding happen first, so such errors now appear on stderr, and the return
3304 code is 1. However, the actual setting up of the sockets still happens in
3305 the disconnected process, so errors there are still only recorded on the
3308 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3309 connections (as happens on some IP stacks) was logged at start up time as
3310 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3311 it from "IPv6 and IPv4", which means that two separate sockets are being
3314 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3315 says whether AF_INET or AF_INET6 was passed as an argument.
3317 42. Exiwhat output was messed up when time zones were included in log
3320 43. Exiwhat now gives more information about the daemon's listening ports,
3321 and whether -tls-on-connect was used.
3323 44. The "port" option of the smtp transport is now expanded.
3325 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3326 silently ignored. Now an error message is written to the main and panic
3329 46. There's a new ACL modifier called "logwrite" which writes to a log file
3330 as soon as it is encountered.
3332 47. Added $local_user_uid and $local_user_gid at routing time.
3334 48. Exim crashed when trying to verify a sender address that was being
3337 49. Exim was recognizing only a space character after ".include". It now also
3338 recognizes a tab character.
3340 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3341 extracting the relevant information from the specification. The man page no
3342 longer contains scrambled data for the -d option, and I've added a section
3343 at the front about calling Exim under different names.
3345 51. Added "extra_headers" argument to the "mail" command in filter files.
3347 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3350 53. Installed eximstats 1.29.
3352 54. Added transport_filter_timeout as a generic transport option.
3354 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3355 Cc: header lines. This was required by RFC 822, but it not required by RFC
3358 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3359 incoming messages that did not have them. Now it does so only if the
3360 message originates locally, that is, if there is no associated remote host
3361 address. When Resent- header lines are present, this applies to the Resent-
3362 lines rather than the non-Resent- lines.
3364 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3365 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3367 58. Messages for configuration errors now include the name of the main
3368 configuration file - useful now that there may be more than one file in a
3369 list (.included file names were always shown).
3371 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3372 for those rare installations that do not start the daemon as root or run it
3373 setuid root. I've cut out the call to initgroups() if the daemon is not
3376 60. The Exim user and group can now be bound into the binary as text strings
3377 that are looked up at the start of Exim's processing.
3379 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3381 62. Added $mailstore_basename variable.
3383 63. Installed patch to sieve.c from Michael Haardt.
3385 64. When Exim failed to open the panic log after failing to open the main log,
3386 the original message it was trying to log was written to stderr and debug
3387 output, but if they were not available (the usual case in production), it
3388 was lost. Now it is written to syslog before the two lines that record the
3389 failures to open the logs.
3391 65. Users' Exim filters run in subprocesses under the user's uid. It is
3392 possible for a "deliver" command or an alias in a "personal" command to
3393 provoke an address rewrite. If logging of address rewriting is configured,
3394 this fails because the process is not running as root or exim. There may be
3395 a better way of dealing with this, but for the moment (because 4.30 needs
3396 to be released), I have disabled address rewrite logging when running a
3397 filter in a non-root, non-exim process.
3403 1. The buildconfig auxiliary program wasn't quoting the value set for
3404 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3405 not defined. This bug was masked in 4.22 by the effect that was fixed in
3408 2. Some messages that were rejected after a message id was allocated were
3409 shown as "incomplete" by exigrep. It no longer does this for messages that
3410 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3412 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3413 have allow_domain_literals set, the ID did not get logged in the <= line.
3414 Domain literals are now always recognized in Message-ID: header lines.
3416 4. The first argument for a ${extract expansion item is the key name or field
3417 number. Leading and trailing spaces in this item were not being ignored,
3418 causing some misleading effects.
3420 5. When deliver_drop_privilege was set, single queue runner processes started
3421 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3422 same command in the process it spins off) were not dropping privilege.
3424 6. When the daemon running as "exim" started a queue runner, it always
3425 re-executed Exim in the spun-off process. This is a waste of effort when
3426 deliver_drop_privilege is set. The new process now just calls the
3427 queue-runner function directly.
3433 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3436 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3437 ignored. Though the use of -oP was forcing the writing of a pid file, it
3438 was always written to the default place.
3440 3. If the message "no IP address found for host xxxx" is generated during
3441 incoming verification, it is now followed by identification of the incoming
3442 connection (so you can more easily find what provoked it).
3444 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3446 5. Added some features to "harden" Exim a bit more against certain attacks:
3448 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3449 be put in Local/Makefile. This is like the never_users runtime option,
3450 but it cannot be overridden. The default setting is "root".
3452 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3453 prefix string with which any file named in a -C command line option
3456 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3457 is retained for -C and -D only if the caller of Exim is root. Without
3458 it, the exim user may also use -C and -D and retain privilege.
3460 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3461 command line option is disabled.
3463 6. Macro names set by the -D option must start with an upper case letter, just
3464 like macro names defined in the configuration file.
3466 7. Added "dereference=" facility to LDAP.
3468 8. Two instances of the typo "uknown" in the source files are fixed.
3470 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3471 the Configure-Makefile script screwed up while processing it.
3473 10. Incorporated PCRE 4.4.
3475 11. The SMTP synchronization check was not operating right at the start of an
3476 SMTP session. For example, it could not catch a HELO sent before the client
3477 waited for the greeting. There is now a check for outstanding input at the
3478 point when the greeting is written. Because of the duplex, asynchronous
3479 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3480 way, but not yet received, when the check is performed.
3482 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3483 on TCP/IP sockets, because this apparently causes some broken clients to
3486 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3487 unchanged) from the Cygwin maintainer.
3489 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3490 was supported instead of testing for mbx. Effectively a typo.
3492 15. The spa authenticator server code was not checking that the input it
3493 received was valid base64.
3495 16. The debug output line for the "set" modifier in ACLs was not showing the
3496 name of the variable that was being set.
3498 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3500 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3501 lookup on its own. Something else was needed to trigger the lookup. For
3502 example, a match in host_lookup or the need for a host name in a host list.
3503 Now, if $sender_host_name is referenced and the host name has not yet been
3504 looked up, a lookup is performed. If the lookup fails, the variable remains
3505 empty, and $host_lookup_failed is set to "1".
3507 19. Added "eqi" as a case-independent comparison operator.
3509 20. The saslauthd authentication condition could segfault if neither service
3510 nor realm was specified.
3512 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3513 error message that was logged was misleading, and incoming SMTP
3514 connections were dropped. The message is now more accurate, and temporary
3515 errors are given to SMTP connections.
3517 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3518 (except RSET) with a 421 error, until QUIT is received. However, it was
3519 failing to send a response to QUIT.
3521 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3522 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3523 ACL is now run after the helo_try_verify_hosts code.
3525 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3526 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3527 case-independent, so other case variants are also recognized). Apparently
3528 some systems use these upper case variants.
3530 25. If more than two messages were waiting for the same host, and a transport
3531 filter was specified for the transport, Exim sent two messages over the
3532 same TCP/IP connection, and then failed with "socket operation on non-
3533 socket" when it tried to send the third.
3535 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3537 27. The extern definition of crypt16() in expand.c was not being excluded when
3538 the OS had its own crypt16() function.
3540 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3541 as a preferred synonym for return_size_limit, both as an option and as an
3544 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3546 30. Changed the default configuration ACL to relax the local part checking rule
3547 for addresses that are not in any local domains. For these addresses,
3548 slashes and pipe symbols are allowed within local parts, but the sequence
3549 /../ is explicitly forbidden.
3551 31. SPA server authentication was not clearing the challenge buffer before
3554 32. log_message in a "warn" ACL statement was writing to the reject log as
3555 well as to the main log, which contradicts the documentation and doesn't
3556 seem right (because no rejection is happening). So I have stopped it.
3558 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3559 However, I am unable to do any testing of this.
3561 34. Fixed an infelicity in the appendfile transport. When checking directories
3562 for a mailbox, to see if any needed to be created, it was accidentally
3563 using path names with one or more superfluous leading slashes; tracing
3564 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3566 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3567 contents are added to the log line that is written for every discarded
3568 recipient. (Previously a log_message setting was ignored.)
3570 36. The ${quote: operator now quotes the string if it is empty.
3572 37. The install script runs exim in order to find its version number. If for
3573 some reason other than non-existence or emptiness, which it checks, it
3574 could not run './exim', it was installing it with an empty version number,
3575 i.e. as "exim-". This error state is now caught, and the installation is
3578 38. An argument was missing from the function that creates an error message
3579 when Exim fails to connect to the socket for saslauthd authentication.
3580 This could cause Exim to crash, or give a corrupted message.
3582 39. Added isip, isip4, and isip6 to ${if conditions.
3584 40. The ACL variables $acl_xx are now saved with the message, and can be
3585 accessed later in routers, transports, and filters.
3587 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3588 strings in the file are not string-expanded.
3590 42. If a MAIL command specified a SIZE value that was too large to fit into an
3591 int variable, the check against message_size_limit failed. Such values are
3592 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3593 day this will have to be increased, but I don't think I want to be around
3594 when emails are that large.
3601 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3602 iconv() is not standard in FreeBSD.
3604 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3605 IPv6 enabled. The observed symptom was a segmentation fault on return from
3606 the function os_common_find_running_interfaces() in src/os.c.
3608 3. In the check_special_case() function in daemon.c I had used "errno" as an
3609 argument name, which causes warnings on some systems. This was basically a
3610 typo, since it was named "eno" in the comments!
3612 4. The code that waits for the clock to tick (at a resolution of some fraction
3613 of a second) so as to ensure message-id uniqueness was always waiting for
3614 at least one whole tick, when it could have waited for less. [This is
3615 almost certainly not relevant at current processor speeds, where it is
3616 unlikely to ever wait at all. But we try to future-proof.]
3618 5. The function that sleeps for a time interval that includes fractions of a
3619 second contained a race. It did not block SIGALRM between setting the
3620 timer, and suspending (a couple of lines later). If the interval was short
3621 and the sigsuspend() was delayed until after it had expired, the suspension
3622 never ended. On busy systems this could lead to processes getting stuck for
3625 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3626 process, before it forks any delivery processes. The open lookup caching
3627 mechanism meant that the open file or database connection was passed into
3628 the delivery process. The problem was that delivery processes always tidy
3629 up cached lookup data. This could cause a problem for the next delivery
3630 process started by the queue runner, because the external queue runner
3631 process does not know about the closure. So the next delivery process
3632 still has data in the lookup cache. In the case of a file lookup, there was
3633 no problem because closing a file descriptor in a subprocess doesn't affect
3634 the parent. However, if the lookup was caching a connection to a database,
3635 the connection was closed, and the second delivery process was likely to
3636 see errors such as "PGSQL: query failed: server closed the connection
3637 unexpectedly". The problem has been fixed by closing all cached lookups
3638 in a queue runner before running a delivery process.
3640 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3641 seem to have the "const" qualifier which it has on other OS. I've
3644 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3645 *of the same type* (client or server) with the same public name that an
3646 error should be diagnosed.
3648 9. When Exim looked up a host name for an IP address, but failed to find the
3649 original IP address when looking up the host name (a safety check), it
3650 output the message "<ip address> does not match any IP for NULL", which was
3651 confusing, to say the least. The bug was that the host name should have
3652 appeared instead of "NULL".
3654 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3655 user that is built into the binary, and the -C or -D options is used, root
3656 privilege is dropped before the configuration file is read. In addition,
3657 logging is switched to stderr instead of the normal log files. If the
3658 configuration then re-defines the Exim user, the unprivileged environment
3659 is probably not what is expected, so Exim logs a panic warning message (but
3662 However, if deliver_drop_privilege is set, the unprivileged state may well
3663 be exactly what is intended, so the warning has been cut out in that case,
3664 and Exim is allowed to try to write to its normal log files.
3670 1. smtp_return_error_details was not giving details for temporary sender
3671 or receiver verification errors.
3673 2. Diagnose a configuration error if two authenticators have the same public
3676 3. Exim used not to create the message log file for a message until the first
3677 delivery attempt. This could be confusing when incoming messages were held
3678 for policy or load reasons. The message log file is now created at the time
3679 the message is received, and an initial "Received" line is written to it.
3681 4. The automatically generated man page for command line options had a minor
3682 bug that caused no ill effects; however, a more serious problem was that
3683 the procedure for building the man page automatically didn't always
3684 operate. Consequently, release 4.20 contains an out-of-date version. This
3685 shouldn't happen again.
3687 5. When building Exim with embedded Perl support, the script that builds the
3688 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3689 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3691 6. The freeze_tell option was not being used for messages that were frozen on
3692 arrival, either by an ACL or by local_scan().
3694 7. Added the smtp_incomplete_transaction log selector.
3696 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3697 was accepting AUTH without a new EHLO.
3699 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3700 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3703 10. Logging of TCP/IP connections (when configured) now happens in the main
3704 daemon process instead of the child process, so that the TCP/IP connection
3705 count is more accurate (but it can never be perfect).
3707 11. The use of "drop" in a nested ACL was not being handled correctly in the
3708 outer ACL. Now, if condition failure induced by the nested "drop" causes
3709 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3710 or "require"), the connection is dropped.
3712 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3713 that yield "discard" can now be used with an "accept" or a "discard" verb,
3714 but an error is generated for any others (because I can't see a useful way
3715 to define what should happen).
3717 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3718 are now processed in the same way as lines in the Exim configuration file.
3719 In particular, continuation lines are supported.
3721 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3723 15. Added -ti meaning -t -i.
3725 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3726 domains, and warn by logging if others are found.
3728 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3729 returned from GIFCONF when Exim is trying to find the list of interfaces on
3730 a host. The code in os.c has been modified to copy each ifreq to an aligned
3731 structure in all cases.
3733 Also, in some cases, the returned ifreq's were being copied to a 'struct
3734 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3735 means the last couple of bytes of an IPv6 address could be chopped if the
3736 ifreq contained only a normal sockaddr (14 bytes storage).
3738 18. Named domain lists were not supported in the hosts_treat_as_local option.
3739 An entry such as +xxxx was not recognized, and was treated as a literal
3742 19. Ensure that header lines added by a DATA ACL are included in the reject log
3743 if the ACL subsequently rejects the message.
3745 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3746 MD5 (which is deprecated).
3748 21. When testing a filter file using -bf, Exim was writing a message when it
3749 took the sender from a "From " line in the message, but it was not doing so
3750 when it took $return_path from a Return-Path: header line. It now does.
3752 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3753 with a valid header line field name (a series of printing characters
3754 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3756 23. Changed "disc" in the source to "disk" to conform to the documentation and
3757 the book and for uniformity.
3759 24. Ignore Sendmail's -Ooption=value command line item.
3761 25. When execve() failed while trying to run a command in a pipe transport,
3762 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3763 could be confused with a return value of 69 from the command itself. This
3764 has been changed to 127, the value the shell returns if it is asked to run
3765 a non-existent command. The wording for the related log line suggests a
3766 non-existent command as the problem.
3768 26. If received_header_text expands to an empty string, do not add a Received:
3769 header line to the message. (Well, it adds a token one on the spool, but
3770 marks it "old" so that it doesn't get used or transmitted.)
3772 27. Installed eximstats 1.28 (addition of -nt option).
3774 28. There was no check for failure on the call to getsockname() in the daemon
3775 code. This can fail if there is a shortage of resources on the system, with
3776 ENOMEM, for example. A temporary error is now given on failure.
3778 29. Contrary to the C standard, it seems that in some environments, the
3779 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3780 program. Exim now does this explicitly; it affects the formatting of
3781 timestamps using strftime().
3783 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3784 complaints. I've now initialized all the variables, to avoid this.
3786 32. Header lines added by a system filter were not being "seen" during
3787 transport-time rewrites.
3789 33. The info_callback() function passed to OpenSSL is set up with type void
3790 (*)(SSL *, int, int), as described somewhere. However, when calling the
3791 function (actually a macro) that sets it up, the type void(*)() is
3792 expected. I've put in a cast to prevent warnings from picky compilers.
3794 34. If a DNS black list lookup found a CNAME record, but there were no A
3795 records associated with the domain it pointed at, Exim crashed.
3797 35. If a DNS black list lookup returned more than one A record, Exim ignored
3798 all but the first. It now scans all returned addresses if a particular IP
3799 value is being sought. In this situation, the contents of the
3800 $dnslist_value variable are a list of all the addresses, separated by a
3803 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3804 to accept a host name and all its aliases if the forward lookup for any of
3805 them yielded the IP address of the incoming connection. Now it accepts only
3806 those names whose forward lookup yields the correct IP address. Any other
3807 names are discarded. This closes a loophole whereby a rogue DNS
3808 administrator could create reverse DNS records to break through a
3809 wildcarded host restriction in an ACL.
3811 37. If a user filter or a system filter that ran in a subprocess used any of
3812 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3813 the wrong values were passed to the pipe command ($thisaddress had the
3814 value of $0, $0 had the value of $1, etc). This bug was introduced by
3815 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3817 38. Improved the line breaking for long SMTP error messages from ACLs.
3818 Previously, if there was no break point between 40 and 75 characters, Exim
3819 left the rest of the message alone. Two changes have been made: (a) I've
3820 reduced the minimum length to 35 characters; (b) if it can't find a break
3821 point between 35 and 75 characters, it looks ahead and uses the first one
3822 that it finds. This may give the occasional overlong line, but at least the
3823 remaining text gets split now.
3825 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3826 file descriptors might be low, and that setting 1000 would always raise it.
3827 It turns out that in some environments, the limit is already over 1000 and
3828 that lowering it causes trouble. So now Exim takes care not to decrease it.
3830 40. When delivering a message, the value of $return_path is set to $sender_
3831 address at the start of routing (routers may change the value). By an
3832 oversight, this default was not being set up when an address was tested by
3833 -bt or -bv, which affected the outcome if any router or filter referred to
3836 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3837 log only when "message" or "log_message" are set. However, if one of the
3838 conditions was an address verification, or a call to a nested ACL, the
3839 messages generated by the underlying test were being passed through. This
3840 no longer happens. The underlying message is available in $acl_verify_
3841 message for both "message" and "log_message" expansions, so it can be
3842 passed through if needed.
3844 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3845 new expansion $bh_ to give the encoded byte string without charset
3846 translation. Translation happens only if iconv() is available; HAVE_ICONV
3847 indicates this at build time. HEADERS_CHARSET gives the charset to
3848 translate to; headers_charset can change it in the configuration, and
3849 "headers charset" can change it in an individual filter file.
3851 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3852 that creates RFC 2047 encoded "words" labels them as that charset instead
3853 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3854 expansion operator; (ii) when Exim creates a From: line for a local
3855 message; (iii) when a header line is rewritten to include a "phrase" part.
3857 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3858 buggy, causing it to skip the first lines of messages whose message ID
3859 ended in 'D'. This would not have bitten before Exim release 4.14, because
3860 message IDs were unlikely to end in 'D' before then. The effect was to have
3861 incorrect size information for certain domains.
3863 45. #include "config.h" was missing at the start of the crypt16.c module. This
3864 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3867 46. If there was a timeout during a "random" callout check, Exim treated it as
3868 a failure of the random address, and carried on sending RSET and the real
3869 address. If the delay was just some slowness somewhere, the response to the
3870 original RCPT would be taken as a response to RSET and so on, causing
3871 mayhem of various kinds.
3873 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3874 when I implemented it. It didn't allow for the fact that some option values
3875 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3876 the right test for positive values.
3878 48. Domain names in DNS records are case-independent. Exim always looks them up
3879 in lower case. Some resolvers return domain names in exactly the case they
3880 appear in the zone file, that is, they may contain uppercase letters. Not
3881 all resolvers do this - some return always lower case. Exim was treating a
3882 change of case by a resolver as a change of domain, similar to a widening
3883 of a domain abbreviation. This triggered its re-routing code and so it was
3884 trying to route what was effectively the same domain again. This normally
3885 caused routing to fail (because the router wouldn't handle the domain
3886 twice). Now Exim checks for this case specially, and just changes the
3887 casing of the domain that it ultimately uses when it transmits the message
3890 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3893 50. If a filter generated a file delivery with a non-absolute name (possible if
3894 no home directory exists for the router), the forbid_file option was not
3897 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3898 the existing equality matching.
3900 52. Exim was using ints instead of ino_t variables in some places where it was
3901 dealing with inode numbers.
3903 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3904 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3905 called TMPDIR, and if it finds it is different, it changes its value.
3907 54. The smtp_printf() function is now made available to local_scan() so
3908 additional output lines can be written before returning. There is also an
3909 smtp_fflush() function to enable the detection of a dropped connection.
3910 The variables smtp_input and smtp_batched_input are exported to
3913 55. Changed the default runtime configuration: the message "Unknown user"
3914 has been removed from the ACL, and instead placed on the localuser router,
3915 using the cannot_route_message feature. This means that any verification
3916 failures that generate their own messages won't get overridden. Similarly,
3917 the "Unrouteable address" message that was in the ACL for unverifiable
3918 relay addresses has also been removed.
3920 56. Added hosts_avoid_esmtp to the smtp transport.
3922 57. The exicyclog script was not checking for the esoteric option
3923 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3924 will work only if exicyclog is run under the appropriate euid.
3926 58. Following a discussion on the list, the rules by which Exim recognises line
3927 endings on incoming messages have been changed. The -dropcr and drop_cr
3928 options are now no-ops, retained only for backwards compatibility. The
3929 following line terminators are recognized: LF CRLF CR. However, special
3930 processing applies to CR:
3932 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3933 nor a local message in the state where . is a terminator.
3935 (ii) If a bare CR is encountered in a header line, an extra space is added
3936 after the line terminator so as not to end the header. The reasoning
3937 behind this is that bare CRs in header lines are most likely either
3938 to be mistakes, or people trying to play silly games.
3940 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3941 was being incorrectly given as 18 bytes larger than it should have been.
3942 This is a VOB (very old bug).
3944 60. This may never have affected anything current, but just in case it has:
3945 When the local host is found other than at the start of a list of hosts,
3946 the local host, those with the same MX, and any that follow, are discarded.
3947 When the list in question was part of a longer list of hosts, the following
3948 hosts (not currently being processed) were also being discarded. This no
3949 longer happens. I'm not sure if this situation could ever has previously
3952 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3955 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3956 header. This is something that is recommended in a new Internet Draft, and
3957 is something that is documented as being done by Sendmail. There are two
3958 possible values. For messages generated by the autoreply transport, Exim
3961 Auto-Submitted: auto-replied
3963 whereas for all other generated messages (e.g. bounces) it adds
3965 Auto-Submitted: auto-generated
3967 63. The "personal" condition in filters now includes a test for the
3968 Auto-Submitted: header. If it contains the string "auto-" the message it
3969 not considered personal.
3971 64. Added rcpt_include_affixes as a generic transport option.
3973 65. Added queue_only_override (default true).
3975 66. Added the syslog_duplication option.
3977 67. If what should have been the first header line of a message consisted of
3978 a space followed by a colon, Exim was mis-interpreting it as a header line.
3979 It isn't of course - it is syntactically invalid and should therefore be
3980 treated as the start of the message body. The misbehaviour could have
3981 caused a number of strange effects, including loss of data in subsequent
3982 header lines, and spool format errors.
3984 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3985 client host had authenticated. This control can now be exercised by an ACL
3986 for more flexibility.
3988 69. By default, callouts do not happen when testing with -bh. There is now a
3989 variant, -bhc, which does actually run the callout code, including
3990 consulting and updating the callout cache.
3992 70. Added support for saslauthd authentication, courtesy of Alexander
3995 71. If statvfs() failed on the spool or log directories while checking their
3996 size for availability, Exim confusingly gave the error "space shortage".
3997 Furthermore, in debugging mode it crashed with a floating point exception.
3998 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3999 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4000 serious problem, Exim now writes to the main and panic logs when this
4001 happens, with details of the failure. It then refuses to accept the
4002 incoming message, giving the message "spool directory problem" or "log
4003 directory problem" with a 421 code for SMTP messages.
4005 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4006 0, 1, and 2 exist, because some OS complain for execs without them (see
4007 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4008 descriptors. However, the code omitted to check that the open succeeded,
4009 causing mysterious errors if for some reason the permissions on /dev/null
4010 got screwed. Now Exim writes a message to the main and panic logs, and
4011 bombs out if it can't open /dev/null.
4013 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4014 interact so that it is all more flexible. It is supposed to remain
4015 backwards compatible. Also added extra_local_interfaces.
4017 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4018 to bomb out with an assertion failure - to the client this appears as a
4019 connection drop. This problem occurs in the part of the code that was taken
4020 from the Samba project. Fortunately, the assertion is in a very simple
4021 function, so I have fixed this by reproducing the function inline in the
4022 one place where it is called, and arranging for authentication to fail
4023 instead of killing the process with assert().
4025 75. The SPA client code was not working when the server requested OEM rather
4026 than Unicode encoding.
4028 76. Added code to make require_files with a specific uid setting more usable in
4029 the case where statting the file as root fails - usually a non-root-mounted
4030 NFS file system. When this happens and the failure is EACCES, Exim now
4031 forks a subprocess and does the per-uid checking as the relevant uid.
4033 77. Added process_log_path.
4035 78. If log_file_path was not explicitly set, a setting of check_log_space or
4036 check_log_inodes was ignored.
4038 79. If a space check for the spool or log partitions fails, the incident is now
4039 logged. Of course, in the latter case the data may get lost...
4041 80. Added the %p formatting code to string_format() so that it can be used to
4042 print addresses in debug_print(). Adjusted all the address printing in the
4043 debugging in store.c to use %p rather than %d.
4045 81. There was a concern that a line of code in smtp_in.c could overflow a
4046 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4047 initially expressed, the concern was not well-founded, because trailing
4048 spaces are removed early. However, if the trailing spaces were followed by
4049 a NULL, they did not get removed, so the overflow was possible. Two fixes
4052 (a) I re-wrote the offending code in a cleaner fashion.
4053 (b) If an incoming SMTP command contains a NULL character, it is rejected
4056 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4057 runs initgroups(), so that if the Exim user is in any additional groups,
4058 they will be used during message reception.
4064 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob