1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.370 2006/07/07 13:36:33 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
90 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
91 other effects) broke the use of negated acl sub-conditions.
93 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
96 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
97 "Deny" causes Exim to reject the incoming connection with a 554 error.
98 Unfortunately, if there is a major crisis, such as a disk failure,
99 tcp-wrappers gives "deny", whereas what one would like would be some
100 kind of temporary error. A kludge has been added to help with this.
101 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
102 554 error is used if errno is still zero or contains ENOENT (which occurs
103 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
106 PH/03 Add -lutil to the default FreeBSD LIBS setting.
108 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
109 errors. Otherwise a message that provokes a temporary error (when other
110 messages do not) can cause a whole host to time out.
112 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
113 addresses were routed directly to files or pipes from a redirect router.
114 File deliveries just didn't batch; pipe deliveries might have suffered
117 PH/06 A failure to get a lock for a hints database would erroneously always say
118 "Failed to get write lock", even when it was really a read lock.
120 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
121 of 0600. This has been changed to use the value of the lockfile_mode
122 option (which defaults to 0600).
124 PH/08 Applied small patch from the Sieve maintainer.
126 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
127 folder from quota calculations, a direct delivery into this folder messed
128 up the contents of the maildirsize file. This was because the regex was
129 used only to exclude .Trash (or whatever) when the size of the mailbox
130 was calculated. There was no check that a delivery was happening into an
131 excluded directory. This bug has been fixed by ignoring all quota
132 processing for deliveries into excluded directories.
134 PH/10 Added the maildirfolder_create_regex option to appendfile.
140 PH/01 The code for finding all the local interface addresses on a FreeBSD
141 system running IPv6 was broken. This may well have applied to all BSD
142 systems, as well as to others that have similar system calls. The broken
143 code found IPv4 interfaces correctly, but gave incorrect values for the
144 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
145 that it would not match correctly against @[] and not recognize the IPv6
148 PH/02 The ipliteral router was not recognizing addresses of the form user@
149 [ipv6:....] because it didn't know about the "ipv6:" prefix.
151 PH/03 Added disable_ipv6.
153 PH/04 Changed $reply_address to use the raw form of the headers instead of the
154 decoded form, because it is most often used to construct To: headers
155 lines in autoreplies, and the decoded form may well be syntactically
156 invalid. However, $reply_address has leading white space removed, and all
157 newlines turned into spaces so that the autoreply transport does not
160 PH/05 If group was specified without a user on a router, and no group or user
161 was specified on a transport, the group from the router was ignored.
163 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
164 for visible compile-time settings that can be used to change these
165 numbers, for those that want even more. Backwards compatibility with old
166 spool files has been maintained. However, going back to a previous Exim
167 release will lost any variables that are in spool files.
169 PH/07 Two small changes when running in the test harness: increase delay when
170 passing a TCP/IP connection to a new process, in case the original
171 process has to generate a bounce, and remove special handling of
172 127.0.0.2 (sic), which is no longer necessary.
174 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
175 be the same on different OS.
177 PH/09 Moved a debug statement in filter processing to avoid a race problem when
180 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
181 whether --show-vars was specified or not
183 JJ/02 exipick: Added support for new ACL variable spool format introduced
186 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
187 syntactically invalid From: or Reply-to: line, and a filter used this to
188 generate an autoreply, and therefore failed to obtain an address for the
189 autoreply, Exim could try to deliver to a non-existent relative file
190 name, causing unrelated and misleading errors. What now happens is that
191 it logs this as a hard delivery error, but does not attempt to create a
194 PH/11 The exinext utility has a -C option for testing purposes, but although
195 the given file was scanned by exinext itself; it wasn't being passed on
196 when Exim was called.
198 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
199 an end-of-file indication when reading a command response.
201 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
202 compiled. In many other places in Exim, IPv6 addresses are always
203 recognized, so I have changed this. It also means that IPv4 domain
204 literals of the form [IPV4:n.n.n.n] are now always recognized.
206 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
207 used if the router is not running as root, for example, when verifying at
208 ACL time, or when using -bh. The debugging output from this situation was
209 non-existent - all you got was a failure to exec. I have made two
212 (a) Failures to set uid/gid, the current directory, or a process leader
213 in a subprocess such as that created by queryprogram now generate
214 suitable debugging ouput when -d is set.
216 (b) The queryprogram router detects when it is not running as root,
217 outputs suitable debugging information if -d is set, and then runs
218 the subprocess without attempting to change uid/gid.
220 PH/15 Minor change to Makefile for building test_host (undocumented testing
223 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
224 additional section of a DNS packet that returns MX or SRV records.
225 Instead, it always explicitly searches for A/AAAA records. This avoids
226 major problems that occur when a DNS server includes only records of one
227 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
228 fixed another bug: if SRV records were looked up and the corresponding
229 address records were *not* found in the additional section, the port
230 values from the SRV records were lost.
232 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
233 using the correct key (the original address) when searching the retry
234 rules in order to find which one to use for generating the retry hint.
236 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
237 adding the default one. Similarly, if it contains a Reply-To: header, the
238 errors_reply_to option, if set, is not used.
240 PH/19 When calculating a retry time, Exim used to measure the "time since
241 failure" by looking at the "first failed" field in the retry record. Now
242 it does not use this if it is later than than the arrival time of the
243 message. Instead it uses the arrival time. This makes for better
244 behaviour in cases where some deliveries succeed, thus re-setting the
245 "first failed" field. An example is a quota failure for a huge message
246 when small messages continue to be delivered. Without this change, the
247 "time since failure" will always be short, possible causing more frequent
248 delivery attempts for the huge message than are intended.
249 [Note: This change was subsequently modified - see PH/04 for 4.62.]
251 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
252 $1, $2, $3) because the numerical variables can be reset during some
253 expansion items (e.g. "match"), thereby losing the authentication data.
255 PH/21 Make -bV show the size of off_t variables so that the test suite can
256 decide whether to run tests for quotas > 2G.
258 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
259 mailbox_size, and mailbox_filecount in the appendfile transport. If a
260 filecount value is greater than 2G or if a quota value is greater than 2G
261 on a system where the size of off_t is not greater than 4, a panic error
264 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
265 never match. The debug and -bh output now contains an explicit error
266 message indicating a malformed IPv4 address or mask.
268 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
269 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
272 PH/25 Do not write to syslog when running in the test harness. The only
273 occasion when this arises is a failure to open the main or panic logs
274 (for which there is an explicit test).
276 PH/26 Added the /no_tell option to "control=freeze".
278 PH/27 If a host name lookup failed very early in a connection, for example, if
279 the IP address matched host_lookup and the reverse lookup yielded a name
280 that did not have a forward lookup, an error message of the form "no IP
281 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
282 could be logged. Now it outputs the IP address instead of "NULL".
284 PH/28 An enabling patch from MH: add new function child_open_exim2() which
285 allows the sender and the authenticated sender to be set when
286 submitting a message from within Exim. Since child_open_exim() is
287 documented for local_scan(), the new function should be too.
289 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
290 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
291 results in an empty string is now treated as unset.
293 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
295 PH/31 Added sender_verify_fail logging option.
297 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
298 needed by RFC 822 but not by RFC 2822 was commented out. I have now
299 tidied the source and removed it altogether.
301 PH/33 When a queue run was abandoned because the load average was too high, a
302 log line was always written; now it is written only if the queue_run log
303 selector is set. In addition, the log line for abandonment now contains
304 information about the queue run such as the pid. This is always present
305 in "start" and "stop" lines but was omitted from the "abandon" line.
307 PH/34 Omit spaces between a header name and the colon in the error message that
308 is given when verify = headers_syntax fails (if there are lots of them,
309 the message gets confusing).
311 PH/35 Change the default for dns_check_names_pattern to allow slashes within
312 names, as there are now some PTR records that contain slashes. This check
313 is only to protect against broken name servers that fall over on strange
314 characters, so the fact that it applies to all lookups doesn't matter.
316 PH/36 Now that the new test suite is complete, we can remove some of the
317 special code in Exim that was needed for the old test suite. For example,
318 sorting DNS records because real resolvers return them in an arbitrary
319 order. The new test suite's fake resolver always returns records in the
322 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
323 bounces) except when queue_only is set, to avoid logging races between
324 the different processes.
326 PH/38 Panic-die if .include specifies a non-absolute path.
328 PH/39 A tweak to the "H" retry rule from its user.
330 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
331 a label. They prevented compilation on older perls.
333 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
334 a warning to be raised on newish perls.
336 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
337 on queue. Changes to match documented behaviour of showing count of
338 messages matching specified criteria.
340 PH/40 Changed the default ident timeout from 30s to 5s.
342 PH/41 Added support for the use of login_cap features, on those BSD systems
343 that have them, for controlling the resources used by pipe deliveries.
345 PH/42 The content-scanning code uses fopen() to create files in which to put
346 message data. Previously it was not paying any attention to the mode of
347 the files. Exim runs with umask(0) because the rest of the code creates
348 files with open(), and sets the required mode explicitly. Thus, these
349 files were ending up world-writeable. This was not a big issue, because,
350 being within the spool directory, they were not world-accessible. I have
351 created a function called modefopen, which takes an additional mode
352 argument. It sets umask(777), creates the file, chmods it to the required
353 mode, then resets the umask. All the relevant calls to fopen() in the
354 content scanning code have been changed to use this function.
356 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
357 to 24 hours. This avoids potential overflow problems when processing G
358 and H retry rules. I suspect nobody ever tinkers with this value.
360 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
362 PH/45 When the plaintext authenticator is running as a client, the server's
363 challenges are checked to ensure they are valid base64 strings. By
364 default, the authentication attempt is cancelled if an invalid string is
365 received. Setting client_ignore_invalid_base64 true ignores these errors.
366 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
367 they are received. Thus, the responses can be made to depend on the
368 challenges. If an invalid string is ignored, an empty string is placed in
371 PH/46 Messages that are created by the autoreply transport now contains a
372 References: header, in accordance with RFCs 2822 and 3834.
374 PH/47 Added authenticated_sender_force to the smtp transport.
376 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
378 PH/49 Installed latest patch from the Sieve maintainer.
380 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
381 was also set, the appendfile transport was unnecessarily scanning a
382 directory of message files (e.g. for maildir delivery) to find the count
383 of files (along with the size), even though it did not need this
384 information. It now does the scan only if it needs to find either the
385 size of the count of files.
387 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
389 PH/52 Two bugs concerned with error handling when the smtp transport is
392 (i) Exim was not creating retry information for temporary errors given
393 for individual recipients after the DATA command when the smtp transport
394 was used in LMTP mode. This meant that they could be retried too
395 frequently, and not timed out correctly.
397 (ii) Exim was setting the flag that allows error details to be returned
398 for LMTP errors on RCPT commands, but not for LMTP errors for individual
399 recipients that were returned after the DATA command.
401 PH/53 This is related to PH/52, but is more general: for any failing address,
402 when detailed error information was permitted to be returned to the
403 sender, but the error was temporary, then after the final timeout, only
404 "retry timeout exceeded" was returned. Now it returns the full error as
405 well as "retry timeout exceeded".
407 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
408 do this, and (what is worse) MTAs that accept it.
410 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
411 will now be deprecated.
413 PH/56 New os.c-cygwin from the Cygwin maintainer.
415 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
416 formats (previously only available in exim formats via -bpr, -bpru,
417 and -bpra. Now also available in native and exiqgrep formats)
419 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
420 with very large, slow to parse queues
422 JJ/08 exipick: added ! as generic prefix to negate any criteria format
424 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
426 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
427 responses to authentication challenges, though it was showing the
428 challenges; (ii) I've removed the CR characters from the debug output for
431 PH/58 Allow for the insertion of a newline as well as a space when a string
432 is turned into more than one encoded-word during RFC 2047 encoding. The
433 Sieve code now uses this.
435 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
436 data_4xx, lost_connection, tls_required.
438 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
439 message was being sent as an SMTP response.
441 PH/61 Add -l and -k options to exicyclog.
443 PH/62 When verifying, if an address was redirected to one new address, so that
444 verification continued, and the new address failed or deferred after
445 having set something in $address_data, the value of $address_data was not
446 passed back to the ACL. This was different to the case when no
447 redirection occurred. The value is now passed back in both cases.
449 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
450 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
451 use login_cap.h, so on its own it isn't the distinguishing feature. The
452 new name refers directly to the setclassresources() function.
454 PH/65 Added configuration files for NetBSD3.
456 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
458 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
459 is preferred over IPv4.
461 PH/68 The bounce_return_message and bounce_return_body options were not being
462 honoured for bounces generated during the reception of non-SMTP messages.
463 In particular, this applied to messages rejected by the ACL. This bug has
464 been fixed. However, if bounce_return_message is true and bounce_return_
465 body is false, the headers that are returned for a non-SMTP message
466 include only those that have been read before the error was detected.
467 (In the case of an ACL rejection, they have all been read.)
469 PH/69 The HTML version of the specification is now built in a directory called
470 spec_html instead of spec.html, because the latter looks like a path with
471 a MIME-type, and this confuses some software.
473 PH/70 Catch two compiler warnings in sieve.c.
475 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
476 function verify_get_ident() calls ip_connect() to connect a socket, but
477 if the "connect()" function timed out, ip_connect() used to close the
478 socket. However, verify_get_ident() also closes the socket later, and in
479 between Exim writes to the log, which may get opened at this point. When
480 the socket was closed in ip_connect(), the log could get the same file
481 descriptor number as the socket. This naturally causes chaos. The fix is
482 not to close the socket in ip_connect(); the socket should be closed by
483 the function that creates it. There was only one place in the code where
484 this was missing, in the iplookup router, which I don't think anybody now
485 uses, but I've fixed it anyway.
487 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
488 well as to direct DNS lookups. Otherwise the handling of names in host
489 lists is inconsistent and therefore confusing.
495 PH/01 Two changes to the default runtime configuration:
497 (1) Move the checks for relay_from_hosts and authenticated clients from
498 after to before the (commented out) DNS black list checks.
500 (2) Add control=submission to the relay_from_hosts and authenticated
501 clients checks, on the grounds that messages accepted by these
502 statements are most likely to be submissions.
504 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
506 (1) Generate an error if the third argument for the ${prvs expansion is
509 (2) Treat a missing third argument of ${prvscheck as if it were an empty
512 (3) Reset the variables that are obtained from the first argument of
513 ${prvscheck and used in the second argument before leaving the code,
514 because their memory is reclaimed, so using them afterwards may do
517 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
518 one (it's much easier than Tom thought :-).
520 (5) Because of (4), we can now allow for the use of $prvscheck_result
521 inside the third argument.
523 PH/03 For some reason, the default setting of PATH when running a command from
524 a pipe transport was just "/usr/bin". I have changed it to
527 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
528 anything to be listed in the output from -bV.
530 PH/05 When a filter generated an autoreply, the entire To: header line was
531 quoted in the delivery log line, like this:
533 => >A.N.Other <ano@some.domain> <original@ddress> ...
535 This has been changed so that it extracts the operative address. There
536 may be more than one such address. If so, they are comma-separated, like
539 => >ano@some.domain,ona@other.domain <original@ddress> ...
541 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
542 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
543 not being looked up in the rDNS to get a host name, Exim was showing the
544 IP address twice in Received: lines, even though the IP addresses were
545 identical. For example:
547 Received: from [1.2.3.4] (helo=[1.2.3.4])
549 However, if the real host name was known, it was omitting the HELO data
550 if it matched the actual IP address. This has been tidied up so that it
551 doesn't show the same IP address twice.
553 PH/07 When both +timestamp and +memory debugging was on, the value given by
554 $tod_xxx expansions could be wrong, because the tod_stamp() function was
555 called by the debug printing, thereby overwriting the timestamp buffer.
556 Debugging no longer uses the tod_stamp() function when +timestamp is set.
558 PH/08 When the original message was included in an autoreply transport, it
559 always said "this is a copy of the message, including all the headers",
560 even if body_only or headers_only was set. It now gives an appropriate
563 PH/09 Applied a patch from the Sieve maintainer which:
565 o fixes some comments
566 o adds the (disabled) notify extension core
567 o adds some debug output for the result of if/elsif tests
568 o points to the current vacation draft in the documentation
569 and documents the missing references header update
573 o fixes a bug in processing the envelope test (when testing
574 multiple envelope elements, the last element determinted the
577 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
578 Electronic Mail") by including:
580 Auto-submitted: auto-generated
582 in the messages that it generates (bounce messages and others, such as
583 warnings). In the case of bounce messages for non-SMTP mesages, there was
584 also a typo: it was using "Auto_submitted" (underscore instead of
585 hyphen). Since every message generated by Exim is necessarily in response
586 to another message, thes have all been changed to:
588 Auto-Submitted: auto-replied
590 in accordance with these statements in the RFC:
592 The auto-replied keyword:
594 - SHOULD be used on messages sent in direct response to another
595 message by an automatic process,
597 - MUST NOT be used on manually-generated messages,
599 - MAY be used on Delivery Status Notifications (DSNs) and Message
600 Disposition Notifications (MDNs),
602 - MUST NOT be used on messages generated by automatic or periodic
603 processes, except for messages which are automatic responses to
606 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
607 to the default Received: header definition.
609 PH/12 Added log selector acl_warn_skipped (default on).
611 PH/13 After a successful wildlsearch lookup, discard the values of numeric
612 variables because (a) they are in the wrong storage pool and (b) even if
613 they were copied, it wouldn't work properly because of the caching.
615 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
616 checking when decoding. Apparently there are clients that generate
617 overlong encoded strings. Why am I not surprised?
619 PH/15 If the first argument of "${if match_address" was not empty, but did not
620 contain an "@" character, Exim crashed. Now it writes a panic log message
621 and treats the condition as false.
623 PH/16 In autoreply, treat an empty string for "once" the same as unset.
625 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
626 extension "envelope-auth". The code is finished and in agreement with
627 other implementations, but there is no documentation so far and in fact,
628 nobody wrote the draft yet. This extension is currently #undef'ed, thus
629 not changing the active code.
631 Print executed "if" and "elsif" statements when debugging is used. This
632 helps a great deal to understand what a filter does.
634 Document more things not specified clearly in RFC3028. I had all this
635 sorted out, when out of a sudden new issues came to my mind. Oops."
637 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
640 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
641 canonical form (as documented). However, after a host name lookup from
642 the IP address, check_host() was doing a simple string comparison with
643 addresses acquired from the DNS when checking that the found name did
644 have the original IP as one of its addresses. Since any found IPv6
645 addresses are likely to be in abbreviated form, the comparison could
646 fail. Luckily, there already exists a function for doing the comparison
647 by converting both addresses to binary, so now that is used instead of
650 PH/20 There was another similar case to PH/19, when a complete host name was
651 given in a host list; looking up its IP address could give an abbreviated
652 form, whereas the current host's name might or might not be abbreviated.
653 The same fix has been applied.
659 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
660 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
663 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
664 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
666 PH/03 Typo: missing ".o" in src/pcre/Makefile.
668 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
669 header line, restrict the check to what is listed in RFCs 2369 and 2929.
670 Also, for "Auto-Submitted", treat anything other than "no" as
671 non-personal, in accordance with RFC 3834. (Previously it treated
672 anything starting "auto-" as non-personal.)
674 TF/01 The control=submission/name=... option had a problem with syntax
675 errors if the name included a slash character. The /name= option
676 now slurps the rest of the string, so it can include any characters
677 but it must come last in the list of options (after /sender_retain
680 PH/05 Some modifications to the interface to the fake nameserver for the new
688 TK/01 Added the "success_on_redirect" address verification option. See
689 NewStuff for rationale and an example.
691 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
693 PH/02 Patch to exigrep to allow it to work on syslog lines.
695 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
696 fread() to skip over the body file's header line, because in Cygwin the
697 header line is locked and is inaccessible.
699 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
700 co-exist for some time) to make it clear that it is the Exim ID that is
701 referenced, not the Message-ID: header line.
703 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
704 string_format() function, because snprintf() does not exist on all
707 PH/06 The use of forbid_filter_existstest now also locks out the use of the
708 ${stat: expansion item.
710 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
711 protocol synchronization error", to keep the pedants happy.
713 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
714 well as for IRIX systems, when gcc is being used. See the host.c source
717 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
719 PH/10 Named domain lists were not working if used in a queue_smtp_domains
722 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
723 transport and to the smtp transport in LMTP mode.
725 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
727 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
728 run a filter in a subprocess. This could lead to confusion in subsequent
729 lookups in the parent process. There should also be a search_tidyup() at
730 the end of the subprocess.
732 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
733 only if the host matched helo_try_verify_hosts, which caused the
734 verification to occur when the EHLO/HELO command was issued. The ACL just
735 tested the remembered result. Now, if a previous verification attempt has
736 not happened, "verify = helo" does it there and then.
738 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
740 TK/03 Fix log output including CR from clamd.
742 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
743 exist provoked a memory error which could cause a segfault.
745 PH/15 Installed PCRE 6.2
747 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
749 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
750 of the problem. Specifically, suggested +O2 rather than +O1 for the
753 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
755 PH/20 If a delivery was routed to a non-standard port by means of an SRV
756 record, the port was not correctly logged when the outgoing_port log
757 selector was set (it logged the transort's default port).
759 PH/21 Added support for host-specific ports to manualroute, queryprogram,
760 fallback_hosts, and "hosts" in the smtp transport.
762 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
763 host errors such as "Connection refused".
765 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
766 authentication with radiusclient 0.4.9:
768 - Error returned from rc_read_config was caught wrongly
769 - Username/password not passed on to radius server due to wrong length.
771 The presumption is that some radiusclient API changes for 4.51/PH/17
772 were not taken care of correctly. The code is still untested by me (my
773 Linux distribution still has 0.3.2 of radiusclient), but it was
774 contributed by a Radius user.
776 PH/24 When doing a callout, the value of $domain wasn't set correctly when
777 expanding the "port" option of the smtp transport.
779 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
780 while reading a MIME header. Thanks to Tom Hughes for a patch.
782 PH/24 Include config.h inside local_scan.h so that configuration settings are
785 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
786 that in an ACL for RCPT (for example), you can examine exactly what was
789 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
790 commands, but it was not correctly comparing the address with the actual
791 client host address. Thus, it would show the EHLO address in Received:
792 header lines when this was not necessary.
794 PH/27 Added the % operator to ${eval:}.
796 PH/28 Exim tries to create and chdir to its spool directory when it starts;
797 it should be ignoring failures (because with -C, for example, it has lost
798 privilege). It wasn't ignoring creation failures other than "already
801 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
804 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
805 because an input line was too long, either on its own, or by virtue of
806 too many continuations, the temporary file was not being removed, and the
807 return code was incorrect.
809 PH/31 Missing "BOOL" in function definition in filtertest.c.
811 PH/32 Applied Sieve patches from the maintainer.
813 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
815 PH/33 Added "verify = not_blind".
817 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
818 Local/Makefile (with some defaults set). These are used in built scripts
819 such as exicyclog, but they have never been used in the exim_install
820 script (though there are many overriding facilities there). I have
821 arranged that the exim_install script now takes note of these two
824 PH/35 Installed configuration files for Dragonfly.
826 PH/36 When a locally submitted message by a trusted user did not contain a
827 From: header, and the sender address was obtained from -f or from an SMTP
828 MAIL command, and the trusted user did not use -F to supply a sender
829 name, $originator_name was incorrectly used when constructing a From:
830 header. Furthermore, $originator_name was used for submission mode
831 messages from external hosts without From: headers in a similar way,
832 which is clearly wrong.
834 PH/37 Added control=suppress_local_fixups.
836 PH/38 When log_selector = +received_sender was set, and the addition of the
837 sender made the log line's construction buffer exactly full, or one byte
838 less than full, an overflow happened when the terminating "\n" was
841 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
842 when the result of a list match is failure because a DNS lookup failed.
844 PH/40 RM_COMMAND is now used in the building process.
846 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
847 the "build-* directories that it finds.
849 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
850 address in a domain literal was a prefix of an interface address.
852 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
853 when verifying a sender address, unless rewrite_headers is false.
855 PH/44 Wrote a long comment about why errors_to addresses are verified as
856 recipients, not senders.
858 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
859 the ratelimit ACL was added.
861 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
863 PH/46 Added extra information about PostgreSQL errors to the error string.
865 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
866 avoiding the need to install special zones in a real server. This is
867 backwards compatible; if it can't find the fake resolver, it drops back.
868 Thus, both old and new test suites can be run.
870 TF/02 Added util/ratelimit.pl
872 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
873 clock is set back in time.
875 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
876 Candler <B.Candler@pobox.com>.
878 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
879 OK for addresses that are the result of redirections.
881 PH/48 A number of further additions for the benefit of the new test suite,
882 including a fake gethostbyname() that interfaces to the fake DNS resolver
885 TF/06 The fix for widen_domains has also been applied to qualify_single and
886 search_parents which are the other dnslookup options that can cause
889 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
892 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
894 TF/07 Exim produced the error message "an SRV record indicated no SMTP
895 service" if it encountered an MX record with an empty target hostname.
896 The message is now "an MX or SRV record indicated no SMTP service".
898 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
899 if the DNS of the sending site is misconfigured. This is quite a
900 common situation. This change restores the behaviour of treating a
901 helo verification defer as a failure.
903 PH/51 If self=fail was set on a router, the bounce message did not include the
904 actual error message.
910 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
912 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
913 command itself ended in error, the underlying message about the transport
914 filter timeout was being overwritten with the pipe command error. Now the
915 underlying error message should be appended to the second error message.
917 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
919 PH/02 Reduce the amount of output that "make" produces by default. Full output
920 can still be requested.
922 PH/03 The warning log line about a condition test deferring for a "warn" verb
923 was being output only once per connection, rather than after each
924 occurrence (because it was using the same function as for successful
925 "warn" verbs). This seems wrong, so I have changed it.
927 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
928 it should not have, which might have caused a crash in the right
929 circumstances, but probably never did.
931 PH/04 Installed a modified version of Tony Finch's patch to make submission
932 mode fix the return path as well as the Sender: header line, and to
933 add a /name= option so that you can make the user's friendly name appear
936 TF/03 Added the control = fakedefer ACL modifier.
938 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
939 Mark Lowes for thorough testing.
941 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
943 TK/03 Merged latest SRS patch from Miles Wilton.
945 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
946 with the definition in sysexits.h (which is #included earlier).
947 Fortunately, Exim does not actually use EX_OK. The code used to try to
948 preserve the sysexits.h value, by assumimg that macro definitions were
949 scanned for macro replacements. I have been disabused of this notion,
950 so now the code just undefines EX_OK before #including unistd.h.
952 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
953 in the smtp transport. When a block could not be written in a single
954 write() function, the timeout was being re-applied to each part-write.
955 This seems wrong - if the receiver was accepting one byte at a time it
956 would take for ever. The timeout is now adjusted when this happens. It
957 doesn't have to be particularly precise.
959 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
960 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
962 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
963 without a domain if the check to <postmaster@domain> fails.
965 SC/01 Eximstats: added -xls and the ability to specify output files
966 (patch written by Frank Heydlauf).
968 SC/02 Eximstats: use FileHandles for outputing results.
970 SC/03 Eximstats: allow any combination of xls, txt, and html output.
972 SC/04 Eximstats: fixed display of large numbers with -nvr option
974 SC/05 Eximstats: fixed merging of reports with empty tables.
976 SC/06 Eximstats: added the -include_original_destination flag
978 SC/07 Eximstats: removed tabs and trailing whitespace.
980 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
982 TK/06 MBOX spool code: Add real "From " MBOX separator line
983 so the .eml file is really in mbox format (even though
984 most programs do not really care). Patch from Alex Miller.
986 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
987 The latter is generated from $received_to and is only set if the
988 message has one envelope recipient. SA can use these headers,
989 obviously out-of-the-box. Patch from Alex Miller.
991 PH/08 The ${def test on a variable was returning false if the variable's
992 value was "0", contrary to what the specification has always said!
993 The result should be true unless the variable is empty.
995 PH/09 The syntax error of a character other than { following "${if
996 def:variable_name" (after optional whitespace) was not being diagnosed.
997 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
998 accidental colon was present, for example, could give incorrect results.
1000 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1001 result is used (not including appendfile, where other changes are about
1004 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1005 This involved changing a lot of size variables from int to off_t. It
1006 should work with maildirs and everything.
1008 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1009 spamd dying while we are connected to it.
1011 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1014 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1015 function for BDB 4.3.
1017 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1019 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1020 the sources for PCRE have been re-arranged and more files are now
1023 PH/15 The code I had for printing potentially long long variables in PH/11
1024 above was not the best (it lost precision). The length of off_t variables
1025 is now inspected at build time, and an appropriate printing format (%ld
1026 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1027 to be "long long int" or "long int". This is needed for the internal
1028 formatting function string_vformat().
1030 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1031 the configuration file to be ":syslog", then the script "guesses" where
1032 the logs files are, rather than using the compiled in default. In our
1033 case the guess is not the same as the compiled default, so the script
1034 suddenly stopped working when I started to use syslog. The patch checks
1035 to see if log_file_path is "". If so, it attempts to read it from exim
1036 with no configuration file to get the compiled in version, before it
1037 falls back to the previous guessing code."
1039 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1040 implementing BATV in an Exim configuration. See NewStuff for the gory
1043 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1044 Makefile that are specific to HP-UX.
1046 PH/18 If the "use_postmaster" option was set for a recipient callout together
1047 with the "random" option, the postmaster address was used as the MAIL
1048 FROM address for the random test, but not for the subsequent recipient
1049 test. It is now used for both.
1051 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1052 patch removes a few documentation additions to RFC 3028, because the
1053 latest draft now contains them. It adds the new en;ascii-case comparator
1054 and a new error check for 8bit text in MIME parts. Comparator and
1055 require names are now matched exactly. I enabled the subaddress
1056 extension, but it is not well tested yet (read: it works for me)."
1058 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1059 rework some of the code of TK/09 above to avoid the hardwired use of
1060 "%lld" and "long long". Replaced the call to snprintf() with a call to
1063 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1064 records point to non-existent hosts", "retry timeout exceeded", and
1065 "retry time not reached for any host after a long failure period".
1067 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1068 experimental DomainKeys support:
1070 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1071 (2) On an error such as an illegally used "control", the wrong name for
1072 the control was given.
1074 These problems did NOT occur unless DomainKeys support was compiled.
1076 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1078 PH/24 Added ${if match_ip condition.
1080 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1081 databases so that it will be absolutely obvious if a crash occurs in the
1082 DB library. This is a regular occurrence (often caused by mis-matched
1085 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1086 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1087 picked up on a user's system that detects such things. There doesn't seem
1088 to be a gcc warning option for this - only an attribute that has to be
1089 put on the function's prototype. It seems that in Fedora Core 4 they have
1090 set this on a number of new functions. No doubt there will be more in due
1093 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1094 specify a transport. However, if an address that was verified by such a
1095 router was the subject of a callout, Exim crashed because it tried to
1096 read the rcpt_include_affixes from the non-existent transport. Now it
1097 just assumes that the setting of that option is false. This bug was
1098 introduced by 4.51/PH/31.
1100 PH/28 Changed -d+all to exclude +memory, because that information is very
1101 rarely of interest, but it makes the output a lot bigger. People tend to
1102 do -d+all out of habit.
1104 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1105 code in os-type was giving problems when libc.so lives in lib64, like on
1108 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1109 aren't the modern standard, and it seems that some systems' include files
1110 don't always have them. Exim was already checking for some of the newer
1111 ones like T_AAAA, and defining it itself. I've added checks for all the
1112 record types that Exim uses.
1114 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1115 not automatically generating a new one, as it is supposed to. This
1116 prevented TLS from working. If the file did exist, but contained invalid
1117 data, a new version was generated, as expected. It was only the case of a
1118 non-existent file that was broken.
1120 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1121 with a change in libdomainkeys > 0.64.
1123 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1124 from DNS. If the selector record carries the flag, it now has
1125 precedence over the domain-wide flag.
1127 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1129 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1130 the use of a port name that isn't defined in /etc/services) occurred, the
1131 message was deferred as in a normal delivery, and thus remained on the
1132 spool, instead of being failed because of the mua_wrapper setting. This
1133 is now fixed, and I tidied up some of the mua_wrapper messages at the
1136 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1137 the messages in a hash of arrays rather than using individual hashes.
1138 This is a bit cleaner and results in dramatic memory savings, albeit
1139 at a slight CPU cost.
1141 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1142 as requested by Marc Sherman.
1144 SC/10 Eximstats: added histograms for user specified patterns as requested
1147 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1149 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1150 fopen() in the content-scanning modules that did not already have it.
1156 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1157 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1159 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1161 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1163 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1164 to test Sieve filters that use "vacation".
1166 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1167 that changes the way the GnuTLS parameters are stored in the cache file.
1168 The new format can be generated externally. For backward compatibility,
1169 if the data in the cache doesn't make sense, Exim assumes it has read an
1170 old-format file, and it generates new data and writes a new file. This
1171 means that you can't go back to an older release without removing the
1174 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1175 work if there are any delivery delays because "one_time" forces the
1176 parent to be marked "delivered", so its unseen clone is never tried
1177 again. For this reason, Exim now forbids the simultaneous setting of
1180 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1181 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1182 entry if you want to know the details. The fix, however, neglected to
1183 consider the case when local delivery batching is involved. The test for
1184 "previously delivered" was not happening when checking to see if an
1185 address could be batched with a previous (undelivered) one; under
1186 certain circumstances this could lead to multiple deliveries to the same
1189 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1190 in its include files, and this causes problems building Exim.
1192 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1193 header_syntax) but Exim was just ignoring anything given after a slash.
1194 In particular, this caused confusion with an attempt to use "verify =
1195 reverse_host_lookup/defer_ok". An error is now given when options are
1196 supplied for verify items that do not have them. (Maybe reverse_host_
1197 lookup should have a defer_ok option, but that's a different point.)
1199 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1200 defined by RFC 821) to 2048, because there were problems with some AUTH
1201 commands, and RFC 1869 says the size should be increased for extended
1202 SMTP commands that take arguments.
1204 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1207 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1208 "unknown" error; now it says that the functionality isn't in the binary.
1210 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1211 an address' error message when a string expansion fails (syntax or
1212 whatever). Otherwise the password may appear in the log. Following change
1213 PH/42 below, there is no longer a chance of it appearing in a bounce
1216 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1218 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1219 of its addresses was ever tried. (Bugzilla bug #2.)
1221 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1222 the result incorrectly in the debug output. (It correctly added a newline
1223 to what was transported.)
1225 TF/01 Added $received_time.
1227 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1228 commented out examples of how to interface to a virus scanner and to
1229 SpamAssassin. Also added commented examples of av_scanner and
1230 spamd_address settings.
1232 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1233 and controls are allowed in which ACLs. There were a couple of minor
1234 errors. Some of the entries in the conditions table (which is a table of
1235 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1236 negation of where the condition IS allowed.
1238 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1240 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1241 header file does not have a version number, so I've had to invent a new
1242 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1243 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1244 radiusclient), but it was contributed by a Radius user.
1246 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1247 files or directories, for OpenSSL.
1249 PH/19 When an Exim process that is running as root has to create an Exim log
1250 file, it does so in a subprocess that runs as exim:exim so as to get the
1251 ownership right at creation (otherwise, other Exim processes might see
1252 the file with the wrong ownership). There was no test for failure of this
1253 fork() call, which would lead to the process getting stuck as it waited
1254 for a non-existent subprocess. Forks do occasionally fail when resources
1255 run out. I reviewed all the other calls to fork(); they all seem to check
1258 PH/20 When checking for unexpected SMTP input at connect time (before writing
1259 the banner), Exim was not dealing correctly with a non-positive return
1260 from the read() function. If the client had disconnected by this time,
1261 the result was a log entry for a synchronization error with an empty
1262 string after "input=" when read() returned zero. If read() returned -1
1263 (an event I could not check), uninitialized data bytes were printed.
1264 There were reports of junk text (parts of files, etc) appearing after
1267 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1269 PH/22 Added support for macro redefinition, and (re)definition in between
1270 driver and ACL definitions.
1272 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1273 forgetting to use the resulting value; it was using the unexpanded value.
1275 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1276 hadn't been configured. The fix is from Juergen Kreileder, who
1277 understands it better than I do:
1279 "Here's what I see happening with three configured cyrus_sasl
1280 authenticators configured (plain, login, cram-md5):
1282 On startup auth_cyrus_sasl_init() gets called for each of these.
1283 This means three calls to sasl_listmech() without a specified mech_list.
1284 => SASL tests which mechs of all available mechs actually work
1285 => three warnings about OTP not working
1286 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1288 With the patch, sasl_listmech() also gets called three times. But now
1289 SASL's mech_list option is set to the server_mech specified in the the
1290 authenticator. Or in other words, the answer from sasl_listmech()
1291 gets limited to just the mech you're testing for (which is different
1293 => the return list contains just 'plain' or 'login', 'cram-md5' or
1294 nothing depending on the value of ob->server_mech.
1296 I've just tested the patch: Authentication still works fine,
1297 unavailable mechs specified in the exim configuration are still
1298 caught, and the auth.log warnings about OTP are gone."
1300 PH/25 When debugging is enabled, the contents of the command line are added
1301 to the debugging output, even when log_selector=+arguments is not
1304 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1305 answer is "GNU", and only if the return is "GNU/something" is the answer
1308 PH/27 $acl_verify_message is now set immediately after the failure of a
1309 verification in an ACL, and so is available in subsequent modifiers. In
1310 particular, the message can be preserved by coding like this:
1312 warn !verify = sender
1313 set acl_m0 = $acl_verify_message
1315 Previously, $acl_verify_message was set only while expanding "message"
1316 and "log_message" when a very denied access.
1318 PH/28 Modified OS/os.c-Linux with
1320 -#ifndef OS_LOAD_AVERAGE
1321 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1323 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1324 nomenclature these days.)
1326 PH/29 Installed patch from the Sieve maintainer that adds the options
1327 sieve_useraddress and sieve_subaddress to the redirect router.
1329 PH/30 In these circumstances:
1330 . Two addresses routed to the same list of hosts;
1331 . First host does not offer TLS;
1332 . First host accepts first address;
1333 . First host gives temporary error to second address;
1334 . Second host offers TLS and a TLS session is established;
1335 . Second host accepts second address.
1336 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1337 and peerdn, if requested) that were in fact used only for the second
1340 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1341 attention to any local part prefix or suffix that was matched by the
1342 router that accepted the address. It now behaves in the same way as it
1343 does for delivery: the affixes are removed from the local part unless
1344 rcpt_include_affixes is set on the transport.
1346 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1347 timeout during the DATA phase of an incoming message.
1349 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1350 applied a patch sanctioned by the Sieve maintainer.
1352 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1353 the uid or gid is negative. A case of a negative gid caused this to be
1354 noticed. The fix allows for either to be negative.
1356 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1357 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1360 PH/36 The change PH/12 above was broken. Fixed it.
1362 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1363 the grounds that routing the same address twice would always produce the
1364 same answer. This might have been true once, but it is certainly no
1365 longer true now. Routing a child address may depend on the previous
1366 routing that produced that child. Some complicated redirection strategies
1367 went wrong when messages had multiple recipients, and made Exim's
1368 behaviour dependent on the order in which the addresses were given.
1370 I have moved the duplicate checking until after the routing is complete.
1371 Exim scans the addresses that are assigned to local and remote
1372 transports, and removes any duplicates. This means that more work will be
1373 done, as duplicates will always all be routed, but duplicates are
1374 presumably rare, so I don't expect this is of any significance.
1376 For deliveries to pipes, files, and autoreplies, the duplicate checking
1377 still happens during the routing process, since they are not going to be
1380 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1381 It corrects a timeout issue with spamd. This is Ian's comment: "The
1382 background is that sometimes spamd either never reads data from a
1383 connection it has accepted, or it never writes response data. The exiscan
1384 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1385 blindly assumes that writes won't block so it may never time out."
1387 PH/39 Allow G after quota size as well as K and M.
1389 PH/40 The value set for $authenticated_id in an authenticator may not contain
1390 binary zeroes or newlines because the value is written to log lines and
1391 to spool files. There was no check on this. Now the value is run through
1392 the string_printing() function so that such characters are converted to
1393 printable escape sequences.
1395 PH/41 $message_linecount is a new variable that contains the total number of
1396 lines in the message. Compare $body_linecount, which is the count for the
1399 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1400 bounce and delay warning messages, except in certain special cases, which
1403 (a) An SMTP error message from a remote host;
1404 (b) A message specified in a :fail: redirection;
1405 (c) A message specified in a "fail" command in a system filter;
1406 (d) A message specified in a FAIL return from the queryprogram router;
1407 (e) A message specified by the cannot_route_message router option.
1409 In these cases only, Exim does include the error details in bounce and
1410 warning messages. There are also a few cases where bland messages such
1411 as "unrouteable address" or "local delivery error" are given.
1413 PH/43 $value is now also set for the "else" part of a ${run expansion.
1415 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1416 being worked on, but at least Exim now implements the latest version to
1419 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1420 process to complete was treated as a delivery failure, a timeout while
1421 writing the message to the pipe was logged, but erroneously treated as a
1422 successful delivery. Such timeouts include transport filter timeouts. For
1423 consistency with the overall process timeout, these timeouts are now
1424 treated as errors, giving rise to delivery failures by default. However,
1425 there is now a new Boolean option for the pipe transport called
1426 timeout_defer, which, if set TRUE, converts the failures into defers for
1427 both kinds of timeout. A transport filter timeout is now identified in
1430 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1431 systems where "make" and "gmake" are different, calling "gmake" at top
1432 level broke things. I've arranged for the value of $(MAKE) to be passed
1433 from the Makefile to this script so that it can call the same version of
1437 A note about Exim versions 4.44 and 4.50
1438 ----------------------------------------
1440 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1441 changes of various kinds. As a consequence, a big documentation update was
1442 needed. This delayed the release for rather longer than seemed good, especially
1443 in the light of a couple of (minor) security issues. Therefore, the changes
1444 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1445 release. So 4.44 and 4.50 are in effect two different branches that both start
1448 I have left the 4.50 change log unchanged; it contains all the changes since
1449 4.43. The change log for 4.44 is below; many of its items are identical to
1450 those for 4.50. This seems to be the most sensible way to preserve the
1451 historical information.
1457 1. Minor wording change to the doc/README.SIEVE file.
1459 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1460 computation of the current number of files was incorrect.
1462 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1463 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1466 4. Give more explanation in the error message when the command for a transport
1467 filter fails to execute.
1469 5. There are several places where Exim runs a non-Exim command in a
1470 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1471 was being done only for the command run by the queryprogram router. It is
1472 now done for all such subprocesses. The other cases are: ${run, transport
1473 filters, and the commands run by the lmtp and pipe transports.
1475 6. Added CONFIGURE_GROUP build-time option.
1477 7. Some older OS have a limit of 256 on the maximum number of file
1478 descriptors. Exim was using setrlimit() to set 1000 as a large value
1479 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1480 systems. I've change it so that if it can't get 1000, it tries for 256.
1482 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1483 was an oversight, and furthermore, ever since the addition of extra
1484 controls (e.g. 4.43/32), the checks on when to allow different forms of
1485 "control" were broken. There should now be diagnostics for all cases when a
1486 control that does not make sense is encountered.
1488 9. Added the /retain_sender option to "control=submission".
1490 10. $recipients is now available in the predata ACL (oversight).
1492 11. Tidy the search cache before the fork to do a delivery from a message
1493 received from the command line. Otherwise the child will trigger a lookup
1494 failure and thereby defer the delivery if it tries to use (for example) a
1495 cached ldap connection that the parent has called unbind on.
1497 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1498 of $address_data from the recipient verification was clobbered by the
1499 sender verification.
1501 13. The value of address_data from a sender verification is now available in
1502 $sender_address_data in subsequent conditions in the ACL statement.
1504 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1506 15. Added a new option "connect=<time>" to callout options, to set a different
1509 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1510 was its contents. (It was OK if the option was not defined at all.)
1512 17. A "Completed" log line is now written for messages that are removed from
1513 the spool by the -Mrm option.
1515 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1516 information about exactly what failed.
1518 19. Added -dd to debug only the daemon process.
1520 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1521 handles timeouts, both on the server side and network timeouts. Renamed the
1522 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1524 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1525 It is now set to "smtps".
1527 22. $host_address is now set to the target address during the checking of
1528 ignore_target_hosts.
1530 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1531 being passed; this would have caused $sender_host_name to have been used if
1532 matching the list had actually called for a host name (not very likely,
1533 since this list is usually IP addresses). A host name is now passed as
1536 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1537 code that specifies a non-restarting handler (typically sigaction() in
1538 modern systems) in an attempt to fix a rare and obscure crash bug.
1540 25. Narrowed the window for a race in the daemon that could cause it to ignore
1541 SIGCHLD signals. This is not a major problem, because they are used only to
1542 wake it up if nothing else does.
1544 26. A malformed maildirsize file could cause Exim to calculate negative values
1545 for the mailbox size or file count. Odd effects could occur as a result.
1546 The maildirsize information is now recalculated if the size or filecount
1549 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1550 support for a long time. Removed HAVE_SYS_VFS_H.
1552 28. Installed the latest version of exipick from John Jetmore.
1554 29. In an address list, if the pattern was not a regular expression, an empty
1555 subject address (from a bounce message) matched only if the pattern was an
1556 empty string. Non-empty patterns were not even tested. This was the wrong
1557 because it is perfectly reasonable to use an empty address as part of a
1558 database query. An empty address is now tested by patterns that are
1559 lookups. However, all the other forms of pattern expect the subject to
1560 contain a local part and a domain, and therefore, for them, an empty
1561 address still always fails if the pattern is not itself empty.
1563 30. Exim went into a mad DNS loop when attempting to do a callout where the
1564 host was specified on an smtp transport, and looking it up yielded more
1565 than one IP address.
1567 31. Re-factored the code for checking spool and log partition space into a
1568 function that finds that data and another that does the check. The former
1569 is then used to implement four new variables: $spool_space, $log_space,
1570 $spool_inodes, and $log_inodes.
1572 32. The RFC2047 encoding function was originally intended for short strings
1573 such as real names; it was not keeping to the 75-character limit for
1574 encoded words that the RFC imposes. It now respects the limit, and
1575 generates multiple encoded words if necessary. To be on the safe side, I
1576 have increased the buffer size for the ${rfc2047: expansion operator from
1579 33. It is now permitted to omit both strings after an "if" condition; if the
1580 condition is true, the result is "true". As before, when the second string
1581 is omitted, a false condition yields an empty string. This makes it less
1582 cumbersome to write custom ACL and router conditions.
1584 34. Failure to deliver a bounce message always caused it to be frozen, even if
1585 there was an errors_to setting on the router. The errors_to setting is now
1588 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1589 canonical form (fully expanded) before being placed in
1590 $sender_host_address.
1592 36. The table in the code that translates DNS record types into text (T_A to
1593 "A" for instance) was missing entries for NS and CNAME. It is just possible
1594 that this could have caused confusion if both these types were looked up
1595 for the same domain, because the text type is used as part of Exim's
1596 per-process caching. But the chance of anyone hitting this buglet seems
1599 37. The dnsdb lookup has been extended in a number of ways.
1601 (1) There is a new type, "zns", which walks up the domain tree until it
1602 finds some nameserver records. It should be used with care.
1604 (2) There is a new type, "mxh", which is like "mx" except that it returns
1605 just the host names, not the priorities.
1607 (3) It is now possible to give a list of domains (or IP addresses) to be
1608 looked up. The behaviour when one of the lookups defers can be
1609 controlled by a keyword.
1611 (4) It is now possible to specify the separator character for use when
1612 multiple records are returned.
1614 38. The dnslists ACL condition has been extended: it is now possible to supply
1615 a list of IP addresses and/or domains to be looked up in a particular DNS
1618 39. Added log_selector=+queue_time_overall.
1620 40. When running the queue in the test harness, wait just a tad after forking a
1621 delivery process, to get repeatability of debugging output.
1623 41. Include certificate and key file names in error message when GnuTLS fails
1624 to set them up, because the GnuTLS error message doesn't include the name
1625 of the failing file when there is a problem reading it.
1627 42. Allow both -bf and -bF in the same test run.
1629 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1631 44. The "Exiscan patch" is now merged into the mainline Exim source.
1633 45. Sometimes the final signoff response after QUIT could fail to get
1634 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1635 < 0 before doing a fflush(). This bug looks as though it goes back to the
1636 introduction of TLS in release 3.20, but "sometimes" must have been rare
1637 because the tests only now provoked it.
1639 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1640 (this can affect the format of dates).
1642 47. exim_tidydb, when checking for the continued existence of a message for
1643 which it has found a message-specific retry record, was not finding
1644 messages that were in split spool directories. Consequently, it was
1645 deleting retry records that should have stayed in existence.
1647 48. Steve fixed some bugs in eximstats.
1649 49. The SPA authentication driver was not abandoning authentication and moving
1650 on to the next authenticator when an expansion was forced to fail,
1651 contradicting the general specification for all authenticators. Instead it
1652 was generating a temporary error. It now behaves as specified.
1654 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1655 (the order specifies the preference for clients). The order is now AES256,
1656 AES128, 3DES, ARCFOUR128.
1658 51. Small patch to Sieve code - explicitly set From: when generating an
1661 52. Exim crashed if a remote delivery caused a very long error message to be
1662 recorded - for instance if somebody sent an entire SpamAssassin report back
1663 as a large number of 550 error lines. This bug was coincidentally fixed by
1664 increasing the size of one of Exim's internal buffers (big_buffer) that
1665 happened as part of the Exiscan merge. However, to be on the safe side, I
1666 have made the code more robust (and fixed the comments that describe what
1669 53. Now that there can be additional text after "Completed" in log lines (if
1670 the queue_time_overall log selector is set), a one-byte patch to exigrep
1671 was needed to allow it to recognize "Completed" as not the last thing in
1674 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1675 patch that reportedly fixes this has been added. I am not expert enough to
1676 create a test for it. This is what the patch creator wrote:
1678 "I found a little strange behaviour of ldap code when working with
1679 Windows 2003 AD Domain, where users was placed in more than one
1680 Organization Units. When I tried to give exim partial DN, the exit code
1681 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1682 But simultaneously result of request was absolutely normal ldap result,
1683 so I produce this patch..."
1685 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1686 REFERENCE, so I have modified the code to exclude the patch when that macro
1689 55. Some experimental protocols are using DNS PTR records for new purposes. The
1690 keys for these records are domain names, not reversed IP addresses. The
1691 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1692 leaves it alone. Component reversal etc. now happens only for IP addresses.
1695 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1697 57. Double the size of the debug message buffer (to 2048) so that more of very
1698 long debug lines gets shown.
1700 58. The exicyclog utility now does better if the number of log files to keep
1701 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1703 59. Two changes related to the smtp_active_hostname option:
1705 (1) $smtp_active_hostname is now available as a variable.
1706 (2) The default for smtp_banner uses $smtp_active_hostname instead
1707 of $primary_hostname.
1709 60. The host_aton() function is supposed to be passed a string that is known
1710 to be a valid IP address. However, in the case of IPv6 addresses, it was
1711 not checking this. This is a hostage to fortune. Exim now panics and dies
1712 if the condition is not met. A case was found where this could be provoked
1713 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1714 components; fortuitously, this particular loophole had already been fixed
1715 by change 4.50/55 above.
1717 If there are any other similar loopholes, the new check in host_aton()
1718 itself should stop them being exploited. The report I received stated that
1719 data on the command line could provoke the exploit when Exim was running as
1720 exim, but did not say which command line option was involved. All I could
1721 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1722 running as the user.
1725 61. There was a buffer overflow vulnerability in the SPA authentication code
1726 (which came originally from the Samba project). I have added a test to the
1727 spa_base64_to_bits() function which I hope fixes it.
1730 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1731 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1733 63. The daemon start-up calls getloadavg() while still root for those OS that
1734 need the first call to be done as root, but it missed one case: when
1735 deliver_queue_load_max is set with deliver_drop_privilege. This is
1736 necessary for the benefit of the queue runner, because there is no re-exec
1737 when deliver_drop_privilege is set.
1739 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1740 This has been fixed.
1742 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1743 was in use, was not putting the data itself into the right store pool;
1744 consequently, it could be overwritten for a subsequent message in the same
1745 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1748 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1750 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1751 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1752 the function were treating the return as a boolean value, which happened to
1753 work because 0=false and not-0=true, but is not correct code.
1755 68. The host_aton() function was not handling scoped IPv6 addresses (those
1756 with, for example, "%eth0" on the end) correctly.
1758 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1759 negated items (that is, ~something) in unsigned ints. Some compilers
1760 apparently mutter when there is no cast.
1762 70. If an address verification called from an ACL failed, and did not produce a
1763 user-specific message (i.e. there was only a "system" message), nothing was
1764 put in $acl_verify_message. In this situation, it now puts the system
1767 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1768 session; change 4.31/43 added the unwanted input to the log line - except
1769 that it did not do this in the start of session case. It now does.
1771 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1772 This is wrong and can cause the other end to generate a synchronization
1773 error if it is another Exim or anything else that does the synchronization
1774 check. A QUIT command is no longer sent after a timeout.
1776 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1777 during host lookups.
1779 74. The defer_ok option of callout verification was not working if it was used
1780 when verifying addresses in header lines, that is, for this case:
1782 verify = header_sender/callout=defer_ok
1784 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1785 those file descriptors could be used for SMTP connections. If anything
1786 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1787 could be sent to the SMTP client, causing chaos. The daemon now opens
1788 stdin, stdout, and stderr to /dev/null when it puts itself into the
1791 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1792 log by default. The user can override this with suitable Perl magic.
1794 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1795 the log message when discard triggers, was not working for the DATA ACL or
1796 for the non-SMTP ACL.
1798 78. Error message wording change in sieve.c.
1800 79. If smtp_accept_max_per_host was set, the number of connections could be
1801 restricted to fewer than expected, because the daemon was trying to set up
1802 a new connection before checking whether the processes handling previous
1803 connections had finished. The check for completed processes is now done
1804 earlier. On busy systems, this bug wouldn't be noticed because something
1805 else would have woken the daemon, and it would have reaped the completed
1808 80. If a message was submitted locally by a user whose login name contained one
1809 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1810 It caused a spool format error. I have fixed the spool reading code. A
1811 related problem was that the "from" clause in the Received: line became
1812 illegal because of the space(s). It is now covered by ${quote_local_part.
1814 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1817 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1819 83. Patch from Sieve maintainer to fix unterminated string problem in
1820 "vacation" handling.
1822 84. Some minor changes to the Linux configuration files to help with other
1823 OS variants using glibc.
1825 85. One more patch for Sieve to update vacation handling to latest spec.
1828 ----------------------------------------------------
1829 See the note above about the 4.44 and 4.50 releases.
1830 ----------------------------------------------------
1836 1. Change 4.43/35 introduced a bug that caused file counts to be
1837 incorrectly computed when quota_filecount was set in an appendfile
1840 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1841 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1844 3. Give more explanation in the error message when the command for a transport
1845 filter fails to execute.
1847 4. There are several places where Exim runs a non-Exim command in a
1848 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1849 was being done only for the command run by the queryprogram router. It is
1850 now done for all such subprocesses. The other cases are: ${run, transport
1851 filters, and the commands run by the lmtp and pipe transports.
1853 5. Some older OS have a limit of 256 on the maximum number of file
1854 descriptors. Exim was using setrlimit() to set 1000 as a large value
1855 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1856 systems. I've change it so that if it can't get 1000, it tries for 256.
1858 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1859 was an oversight, and furthermore, ever since the addition of extra
1860 controls (e.g. 4.43/32), the checks on when to allow different forms of
1861 "control" were broken. There should now be diagnostics for all cases when a
1862 control that does not make sense is encountered.
1864 7. $recipients is now available in the predata ACL (oversight).
1866 8. Tidy the search cache before the fork to do a delivery from a message
1867 received from the command line. Otherwise the child will trigger a lookup
1868 failure and thereby defer the delivery if it tries to use (for example) a
1869 cached ldap connection that the parent has called unbind on.
1871 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1872 of $address_data from the recipient verification was clobbered by the
1873 sender verification.
1875 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1876 was its contents. (It was OK if the option was not defined at all.)
1878 11. A "Completed" log line is now written for messages that are removed from
1879 the spool by the -Mrm option.
1881 12. $host_address is now set to the target address during the checking of
1882 ignore_target_hosts.
1884 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1885 being passed; this would have caused $sender_host_name to have been used if
1886 matching the list had actually called for a host name (not very likely,
1887 since this list is usually IP addresses). A host name is now passed as
1890 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1891 code that specifies a non-restarting handler (typically sigaction() in
1892 modern systems) in an attempt to fix a rare and obscure crash bug.
1894 15. Narrowed the window for a race in the daemon that could cause it to ignore
1895 SIGCHLD signals. This is not a major problem, because they are used only to
1896 wake it up if nothing else does.
1898 16. A malformed maildirsize file could cause Exim to calculate negative values
1899 for the mailbox size or file count. Odd effects could occur as a result.
1900 The maildirsize information is now recalculated if the size or filecount
1903 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1904 support for a long time. Removed HAVE_SYS_VFS_H.
1906 18. Updated exipick to current release from John Jetmore.
1908 19. Allow an empty sender to be matched against a lookup in an address list.
1909 Previously the only cases considered were a regular expression, or an
1912 20. Exim went into a mad DNS lookup loop when doing a callout where the
1913 host was specified on the transport, if the DNS lookup yielded more than
1916 21. The RFC2047 encoding function was originally intended for short strings
1917 such as real names; it was not keeping to the 75-character limit for
1918 encoded words that the RFC imposes. It now respects the limit, and
1919 generates multiple encoded words if necessary. To be on the safe side, I
1920 have increased the buffer size for the ${rfc2047: expansion operator from
1923 22. Failure to deliver a bounce message always caused it to be frozen, even if
1924 there was an errors_to setting on the router. The errors_to setting is now
1927 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1928 canonical form (fully expanded) before being placed in
1929 $sender_host_address.
1931 24. Updated eximstats to version 1.33
1933 25. Include certificate and key file names in error message when GnuTLS fails
1934 to set them up, because the GnuTLS error message doesn't include the name
1935 of the failing file when there is a problem reading it.
1937 26. Expand error message when OpenSSL has problems setting up cert/key files.
1940 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1941 (this can affect the format of dates).
1943 28. exim_tidydb, when checking for the continued existence of a message for
1944 which it has found a message-specific retry record, was not finding
1945 messages that were in split spool directories. Consequently, it was
1946 deleting retry records that should have stayed in existence.
1948 29. eximstats updated to version 1.35
1949 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1950 1.35 - bugfix such that pie charts by volume are generated correctly
1952 30. The SPA authentication driver was not abandoning authentication and moving
1953 on to the next authenticator when an expansion was forced to fail,
1954 contradicting the general specification for all authenticators. Instead it
1955 was generating a temporary error. It now behaves as specified.
1957 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1958 (the order specifies the preference for clients). The order is now AES256,
1959 AES128, 3DES, ARCFOUR128.
1961 31. Small patch to Sieve code - explicitly set From: when generating an
1964 32. Exim crashed if a remote delivery caused a very long error message to be
1965 recorded - for instance if somebody sent an entire SpamAssassin report back
1966 as a large number of 550 error lines. This bug was coincidentally fixed by
1967 increasing the size of one of Exim's internal buffers (big_buffer) that
1968 happened as part of the Exiscan merge. However, to be on the safe side, I
1969 have made the code more robust (and fixed the comments that describe what
1972 33. Some experimental protocols are using DNS PTR records for new purposes. The
1973 keys for these records are domain names, not reversed IP addresses. The
1974 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1975 leaves it alone. Component reversal etc. now happens only for IP addresses.
1978 34. The host_aton() function is supposed to be passed a string that is known
1979 to be a valid IP address. However, in the case of IPv6 addresses, it was
1980 not checking this. This is a hostage to fortune. Exim now panics and dies
1981 if the condition is not met. A case was found where this could be provoked
1982 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1983 components; fortuitously, this particular loophole had already been fixed
1984 by change 4.50/55 or 4.44/33 above.
1986 If there are any other similar loopholes, the new check in host_aton()
1987 itself should stop them being exploited. The report I received stated that
1988 data on the command line could provoke the exploit when Exim was running as
1989 exim, but did not say which command line option was involved. All I could
1990 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1991 running as the user.
1994 35. There was a buffer overflow vulnerability in the SPA authentication code
1995 (which came originally from the Samba project). I have added a test to the
1996 spa_base64_to_bits() function which I hope fixes it.
1999 36. The daemon start-up calls getloadavg() while still root for those OS that
2000 need the first call to be done as root, but it missed one case: when
2001 deliver_queue_load_max is set with deliver_drop_privilege. This is
2002 necessary for the benefit of the queue runner, because there is no re-exec
2003 when deliver_drop_privilege is set.
2005 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2006 was in use, was not putting the data itself into the right store pool;
2007 consequently, it could be overwritten for a subsequent message in the same
2008 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2011 38. Sometimes the final signoff response after QUIT could fail to get
2012 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2013 < 0 before doing a fflush(). This bug looks as though it goes back to the
2014 introduction of TLS in release 3.20, but "sometimes" must have been rare
2015 because the tests only now provoked it.
2021 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2022 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2023 Now it returns an integer. A number of calls were still expecting a T/F
2024 return. Fortuitously, in all cases, the tests worked in OK situations,
2025 which is the norm. However, things would have gone wrong on any write
2026 failures on the smtp file descriptor. This function is used when sending
2027 messages over SMTP and also when doing verify callouts.
2029 2. When Exim is called to do synchronous delivery of a locally submitted
2030 message (the -odf or -odi options), it no longer closes stderr before doing
2033 3. Implemented the mua_wrapper option.
2035 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2037 5. Implemented the functions header_remove(), header_testname(),
2038 header_add_at_position(), and receive_remove_recipient(), and exported them
2041 6. If an ACL "warn" statement specified the addition of headers, Exim already
2042 inserted X-ACL-Warn: at the start if there was no header name. However, it
2043 was not making this test for the second and subsequent header lines if
2044 there were newlines in the string. This meant that an invalid header could
2045 be inserted if Exim was badly configured.
2047 7. Allow an ACL "warn" statement to add header lines at the start or after all
2048 the Received: headers, as well as at the end.
2050 8. Added the rcpt_4xx retry error code.
2052 9. Added postmaster_mailfrom=xxx to callout verification option.
2054 10. Added mailfrom=xxxx to the callout verification option, for verify=
2057 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2058 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2060 12. Inserted some casts to stop certain compilers warning when using pointer
2061 differences as field lengths or precisions in printf-type calls (mostly
2062 affecting debugging statements).
2064 13. Added optional readline() support for -be (dynamically loaded).
2066 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2067 same clock tick as a message's arrival, so that its received time was the
2068 same as the "first fail" time on the retry record, and that message
2069 remained on the queue past the ultimate address timeout, every queue runner
2070 would try a delivery (because it was past the ultimate address timeout) but
2071 after another failure, the ultimate address timeout, which should have then
2072 bounced the address, did not kick in. This was a "< instead of <=" error;
2073 in most cases the first failure would have been in the next clock tick
2074 after the received time, and all would be well.
2076 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2077 being recognized when the domain list was tested by the match_domain
2078 condition in an expansion string.
2080 16. Added the ${str2b64: operator.
2082 17. Exim was always calling setrlimit() to set a large limit for the number of
2083 processes, without checking whether the existing limit was already
2084 adequate. (It did check for the limit on file descriptors.) Furthermore,
2085 errors from getrlimit() and setrlimit() were being ignored. Now they are
2086 logged to the main and panic logs, but Exim does carry on, to try to do its
2087 job under whatever limits there are.
2089 18. Imported PCRE 5.0.
2091 19. Trivial typo in log message " temporarily refused connection" (the leading
2094 20. If the log selector return_path_on_delivery was set and an address was
2095 redirected to /dev/null, the delivery process crashed because it assumed
2096 that a return path would always be set for a "successful" delivery. In this
2097 case, the whole delivery is bypassed as an optimization, and therefore no
2100 21. Internal re-arrangement: the function for sending a challenge and reading
2101 a response while authentication was assuming a zero-terminated challenge
2102 string. It's now changed to take a pointer and a length, to allow for
2103 binary data in such strings.
2105 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2107 23. Exim was not respecting finduser_retries when seeking the login of the
2108 uid under which it was called; it was always trying 10 times. (The default
2109 setting of finduser_retries is zero.) Also, it was sleeping after the final
2110 failure, which is pointless.
2112 24. Implemented tls_on_connect_ports.
2114 25. Implemented acl_smtp_predata.
2116 26. If the domain in control=submission is set empty, Exim assumes that the
2117 authenticated id is a complete email address when it generates From: or
2118 Sender: header lines.
2120 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2121 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2122 chown and chgrp in /bin and hostname in /usr/bin.
2124 28. Exim was keeping the "process log" file open after each use, just as it
2125 does for the main log. This opens the possibility of it remaining open for
2126 long periods when the USR1 signal hits a daemon. Occasional processlog
2127 errors were reported, that could have been caused by this. Anyway, it seems
2128 much more sensible not to leave this file open at all, so that is what now
2131 29. The long-running daemon process does not normally write to the log once it
2132 has entered its main loop, and it closes the log before doing so. This is
2133 so that log files can straightforwardly be renamed and moved. However,
2134 there are a couple of unusual error situations where the daemon does write
2135 log entries, and I had neglected to close the log afterwards.
2137 30. The text of an SMTP error response that was received during a remote
2138 delivery was being truncated at 512 bytes. This is too short for some of
2139 the long messages that one sometimes sees. I've increased the limit to
2142 31. It is now possible to make retry rules that apply only when a message has a
2143 specific sender, in particular, an empty sender.
2145 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2146 it possible to be selective about when SMTP synchronization is enforced.
2148 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2150 32. Implemented hosts_connection_nolog.
2152 33. Added an ACL for QUIT.
2154 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2157 35. Added mailbox_size and mailbox_filecount to appendfile.
2159 36. Added control = no_multiline_responses to ACLs.
2161 37. There was a bug in the logic of the code that waits for the clock to tick
2162 in the case where the clock went backwards by a substantial amount such
2163 that the microsecond fraction of "now" was more than the microsecond
2164 fraction of "then" (but the whole seconds number was less).
2166 38. Added support for the libradius Radius client library this is found on
2167 FreeBSD (previously only the radiusclient library was supported).
2173 1. When certain lookups returned multiple values in the form name=value, the
2174 quoting of the values was not always being done properly. Specifically:
2175 (a) If the value started with a double quote, but contained no whitespace,
2177 (b) If the value contained whitespace other than a space character (i.e.
2178 tabs or newlines or carriage returns) it was not quoted.
2179 This fix has been applied to the mysql and pgsql lookups by writing a
2180 separate quoting function and calling it from the lookup code. The fix
2181 should probably also be applied to nisplus, ibase and oracle lookups, but
2182 since I cannot test any of those, I have not disturbed their existing code.
2184 2. A hit in the callout cache for a specific address caused a log line with no
2185 reason for rejecting RCPT. Now it says "Previous (cached) callout
2186 verification failure".
2188 3. There was an off-by-one bug in the queryprogram router. An over-long
2189 return line was truncated at 256 instead of 255 characters, thereby
2190 overflowing its buffer with the terminating zero. As well as fixing this, I
2191 have increased the buffer size to 1024 (and made a note to document this).
2193 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2194 when Exim is waiting for an SMTP response from a remote server, Exim
2195 restarts its select() call on the socket, thereby resetting its timeout.
2196 This is not a problem when such interrupts are rare. Somebody set up a cron
2197 job to run exiwhat every 2 minutes, which is less than the normal select()
2198 timeout (5 or 10 minutes). This meant that the select() timeout never
2199 kicked in because it was always reset. I have fixed this by comparing the
2200 time when an interrupt arrives with the time at the start of the first call
2201 to select(). If more time than the timeout has elapsed, the interrupt is
2202 treated as a timeout.
2204 5. Some internal re-factoring in preparation for the addition of Sieve
2205 extensions (by MH). In particular, the "personal" test is moved to a
2206 separate function, and given an option for scanning Cc: and Bcc: (which is
2207 not set for Exim filters).
2209 6. When Exim created an email address using the login of the caller as the
2210 local part (e.g. when creating a From: or Sender: header line), it was not
2211 quoting the local part when it contained special characters such as @.
2213 7. Installed new OpenBSD configuration files.
2215 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2216 try to make them clearer.
2218 9. Callout options, other than the timeout value, were being ignored when
2219 verifying sender addresses in header lines. For example, when using
2221 verify = header_sender/callout=no_cache
2223 the cache was (incorrectly) being used.
2225 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2226 only the Cygwin environment.
2228 11. When return_path_on_delivery was set as a log selector, if different remote
2229 addresses in the same message used different return paths and parallel
2230 remote delivery occurred, the wrong values would sometimes be logged.
2231 (Whenever a remote delivery process finished, the return path value from
2232 the most recently started remote delivery process was logged.)
2234 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2235 lines when AUTH and/or TLS are in use. This is the "received protocol"
2236 field. Exim used to use "asmtp" for authenticated SMTP, without any
2237 indication (in the protocol name) for TLS use. Now it follows the RFC and
2238 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2239 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2240 names appear in log lines as well as in Received: header lines.
2242 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2243 extensions, and comparison tests, and to fix some bugs.
2245 14. Changes to the "personal" filter test:
2247 (1) The test was buggy in that it was just doing the equivalent of
2248 "contains" tests on header lines. For example, if a user's address was
2249 anne@some.where, the "personal" test would incorrectly be true for
2251 To: susanne@some.where
2253 This test is now done by extracting each address from the header in turn,
2254 and checking the entire address. Other tests that are part of "personal"
2255 are now done using regular expressions (for example, to check local parts
2256 of addresses in From: header lines).
2258 (2) The list of non-personal local parts in From: addresses has been
2259 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2260 taken from the Sieve specification recommendations.
2262 (3) If the message contains any header line starting with "List-" it is
2263 treated as non-personal.
2265 (4) The test for "circular" in the Subject: header line has been removed
2266 because it now seems ill-conceived.
2268 15. Minor typos in src/EDITME comments corrected.
2270 16. Installed latest exipick from John Jetmore.
2272 17. If headers_add on a router specified a text string that was too long for
2273 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2274 of string_sprintf() is now avoided.
2276 18. $message_body_size was not set (it was always zero) when running the DATA
2277 ACL and the local_scan() function.
2279 19. For the "mail" command in an Exim filter, no default was being set for
2280 the once_repeat time, causing a random time value to be used if "once" was
2281 specified. (If the value happened to be <= 0, no repeat happened.) The
2282 default is now 0s, meaning "never repeat". The "vacation" command was OK
2283 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2284 (I found it when inspecting the code).
2286 20. There is now an overall timeout for performing a callout verification. It
2287 defaults to 4 times the callout timeout, which applies to individual SMTP
2288 commands during the callout. The overall timeout applies when there is more
2289 than one host that can be tried. The timeout is checked before trying the
2290 next host. This prevents very long delays if there are a large number of
2291 hosts and all are timing out (e.g. when the network connections are timing
2292 out). The value of the overall timeout can be changed by specifying an
2293 additional sub-option for "callout", called "maxwait". For example:
2295 verify = sender/callout=5s,maxwait=20s
2297 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2298 to the end before writing, but this should make it even safer).
2300 22. Exim was forgetting that it had advertised PIPELINING for the second and
2301 subsequent messages on an SMTP connection. It was also not resetting its
2302 memory on STARTTLS and an internal HELO.
2304 23. When Exim logs an SMTP synchronization error within a session, it now
2305 records whether PIPELINING has been advertised or not.
2307 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2308 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2309 rather than long int.
2311 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2313 26. Added the never_mail option to autoreply.
2319 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2320 crash if the getsockname() call failed; this can happen if a connection is
2321 closed very soon after it is established. The problem was simply in the
2322 order in which certain operations were done, causing Exim to try to write
2323 to the SMTP stream before it had set up the file descriptor. The bug has
2324 been fixed by making things happen in the correct order.
2330 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2331 before the connection was closed, thus losing the rejection response.
2333 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2334 some early Solaris releases, but causes trouble in current releases where
2335 socklen_t is defined.
2337 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2340 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2343 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2344 that releasing the top of it at the end releases what was used for sub-
2345 expansions (unless the block got too big). However, discard this block if
2346 the first thing is a variable or header, so that we can use its block when
2347 it is dynamic (useful for very large $message_headers, for example).
2349 6. Lookups now cache *every* query, not just the most recent. A new, separate
2350 store pool is used for this. It can be recovered when all lookup caches are
2351 flushed. Lookups now release memory at the end of their result strings.
2352 This has involved some general refactoring of the lookup sources.
2354 7. Some code has been added to the store_xxx() functions to reduce the amount
2355 of flapping under certain conditions.
2357 8. log_incoming_interface used to affect only the <= reception log lines. Now
2358 it causes the local interface and port to be added to several more SMTP log
2359 lines, for example "SMTP connection from", and rejection lines.
2361 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2363 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2365 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2366 could be overwritten at the end of the current message (or the start of a
2367 new message if it was set in a HELO ACL). The value is now preserved for
2368 the duration of the SMTP connection.
2370 12. If a transport had a headers_rewrite setting, and a matching header line
2371 contained an unqualified address, that address was qualified, even if it
2372 did not match any rewriting rules. The underlying bug was that the values
2373 of the flags that permit the existence of unqualified sender and recipient
2374 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2375 non-local messages, and by -bnq for local messages) were not being
2376 preserved with the message after it was received.
2378 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2379 "next input=" as part of the text comprising the host identity instead of
2380 the correct text. The code was using the same buffer for two different
2381 strings. However, depending on which order the printing function evaluated
2382 its arguments, the bug did not always show up. Under Linux, for example, my
2383 test suite worked just fine.
2385 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2386 doesn't work with some older versions of Perl. It has been changed to "my",
2387 which in any case is probably the better facility to use.
2389 15. A really picky compiler found some instances of statements for creating
2390 error messages that either had too many or two few arguments for the format
2393 16. The size of the buffer for calls to the DNS resolver has been increased
2394 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2395 for addresses that have a lot of PTR records. This alleviates a problem; it
2396 does not fully solve it.
2398 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2399 buffer now truncates the list and logs the incident, which is the same
2400 action as happens when Exim is looking up a host name and its aliases.
2401 Previously in this situation something unpredictable would happen;
2402 sometimes it was "internal error: store_reset failed".
2404 18. If a server dropped the connection unexpectedly when an Exim client was
2405 using GnuTLS and trying to read a response, the client delivery process
2406 crashed while trying to generate an error log message.
2408 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2409 string, for example:
2411 warn message = H1: something\nH2: something
2413 the text was added as a single header line from Exim's point of view
2414 though it ended up OK in the delivered message. However, searching for the
2415 second and subsequent header lines using $h_h2: did not work. This has been
2416 fixed. Similarly, if a system filter added multiple headers in this way,
2417 the routers could not see them.
2419 20. Expanded the error message when iplsearch is called with an invalid key to
2420 suggest using net-iplsearch in a host list.
2422 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2423 ACLs are no longer actually imposed (and a message to that effect is
2426 22. If a "gecos" field in a passwd entry contained escaped characters, in
2427 particular, if it contained a \" sequence, Exim got it wrong when building
2428 a From: or a Sender: header from that name. A second bug also caused
2429 incorrect handling when an unquoted " was present following a character
2430 that needed quoting.
2432 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2433 was not being matched caselessly.
2435 24. Arranged for all hyphens in the exim.8 source to be escaped with
2438 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2439 properly. Recipient callout cache records were still being keyed to include
2440 the sender, even when use_sender was set false. This led to far more
2441 callouts that were necessary. The sender is no longer included in the key
2442 when use_sender is false.
2444 26. Added "control = submission" modifier to ACLs.
2446 27. Added the ${base62d: operator to decode base 62 numbers.
2448 28. dnsdb lookups can now access SRV records.
2450 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2451 the configuration file.
2453 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2454 (-v) mode. This makes the output for a verbose queue run more intelligible.
2456 31. Added a use_postmaster feature to recipient callouts.
2458 32. Added the $body_zerocount variable, containing the number of binary zero
2459 bytes in the message body.
2461 33. The time of last modification of the "new" subdirectory is now used as the
2462 "mailbox time last read" when there is a quota error for a maildir
2465 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2467 35. Added +ignore_unknown as a special item in host lists.
2469 36. Code for decoding IPv6 addresses in host lists is now included, even if
2470 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2471 address was recognized as an IP address, but was then not correctly decoded
2472 into binary, causing unexpected and incorrect effects when compared with
2479 1. Very minor rewording of debugging text in manualroute to say "list of
2480 hosts" instead of "hostlist".
2482 2. If verify=header_syntax was set, and a header line with an unqualified
2483 address (no domain) and a large number of spaces between the end of the
2484 name and the colon was received, the reception process suffered a buffer
2485 overflow, and (when I tested it) crashed. This was caused by some obsolete
2486 code that should have been removed. The fix is to remove it!
2488 3. When running in the test harness, delay a bit after writing a bounce
2489 message to get a bit more predictability in the log output.
2491 4. Added a call to search_tidyup() just before forking a reception process. In
2492 theory, someone could use a lookup in the expansion of smtp_accept_max_
2493 per_host which, without the tidyup, could leave open a database connection.
2495 5. Added the variables $recipient_data and $sender_data which get set from a
2496 lookup success in an ACL "recipients" or "senders" condition, or a router
2497 "senders" option, similar to $domain_data and $local_part_data.
2499 6. Moved the writing of debug_print from before to after the "senders" test
2502 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2503 problems for message scanning, either using a data ACL, or using
2504 local_scan() because the Received: header was not generated till after they
2505 were called (in order to set the time as the time of reception completion).
2506 I have revised the way this works. The header is now generated after the
2507 body is received, but before the ACL or local_scan() are called. After they
2508 are run, the timestamp in the header is updated.
2514 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2515 before starting a queue runner without re-exec. This happened only when
2516 deliver_drop_privilege was set or when the Exim user was set to root. The
2517 effect of the bug was that timeouts during subsequent deliveries caused
2518 crashes instead of being properly handled. The handler is now left at its
2519 default (and expected) setting.
2521 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2522 message, again when deliver_drop_privilege is set or Exim is run as root.
2523 The bug described in (1) was not present in this case, but the tidying up
2524 of the other signals was missing. I have made the two cases consistent.
2526 3. The ignore_target_hosts setting on a manualroute router was being ignored
2527 for hosts that were looked up using the /MX notation.
2529 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2532 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2533 operated on the sender address. After changing the $sender_address to <>
2534 for the sender address verify, Exim was re-instated it as the original
2535 (before rewriting) address, but remembering that it had rewritten it, so it
2536 wasn't rewriting it again. This bug also had the effect of breaking the
2537 sender address verification caching when the sender address was rewritten.
2539 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2540 This has been changed so that if the ip literal address matches
2541 ignore_target_hosts, the router declines.
2543 7. Added expansion conditions match_domain, match_address, and match_local_
2544 part (NOT match_host).
2546 8. The placeholder for the Received: header didn't have a length field set.
2548 9. Added code to Exim itself and to exim_lock to test for a specific race
2549 condition that could lead to file corruption when using MBX delivery. The
2550 issue is with the lockfile that is created in /tmp. If this file is removed
2551 after a process has opened it but before that process has acquired a lock,
2552 there is the potential for a second process to recreate the file and also
2553 acquire a lock. This could lead to two Exim processes writing to the file
2554 at the same time. The added code performs the same test as UW imapd; it
2555 checks after acquiring the lock that its file descriptor still refers to
2556 the same named file.
2558 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2559 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2562 11. Added the smtp_active_hostname option. If used, this will typically be made
2563 to depend on the incoming interface address. Because $interface_address is
2564 not set up until the daemon has forked a reception process, error responses
2565 that can happen earlier (such as "too many connections") no longer contain
2568 12. If an expansion in a condition on a "warn" statement fails because a lookup
2569 defers, the "warn" statement is abandoned, and the next ACL statement is
2570 processed. Previously this caused the whole ACL to be aborted.
2572 13. Added the iplsearch lookup type.
2574 14. Added ident_timeout as a log selector.
2576 15. Added tls_certificate_verified as a log selector.
2578 16. Added a global option tls_require_ciphers (compare the smtp transport
2579 option of the same name). This controls incoming TLS connections.
2581 17. I finally figured out how to make tls_require_ciphers do a similar thing
2582 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2583 before starting the TLS session.
2585 18. Tabs are now shown as \t in -bP output.
2587 19. If the log selector return_path_on_delivery was set, Exim crashed when
2588 bouncing a message because it had too many Received: header lines.
2590 20. If two routers both had headers_remove settings, and the first one included
2591 a superfluous trailing colon, the final name in the first list and the
2592 first name in the second list were incorrectly joined into one item (with a
2593 colon in the middle).
2599 1. Added -C and -D options to the exinext utility, mainly to make it easier
2600 to include in the automated testing, but these could be helpful when
2601 multiple configurations are in use.
2603 2. The exinext utility was not formatting the output nicely when there was
2604 an alternate port involved in the retry record key, nor when there was a
2605 message id as well (for retries that were specific to a specific message
2606 and a specific host). It was also confused by IPv6 addresses, because of
2607 the additional colons they contain. I have fixed the IPv4 problem, and
2608 patched it up to do a reasonable job for IPv6.
2610 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2611 delivery, the log line now contains "pipelined" if PIPELINING was used.
2613 4. An SMTP transport process used to panic and die if the bind() call to set
2614 an explicit outgoing interface failed. This has been changed; it is now
2615 treated in the same way as a connect() failure.
2617 5. A reference to $sender_host_name in the part of a conditional expansion
2618 that was being skipped was still causing a DNS lookup. This no longer
2621 6. The def: expansion condition was not recognizing references to header lines
2622 that used bh_ and bheader_.
2624 7. Added the _cache feature to named lists.
2626 8. The code for checking quota_filecount in the appendfile transport was
2627 allowing one more file than it should have been.
2629 9. For compatibility with Sendmail, the command line option
2637 and sets the incoming protocol and host name (for trusted callers). The
2638 host name and its colon can be omitted when only the protocol is to be set.
2639 Note the Exim already has two private options, -pd and -ps, that refer to
2640 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2641 "s", but I don't think that's a major issue.
2643 10. A number of refactoring changes to the code, none of which should affect
2646 (a) The number of logging options was getting close to filling up the
2647 32-bit word that was used as a bit map. I have split them into two classes:
2648 those that are passed in the argument to log_write(), and those that are
2649 only ever tested independently outside of that function. These are now in
2650 separate 32-bit words, so there is plenty of room for expansion again.
2651 There is no change in the user interface or the logging behaviour.
2653 (b) When building, for example, log lines, the code previously used a
2654 macro that called string_cat() twice, in order to add two strings. This is
2655 not really sufficiently general. Furthermore, there was one instance where
2656 it was actually wrong because one of the argument was used twice, and in
2657 one call a function was used. (As it happened, calling the function twice
2658 did not affect the overall behaviour.) The macro has been replaced by a
2659 function that can join an arbitrary number of extra strings onto a growing
2662 (c) The code for expansion conditions now uses a table and a binary chop
2663 instead of a serial search (which was left over from when there were very
2664 few conditions). Also, it now recognizes conditions like "pam" even when
2665 the relevant support is not compiled in: a suitably worded error message is
2666 given if an attempt is made to use such a condition.
2668 11. Added ${time_interval:xxxxx}.
2670 12. A bug was causing one of the ddress fields not to be passed back correctly
2671 from remote delivery subprocesses. The field in question was not being
2672 subsequently used, so this caused to problems in practice.
2674 13. Added new log selectors queue_time and deliver_time.
2676 14. Might have fixed a bug in maildirsizefile handling that threw up
2677 "unexpected character" debug warnings, and recalculated the data
2678 unnecessarily. In any case, I expanded the warning message to give more
2681 15. Added the message "Restricted characters in address" to the statements in
2682 the default ACL that block characters like @ and % in local parts.
2684 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2685 Three changes have been made:
2687 (a) There was a serious bug; a negative response to MAIL caused the whole
2688 recipient domain to be cached as invalid, thereby blocking all messages
2689 to all local parts at the same domain, from all senders. This bug has
2690 been fixed. The domain is no longer cached after a negative response to
2691 MAIL if the sender used is not empty.
2693 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2696 (c) A new callout option, "use_sender" has been added for people who want
2697 the modified behaviour.
2703 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2706 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2707 indeed breaks things for older releases.
2709 3. Added additional logging to the case where there is a problem reading data
2710 from a filter that is running in a subprocess using a pipe, in order to
2711 try to track down a specific problem.
2713 4. Testing facility fudge: when running in the test harness and attempting
2714 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2715 getting "No route to host". Convert this to a timeout.
2717 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2720 6. Some OS don't have socklen_t but use size_t instead. This affects the
2721 fifth argument of getsockopt() amongst other things. This is now
2722 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2723 can be set for individual OS. I have set it for SunOS5, OSF1, and
2724 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2725 some earlier ones do not.
2727 7. Change 4.30/15 was not doing the test caselessly.
2729 8. The standard form for an IPv6 address literal was being rejected by address
2730 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2731 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2732 this, as well as the form without the "IPv6" on the front (but only when
2733 address literals are enabled, of course).
2735 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2737 10. Exim crashed if a message with an empty sender address specified by -f
2738 encountered a router with an errors_to setting. This could be provoked only
2739 by a command such as
2743 where an empty string was supplied; "<>" did not hit this bug.
2745 11. Installed PCRE release 4.5.
2747 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2748 remained set. It is now erased.
2750 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2751 times from message ids (which are base 36 rather than the normal 62).
2753 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2754 were being counted as actual protocol errors, and logged if the log
2755 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2756 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2757 and DATA following a set of rejected RCPTs do not count as protocol errors.
2758 In other words, Exim assumes they were pipelined, though this may not
2759 actually be the case. Of course, in all cases the client gets an
2760 appropriate error code.
2762 15. If a lookup fails in an ACL condition, a message about the failure may
2763 be available; it is used if testing the ACL cannot continue, because most
2764 such messages specify what the cause of the deferral is. However, some
2765 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2766 that caused an old message to be retained and used if a later statement
2767 caused a defer, replacing the real cause of the deferral.
2769 16. If an IP address had so many PTR records that the DNS lookup buffer
2770 was not large enough to hold them, Exim could crash while trying to process
2771 the truncated data. It now detects and logs this case.
2773 17. Further to 4.21/58, another change has been made: if (and only if) the
2774 first line of a message (the first header line) ends with CRLF, a bare LF
2775 in a subsequent header line has a space inserted after it, so as not to
2776 terminate the header.
2778 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2779 unnecessarily, used atoi() instead of strtol(), and didn't check the
2780 termination when getting file sizes from file names by regex.
2782 19. Completely re-implemented the support for maildirsize files, in the light
2783 of a number of problems with the previous contributed implementation
2784 (4.30/29). In particular:
2786 . If the quota is zero, the maildirsize file is maintained, but no quota is
2789 . If the maildir directory does not exist, it is created before any attempt
2790 to write a maildirsize file.
2792 . The quota value in the file is just a cache; if the quota is changed in
2793 the transport, the new value overrides.
2795 . A regular expression is available for excluding directories from the
2798 20. The autoreply transport checks the characters in options that define the
2799 message's headers; it allows continued headers, but it was checking with
2800 isspace() after an embedded newline instead of explicitly looking for a
2803 21. If all the "regular" hosts to which an address was routed had passed their
2804 expiry times, and had not reached their retry times, the address was
2805 bounced, even if fallback hosts were defined. Now Exim should go on to try
2808 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2809 equivalent code in the SMTP transport. Some hosts send humungous responses
2810 to HELO/EHLO, more than 1024 it seems.
2812 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2813 gives compiler warnings in some environments. I've now done it "properly",
2816 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2817 (because of problems with the built-in one) was declared to return uschar *
2818 instead of char *, causing compiler failure.
2820 25. Fixed a file descriptor leak when processing alias/forward files.
2822 26. Fixed a minor format string issue in dbfn.c.
2824 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2826 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2827 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2830 29. When a sender address is verified, it is cached, to save repeating the test
2831 when there is more than one recipient in a message. However, when the
2832 verification involves a callout, it is possible for different callout
2833 options to be set for different recipients. It is too complicated to keep
2834 track of this in the cache, so now Exim always runs a verification when a
2835 callout is required, relying on the callout cache for the optimization.
2836 The overhead is duplication of the address routing, but this should not be
2839 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2840 to be verified with callout=postmaster, and the main callout worked but the
2841 postmaster check failed, the verification correctly failed. However, if a
2842 subsequent RCPT command asked for sender verification *without* the
2843 postmaster check, incorrect caching caused this verification also to fail,
2846 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2847 it was not caching the DNS options (qualify_single, search_parents) that
2848 were used when the lookup failed. A subsequent lookup with different
2849 options therefore always gave the same answer, though there were cases
2850 where it should not have. (Example: a "domains = !$mx_any" option on a
2851 dnslookup router: the "domains" option is always processed without any
2852 widening, but the router might have qualify_single set.) Now Exim uses the
2853 cached value only when the same options are set.
2855 32. Added John Jetmore's "exipick" utility to the distribution.
2857 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2858 than a timeout (e.g. a certificate is required, and is not provided), an
2859 Exim server now closes the connection immediately. Previously it waited for
2860 the client to close - but if the client is SSL, it seems that they each
2861 wait for each other, leading to a delay before one of them times out.
2863 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2864 maintained 0.8.x compatibility because I don't think many are using it, and
2865 it is clearly obsolete.
2867 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2870 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2871 client certificate was expired. A simple patch fixes this, though I don't
2872 understand the full logic of why the verify callback is called multiple
2875 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2876 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2877 which causes problems with some clients (such as the Certicom SSL Plus
2878 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2879 disables the coutermeasure allowing Eudora to connect."
2881 38. Exim was not checking that a write() to a log file succeeded. This could
2882 lead to Bad Things if a log got too big, in particular if it hit a file
2883 size limit. Exim now panics and dies if it cannot write to a log file, just
2884 as it does if it cannot open a log file.
2886 39. Modified OS/Makefile-Linux so that it now contains
2888 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2890 The two -D definitions ensure that Exim is compiled with large file
2891 support, which makes it possible to handle log files that are bigger than
2894 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2895 instance) a domain was checked against a named list that involved a lookup,
2896 causing $domain_data to be set, then another domain was checked against the
2897 same list, then the first domain was re-checked, the value of $domain_data
2898 after the final check could be wrong. In particular, if the second check
2899 failed, it could be set empty. This bug probably also applied to
2902 41. The strip_trailing_dot option was not being applied to the address given
2903 with the -f command-line option.
2905 42. The code for reading a message's header from the spool was incrementing
2906 $received_count, but never initializing it. This meant that the value was
2907 incorrect (doubled) while delivering a message in the same process in which
2908 it was received. In the most common configuration of Exim, this never
2909 happens - a fresh exec is done - but it can happen when
2910 deliver_drop_privilege is set.
2912 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2913 it now includes up to 150 characters of the unexpected data in the log
2916 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2917 and building data strings. The size of both of these buffers was 10 000
2918 bytes - far larger than anybody would *ever* want, thought I. Needless to
2919 say, somebody hit the limit. I have increased the maximum line length to
2920 20 000 and the maximum data length of concatenated lines to 100 000. I have
2921 also fixed two bugs, because there was no checking on these buffers. Tsk,
2922 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2923 buffer is too small.
2925 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2926 lsearch lookups. Now it does.
2928 46. When parsing a route_list item in a manualroute router, a fixed-length
2929 buffer was used for the list of hosts. I made this 1024 bytes long,
2930 thinking that nobody would ever have a list of hosts that long. Wrong.
2931 Somebody had a whole pile of complicated expansion conditions, and the
2932 string was silently truncated, leading to an expansion error. It turns out
2933 that it is easier to change to an unlimited length (owing to other changes
2934 that have happened since this code was originally written) than to build
2935 structure for giving a limitation error. The length of the item that
2936 expands into the list of hosts is now unlimited.
2938 47. The lsearch lookup could not handle data where the length of text line was
2939 more than 4095 characters. Such lines were truncated, leading to shortened
2940 data being returned. It should now handle lines of any length.
2942 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2943 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2946 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2947 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2948 no longer happens in comment lines. A list of replacements is now placed
2949 at the head of all of the source files, except those whose only change is
2950 to replace PERL_COMMAND in the very first #! line.
2952 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2953 messages on the queue, with a bottom-up merge sort, using code contributed
2954 by Michael Haardt. This should make operations like -bp somewhat faster on
2955 large queues. It won't affect queue runners, except when queue_run_in_order
2958 51. Installed eximstats 1.31 in the distribution.
2960 52. Added support for SRV lookups to the dnslookup router.
2962 53. If an ACL referred to $message_body or $message_body_end, the value was not
2963 reset for any messages that followed in the same SMTP session.
2965 54. The store-handling optimization for building very long strings was not
2966 differentiating between the different store pools. I don't think this
2967 actually made any difference in practice, but I've tidied it.
2969 55. While running the routers to verify a sender address, $sender_address
2970 was still set to the sender address. This is wrong, because when routing to
2971 send a bounce to the sender, it would be empty. Therefore, I have changed
2972 it so that, while verifying a sender address, $sender_address is set to <>.
2973 (There is no change to what happens when verifying a recipient address.)
2975 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2976 target A or AAAA records (if not already returned) without resetting the
2977 qualify_single or search_parents options of the DNS resolver. These are
2978 inappropriate in this case because the targets of MX and SRV records must
2979 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2980 target that, when qualified, matched something in the local domain. These
2981 two options are now turned off when doing these lookups.
2983 57. It seems that at least some releases of Reiserfs (which does not have the
2984 concept of a fixed number of inodes) returns zero and not -1 for the
2985 number of available inodes. This interacted badly with check_spool_inodes,
2986 which assumed that -1 was the "no such thing" setting. What I have done is
2987 to check that the total number of inodes is greater than zero before doing
2988 the test of how many are available.
2990 58. When a "warn" ACL statement has a log_message modifier, the message is
2991 remembered, and not repeated. This is to avoid a lot of repetition when a
2992 message has many recipients that cause the same warning to be written.
2993 Howewer, Exim was preserving the list of already written lines for an
2994 entire SMTP session, which doesn't seem right. The memory is now reset if a
2995 new message is started.
2997 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2998 debugging output unless log_rewrite was also set.
3000 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3001 of (int)(handle) when we know that handle contains (void *)(-1).
3003 61. The Exim daemon panic-logs an error return when it closes the incoming
3004 connection. However "connection reset by peer" seems to be common, and
3005 isn't really an error worthy of noting specially, so that particular error
3008 62. When Exim is trying to find all the local interfaces, it used to panic and
3009 die if the ioctl to get the interface flags failed. However, it seems that
3010 on at least one OS (Solaris 9) it is possible to have an interface that is
3011 included in the list of interfaces, but for which you get a failure error
3012 for this call. This happens when the interface is not "plumbed" into a
3013 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3014 failure of the "get flags" call assumes that the interface is down.
3016 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3017 makes life easier for people who are doing arithmetic on fields extracted
3018 from dates, where you often get leading zeros that should not be
3019 interpreted as octal.
3021 64. Added qualify_domain to the redirect router, to override the global
3024 65. If a pathologically long header line contained very many addresses (the
3025 report of this problem mentioned 10 000) and each of them was rewritten,
3026 Exim could use up a very large amount of memory. (It kept on making new
3027 copies of the header line as it rewrote, and never released the old ones.)
3028 At the expense of a bit more processing, the header rewriting function has
3029 been changed so that it no longer eats memory in this way.
3031 66. The generation of the Received: header has been moved from the time that a
3032 message starts to be received, to the time that it finishes. The timestamp
3033 in the Received: header should now be very close to that of the <= log
3034 line. There are two side-effects of this change:
3036 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3037 logged header lines no longer include the local Received: line, because
3038 it has not yet been created. The same applies to a copy of the message
3039 that is returned to a non-SMTP sender when a message is rejected.
3041 (b) When a filter file is tested using -bf, no additional Received: header
3042 is added to the test message. After some thought, I decided that this
3045 This change does not affect the value of $received_for. It is still set
3046 after address rewriting, but before local_scan() is called.
3048 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3050 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3051 gave an unhelpful panic error message, and a defer error. I have managed to
3052 change this behaviour so that it now rejects any supplied certificate,
3053 which seems right, as the list of acceptable certificates is empty.
3055 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3056 gave an unhelpful defer error. I have not managed to make this reject any
3057 supplied certificates, but the error message it gives is "no certificate
3058 supplied", which is not helpful.
3060 70. exigrep's output now also includes lines that are not associated with any
3061 message, but which match the given pattern. Implemented by a patch from
3062 Martin Sluka, which also tidied up the Perl a bit.
3064 71. Recipient callout verification, like sender verification, was using <> in
3065 the MAIL FROM command. This isn't really the right thing, since the actual
3066 sender may affect whether the remote host accepts the recipient or not. I
3067 have changed it to use the actual sender in the callout; this means that
3068 the cache record is now keyed on a recipient/sender pair, not just the
3069 recipient address. There doesn't seem to be a real danger of callout loops,
3070 since a callout by the remote host to check the sender would use <>.
3071 [SEE ABOVE: changed after hitting problems.]
3073 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3074 temporary errors. However, in the case of such a code being given after
3075 the end of a data transmission (i.e. after ".") Exim was failing to write
3076 a retry record for the message. (Yes, there was some broken host that was
3077 actually sending 8xx at this point.)
3079 73. An unknown lookup type in a host list could cause Exim to panic-die when
3080 the list was checked. (An example that provoked this was putting <; in the
3081 middle of a list instead of at the start.) If this happened during a DATA
3082 ACL check, a -D file could be left lying around. This kind of configuration
3083 error no longer causes Exim to die; instead it causes a defer errror. The
3084 incident is still logged to the main and panic logs.
3086 74. Buglet left over from Exim 3 conversion. The message "too many messages
3087 in one connection" was written to the rejectlog but not the mainlog, except
3088 when address rewriting (yes!) was being logged.
3090 75. Added write_rejectlog option.
3092 76. When a system filter was run not as root (that is, when system_filter_user
3093 was set), the values of the $n variables were not being returned to the
3094 main process; thus, they were not subsequently available in the $sn
3097 77. Added +return_path_on_delivery log selector.
3099 78. A connection timeout was being treated differently from recipients deferred
3100 when testing hosts_max_try with a message that was older than the host's
3101 retry timeout. (The host should not be counted, thus allowing all hosts to
3102 be tried at least once before bouncing.) This may have been the cause of an
3103 occasionally reported bug whereby a message would remain on the queue
3104 longer than the retry timeout, but would be bounced if a delivery was
3105 forced. I say "may" because I never totally pinned down the problem;
3106 setting up timeout/retry tests is difficult. See also the next item.
3108 79. The ultimate address timeout was not being applied to errors that involved
3109 a combination of host plus message (for example, a timeout on a MAIL
3110 command). When an address resolved to a number of possible hosts, and they
3111 were not all tried for each delivery (e.g. because of hosts_max_try), a
3112 message could remain on the queue longer than the retry timeout.
3114 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3117 81. Fixed an obscure SMTP outgoing bug which required at least the following
3118 conditions: (a) there was another message waiting for the same server;
3119 (b) the server returned 5xx to all RCPT commands in the first message so
3120 that the message was not completed; (c) the server dropped the connection
3121 or gave a negative response to the RSET that Exim sends to abort the
3122 transaction. The observed case was a dropped connection after DATA that had
3123 been sent in pipelining mode. That is, the server had advertised PIPELINING
3124 but was not implementing it correctly. The effect of the bug was incorrect
3125 behaviour, such as trying another host, and this could lead to a crash.
3131 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3132 and daemon.c were passed as pointers to ints; they should have been
3133 pointers to socklen_t variables (which are typically unsigned ints).
3135 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3138 3. Fixed a really odd bug that affected only the testing scheme; patching a
3139 certain fixed string in the binary changed the value of another string that
3140 happened to be identical to the end of the original first string.
3142 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3143 name", it returns that address as the IP address. On some operating
3144 systems (e.g. Solaris), it also passes back the IP address string as the
3145 "host name". However, on others (e.g. Linux), it passes back an empty
3146 string. Exim wasn't checking for this, and was changing the host name to an
3147 empty string, assuming it had been canonicized.
3149 5. Although rare, it is permitted to have more than one PTR record for a given
3150 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3151 all the names associated with an address, because they do in Solaris.
3152 However, it seems that they do not in Linux for data that comes from the
3153 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3154 I found this out when I moved to a new Linux workstation and tried to run
3155 the Exim test suite.
3157 To get round this problem I have changed the code so that it now does its
3158 own call to the DNS to look up PTR records when searching for a host name.
3159 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3160 addresses that are only in /etc/hosts are still found.
3162 This behaviour is, however, controlled by an option called host_lookup_
3163 order, which defaults to "bydns:byaddr". If people want to use the other
3164 order, or indeed, just use one or the other means of lookup, they can
3165 specify it in this variable.
3167 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3168 some operating systems, this comes back from gethostbyaddr() as an empty
3169 string, and this is what Exim used to test for. However, it seems that in
3170 other systems, "." is yielded. Exim now tests for this case too.
3172 7. The values of check_spool_space and check_log_space are now held internally
3173 as a number of kilobytes instead of an absolute number of bytes. If a
3174 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3175 kilobyte. This means that much larger values can be stored.
3177 8. Exim monitor: an attempt to get the action menu when not actually pointing
3178 at a message produces an empty menu entitled "No message selected". This
3179 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3180 no entries in it ("Shell widget menu has zero width and/or height"). So I
3181 have added a single, blank menu entry in this case.
3183 9. Added ${quote_local_part.
3185 10. MIME decoding is now applied to the contents of Subject: header lines when
3188 11. Now that a reference to $sender_host_address automatically causes a reverse
3189 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3190 host lookup before query-style lookups in lists that might use this
3191 variable. This has therefore been abolished, and the "net-" prefix is no
3192 longer necessary for query-style lookups.
3194 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3195 have been a typo for LFLAGS, so it has been changed.
3197 13. The install script calls Exim with "-C /dev/null" in order to find the
3198 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3199 to be output. Howeve, since Exim outputs its version number before the
3200 error, it didn't break the script. It just looked ugly. I fixed this by
3201 always allowing "-C /dev/null" if the caller is root.
3203 14. Ignore overlarge ACL variable number when reading spool file - insurance
3204 against a later release with more variables having written the file.
3206 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3207 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3208 this, as well as the form without the "IPv6" on the front.
3210 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3211 OS/Makefile-Darwin file.
3213 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3214 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3216 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3217 variable. However, some OS use ints for inodes. I've added cast to long int
3218 to get rid of the compiler warning.
3220 19. I had forgotten to lock out "/../" in configuration file names when
3221 ALT_CONFIG_PREFIX was set.
3223 20. Routers used for verification do not need to specify transports. However,
3224 if such a router generated a host list, and callout was configured, Exim
3225 crashed, because it could not find a port number from the (non-existent)
3226 transport. It now assumes port 25 in this circumstance.
3228 21. Added the -t option to exigrep.
3230 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3231 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3232 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3233 them is set, LOOKUP_LSEARCH is forced.
3235 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3237 24. Added sender and host information to the "rejected by local_scan()" log
3238 line; previously there was no indication of these.
3240 25. Added .include_if_exists.
3242 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3243 Linux. It turns out that not all file systems support this. Apparently some
3244 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3245 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3246 not supported on the file descriptor, is now ignored when Exim is trying to
3247 sync a directory. This applies only to Linux.
3249 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3251 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3252 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3255 29. Added support for maildirsize files from supplied patch (modified a bit).
3257 30. The use of :fail: followed by an empty string could lead Exim to respond to
3258 sender verification failures with (e.g.):
3260 550 Verification failed for <xxx>
3261 550 Sender verify failed
3263 where the first response line was missing the '-' that indicates it is not
3264 the final line of the response.
3266 31. The loop for finding the name of the user that called Exim had a hardwired
3267 limit of 10; it now uses the value of finduser_retries, which is used for
3268 all other user lookups.
3270 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3273 33. Exim was neglecting to zero errno before one call of strtol() when
3274 expanding a string and expecting an integer value. On some systems this
3275 resulted in spurious "integer overflow" errors. Also, it was casting the
3276 result into an int without checking.
3278 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3279 did not work. The code looks as if it has *never* worked, though it appears
3280 to have been documented since at least releast 1.62. I have made it work.
3282 35. The "timeout_DNS" error in retry rules, also documented since at least
3283 1.62, also never worked. As it isn't clear exactly what this means, and
3284 clearly it isn't a major issue, I have abolished the feature by treating it
3285 as "timeout", and writing a warning to the main and panic logs.
3287 36. The display of retry rules for -brt wasn't always showing the error code
3290 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3291 timeout_connect_A, timeout_connect_MX.
3293 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3294 to the empty sender.
3296 39. The daemon was not analysing the content of -oX till after it had closed
3297 stderr and disconnected from the controlling terminal. This meant that any
3298 syntax errors were only noted on the panic log, and the return code from
3299 the command was 0. By re-arranging the code a little, I've made the
3300 decoding happen first, so such errors now appear on stderr, and the return
3301 code is 1. However, the actual setting up of the sockets still happens in
3302 the disconnected process, so errors there are still only recorded on the
3305 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3306 connections (as happens on some IP stacks) was logged at start up time as
3307 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3308 it from "IPv6 and IPv4", which means that two separate sockets are being
3311 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3312 says whether AF_INET or AF_INET6 was passed as an argument.
3314 42. Exiwhat output was messed up when time zones were included in log
3317 43. Exiwhat now gives more information about the daemon's listening ports,
3318 and whether -tls-on-connect was used.
3320 44. The "port" option of the smtp transport is now expanded.
3322 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3323 silently ignored. Now an error message is written to the main and panic
3326 46. There's a new ACL modifier called "logwrite" which writes to a log file
3327 as soon as it is encountered.
3329 47. Added $local_user_uid and $local_user_gid at routing time.
3331 48. Exim crashed when trying to verify a sender address that was being
3334 49. Exim was recognizing only a space character after ".include". It now also
3335 recognizes a tab character.
3337 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3338 extracting the relevant information from the specification. The man page no
3339 longer contains scrambled data for the -d option, and I've added a section
3340 at the front about calling Exim under different names.
3342 51. Added "extra_headers" argument to the "mail" command in filter files.
3344 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3347 53. Installed eximstats 1.29.
3349 54. Added transport_filter_timeout as a generic transport option.
3351 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3352 Cc: header lines. This was required by RFC 822, but it not required by RFC
3355 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3356 incoming messages that did not have them. Now it does so only if the
3357 message originates locally, that is, if there is no associated remote host
3358 address. When Resent- header lines are present, this applies to the Resent-
3359 lines rather than the non-Resent- lines.
3361 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3362 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3364 58. Messages for configuration errors now include the name of the main
3365 configuration file - useful now that there may be more than one file in a
3366 list (.included file names were always shown).
3368 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3369 for those rare installations that do not start the daemon as root or run it
3370 setuid root. I've cut out the call to initgroups() if the daemon is not
3373 60. The Exim user and group can now be bound into the binary as text strings
3374 that are looked up at the start of Exim's processing.
3376 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3378 62. Added $mailstore_basename variable.
3380 63. Installed patch to sieve.c from Michael Haardt.
3382 64. When Exim failed to open the panic log after failing to open the main log,
3383 the original message it was trying to log was written to stderr and debug
3384 output, but if they were not available (the usual case in production), it
3385 was lost. Now it is written to syslog before the two lines that record the
3386 failures to open the logs.
3388 65. Users' Exim filters run in subprocesses under the user's uid. It is
3389 possible for a "deliver" command or an alias in a "personal" command to
3390 provoke an address rewrite. If logging of address rewriting is configured,
3391 this fails because the process is not running as root or exim. There may be
3392 a better way of dealing with this, but for the moment (because 4.30 needs
3393 to be released), I have disabled address rewrite logging when running a
3394 filter in a non-root, non-exim process.
3400 1. The buildconfig auxiliary program wasn't quoting the value set for
3401 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3402 not defined. This bug was masked in 4.22 by the effect that was fixed in
3405 2. Some messages that were rejected after a message id was allocated were
3406 shown as "incomplete" by exigrep. It no longer does this for messages that
3407 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3409 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3410 have allow_domain_literals set, the ID did not get logged in the <= line.
3411 Domain literals are now always recognized in Message-ID: header lines.
3413 4. The first argument for a ${extract expansion item is the key name or field
3414 number. Leading and trailing spaces in this item were not being ignored,
3415 causing some misleading effects.
3417 5. When deliver_drop_privilege was set, single queue runner processes started
3418 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3419 same command in the process it spins off) were not dropping privilege.
3421 6. When the daemon running as "exim" started a queue runner, it always
3422 re-executed Exim in the spun-off process. This is a waste of effort when
3423 deliver_drop_privilege is set. The new process now just calls the
3424 queue-runner function directly.
3430 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3433 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3434 ignored. Though the use of -oP was forcing the writing of a pid file, it
3435 was always written to the default place.
3437 3. If the message "no IP address found for host xxxx" is generated during
3438 incoming verification, it is now followed by identification of the incoming
3439 connection (so you can more easily find what provoked it).
3441 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3443 5. Added some features to "harden" Exim a bit more against certain attacks:
3445 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3446 be put in Local/Makefile. This is like the never_users runtime option,
3447 but it cannot be overridden. The default setting is "root".
3449 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3450 prefix string with which any file named in a -C command line option
3453 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3454 is retained for -C and -D only if the caller of Exim is root. Without
3455 it, the exim user may also use -C and -D and retain privilege.
3457 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3458 command line option is disabled.
3460 6. Macro names set by the -D option must start with an upper case letter, just
3461 like macro names defined in the configuration file.
3463 7. Added "dereference=" facility to LDAP.
3465 8. Two instances of the typo "uknown" in the source files are fixed.
3467 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3468 the Configure-Makefile script screwed up while processing it.
3470 10. Incorporated PCRE 4.4.
3472 11. The SMTP synchronization check was not operating right at the start of an
3473 SMTP session. For example, it could not catch a HELO sent before the client
3474 waited for the greeting. There is now a check for outstanding input at the
3475 point when the greeting is written. Because of the duplex, asynchronous
3476 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3477 way, but not yet received, when the check is performed.
3479 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3480 on TCP/IP sockets, because this apparently causes some broken clients to
3483 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3484 unchanged) from the Cygwin maintainer.
3486 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3487 was supported instead of testing for mbx. Effectively a typo.
3489 15. The spa authenticator server code was not checking that the input it
3490 received was valid base64.
3492 16. The debug output line for the "set" modifier in ACLs was not showing the
3493 name of the variable that was being set.
3495 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3497 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3498 lookup on its own. Something else was needed to trigger the lookup. For
3499 example, a match in host_lookup or the need for a host name in a host list.
3500 Now, if $sender_host_name is referenced and the host name has not yet been
3501 looked up, a lookup is performed. If the lookup fails, the variable remains
3502 empty, and $host_lookup_failed is set to "1".
3504 19. Added "eqi" as a case-independent comparison operator.
3506 20. The saslauthd authentication condition could segfault if neither service
3507 nor realm was specified.
3509 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3510 error message that was logged was misleading, and incoming SMTP
3511 connections were dropped. The message is now more accurate, and temporary
3512 errors are given to SMTP connections.
3514 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3515 (except RSET) with a 421 error, until QUIT is received. However, it was
3516 failing to send a response to QUIT.
3518 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3519 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3520 ACL is now run after the helo_try_verify_hosts code.
3522 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3523 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3524 case-independent, so other case variants are also recognized). Apparently
3525 some systems use these upper case variants.
3527 25. If more than two messages were waiting for the same host, and a transport
3528 filter was specified for the transport, Exim sent two messages over the
3529 same TCP/IP connection, and then failed with "socket operation on non-
3530 socket" when it tried to send the third.
3532 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3534 27. The extern definition of crypt16() in expand.c was not being excluded when
3535 the OS had its own crypt16() function.
3537 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3538 as a preferred synonym for return_size_limit, both as an option and as an
3541 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3543 30. Changed the default configuration ACL to relax the local part checking rule
3544 for addresses that are not in any local domains. For these addresses,
3545 slashes and pipe symbols are allowed within local parts, but the sequence
3546 /../ is explicitly forbidden.
3548 31. SPA server authentication was not clearing the challenge buffer before
3551 32. log_message in a "warn" ACL statement was writing to the reject log as
3552 well as to the main log, which contradicts the documentation and doesn't
3553 seem right (because no rejection is happening). So I have stopped it.
3555 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3556 However, I am unable to do any testing of this.
3558 34. Fixed an infelicity in the appendfile transport. When checking directories
3559 for a mailbox, to see if any needed to be created, it was accidentally
3560 using path names with one or more superfluous leading slashes; tracing
3561 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3563 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3564 contents are added to the log line that is written for every discarded
3565 recipient. (Previously a log_message setting was ignored.)
3567 36. The ${quote: operator now quotes the string if it is empty.
3569 37. The install script runs exim in order to find its version number. If for
3570 some reason other than non-existence or emptiness, which it checks, it
3571 could not run './exim', it was installing it with an empty version number,
3572 i.e. as "exim-". This error state is now caught, and the installation is
3575 38. An argument was missing from the function that creates an error message
3576 when Exim fails to connect to the socket for saslauthd authentication.
3577 This could cause Exim to crash, or give a corrupted message.
3579 39. Added isip, isip4, and isip6 to ${if conditions.
3581 40. The ACL variables $acl_xx are now saved with the message, and can be
3582 accessed later in routers, transports, and filters.
3584 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3585 strings in the file are not string-expanded.
3587 42. If a MAIL command specified a SIZE value that was too large to fit into an
3588 int variable, the check against message_size_limit failed. Such values are
3589 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3590 day this will have to be increased, but I don't think I want to be around
3591 when emails are that large.
3598 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3599 iconv() is not standard in FreeBSD.
3601 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3602 IPv6 enabled. The observed symptom was a segmentation fault on return from
3603 the function os_common_find_running_interfaces() in src/os.c.
3605 3. In the check_special_case() function in daemon.c I had used "errno" as an
3606 argument name, which causes warnings on some systems. This was basically a
3607 typo, since it was named "eno" in the comments!
3609 4. The code that waits for the clock to tick (at a resolution of some fraction
3610 of a second) so as to ensure message-id uniqueness was always waiting for
3611 at least one whole tick, when it could have waited for less. [This is
3612 almost certainly not relevant at current processor speeds, where it is
3613 unlikely to ever wait at all. But we try to future-proof.]
3615 5. The function that sleeps for a time interval that includes fractions of a
3616 second contained a race. It did not block SIGALRM between setting the
3617 timer, and suspending (a couple of lines later). If the interval was short
3618 and the sigsuspend() was delayed until after it had expired, the suspension
3619 never ended. On busy systems this could lead to processes getting stuck for
3622 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3623 process, before it forks any delivery processes. The open lookup caching
3624 mechanism meant that the open file or database connection was passed into
3625 the delivery process. The problem was that delivery processes always tidy
3626 up cached lookup data. This could cause a problem for the next delivery
3627 process started by the queue runner, because the external queue runner
3628 process does not know about the closure. So the next delivery process
3629 still has data in the lookup cache. In the case of a file lookup, there was
3630 no problem because closing a file descriptor in a subprocess doesn't affect
3631 the parent. However, if the lookup was caching a connection to a database,
3632 the connection was closed, and the second delivery process was likely to
3633 see errors such as "PGSQL: query failed: server closed the connection
3634 unexpectedly". The problem has been fixed by closing all cached lookups
3635 in a queue runner before running a delivery process.
3637 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3638 seem to have the "const" qualifier which it has on other OS. I've
3641 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3642 *of the same type* (client or server) with the same public name that an
3643 error should be diagnosed.
3645 9. When Exim looked up a host name for an IP address, but failed to find the
3646 original IP address when looking up the host name (a safety check), it
3647 output the message "<ip address> does not match any IP for NULL", which was
3648 confusing, to say the least. The bug was that the host name should have
3649 appeared instead of "NULL".
3651 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3652 user that is built into the binary, and the -C or -D options is used, root
3653 privilege is dropped before the configuration file is read. In addition,
3654 logging is switched to stderr instead of the normal log files. If the
3655 configuration then re-defines the Exim user, the unprivileged environment
3656 is probably not what is expected, so Exim logs a panic warning message (but
3659 However, if deliver_drop_privilege is set, the unprivileged state may well
3660 be exactly what is intended, so the warning has been cut out in that case,
3661 and Exim is allowed to try to write to its normal log files.
3667 1. smtp_return_error_details was not giving details for temporary sender
3668 or receiver verification errors.
3670 2. Diagnose a configuration error if two authenticators have the same public
3673 3. Exim used not to create the message log file for a message until the first
3674 delivery attempt. This could be confusing when incoming messages were held
3675 for policy or load reasons. The message log file is now created at the time
3676 the message is received, and an initial "Received" line is written to it.
3678 4. The automatically generated man page for command line options had a minor
3679 bug that caused no ill effects; however, a more serious problem was that
3680 the procedure for building the man page automatically didn't always
3681 operate. Consequently, release 4.20 contains an out-of-date version. This
3682 shouldn't happen again.
3684 5. When building Exim with embedded Perl support, the script that builds the
3685 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3686 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3688 6. The freeze_tell option was not being used for messages that were frozen on
3689 arrival, either by an ACL or by local_scan().
3691 7. Added the smtp_incomplete_transaction log selector.
3693 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3694 was accepting AUTH without a new EHLO.
3696 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3697 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3700 10. Logging of TCP/IP connections (when configured) now happens in the main
3701 daemon process instead of the child process, so that the TCP/IP connection
3702 count is more accurate (but it can never be perfect).
3704 11. The use of "drop" in a nested ACL was not being handled correctly in the
3705 outer ACL. Now, if condition failure induced by the nested "drop" causes
3706 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3707 or "require"), the connection is dropped.
3709 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3710 that yield "discard" can now be used with an "accept" or a "discard" verb,
3711 but an error is generated for any others (because I can't see a useful way
3712 to define what should happen).
3714 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3715 are now processed in the same way as lines in the Exim configuration file.
3716 In particular, continuation lines are supported.
3718 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3720 15. Added -ti meaning -t -i.
3722 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3723 domains, and warn by logging if others are found.
3725 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3726 returned from GIFCONF when Exim is trying to find the list of interfaces on
3727 a host. The code in os.c has been modified to copy each ifreq to an aligned
3728 structure in all cases.
3730 Also, in some cases, the returned ifreq's were being copied to a 'struct
3731 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3732 means the last couple of bytes of an IPv6 address could be chopped if the
3733 ifreq contained only a normal sockaddr (14 bytes storage).
3735 18. Named domain lists were not supported in the hosts_treat_as_local option.
3736 An entry such as +xxxx was not recognized, and was treated as a literal
3739 19. Ensure that header lines added by a DATA ACL are included in the reject log
3740 if the ACL subsequently rejects the message.
3742 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3743 MD5 (which is deprecated).
3745 21. When testing a filter file using -bf, Exim was writing a message when it
3746 took the sender from a "From " line in the message, but it was not doing so
3747 when it took $return_path from a Return-Path: header line. It now does.
3749 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3750 with a valid header line field name (a series of printing characters
3751 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3753 23. Changed "disc" in the source to "disk" to conform to the documentation and
3754 the book and for uniformity.
3756 24. Ignore Sendmail's -Ooption=value command line item.
3758 25. When execve() failed while trying to run a command in a pipe transport,
3759 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3760 could be confused with a return value of 69 from the command itself. This
3761 has been changed to 127, the value the shell returns if it is asked to run
3762 a non-existent command. The wording for the related log line suggests a
3763 non-existent command as the problem.
3765 26. If received_header_text expands to an empty string, do not add a Received:
3766 header line to the message. (Well, it adds a token one on the spool, but
3767 marks it "old" so that it doesn't get used or transmitted.)
3769 27. Installed eximstats 1.28 (addition of -nt option).
3771 28. There was no check for failure on the call to getsockname() in the daemon
3772 code. This can fail if there is a shortage of resources on the system, with
3773 ENOMEM, for example. A temporary error is now given on failure.
3775 29. Contrary to the C standard, it seems that in some environments, the
3776 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3777 program. Exim now does this explicitly; it affects the formatting of
3778 timestamps using strftime().
3780 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3781 complaints. I've now initialized all the variables, to avoid this.
3783 32. Header lines added by a system filter were not being "seen" during
3784 transport-time rewrites.
3786 33. The info_callback() function passed to OpenSSL is set up with type void
3787 (*)(SSL *, int, int), as described somewhere. However, when calling the
3788 function (actually a macro) that sets it up, the type void(*)() is
3789 expected. I've put in a cast to prevent warnings from picky compilers.
3791 34. If a DNS black list lookup found a CNAME record, but there were no A
3792 records associated with the domain it pointed at, Exim crashed.
3794 35. If a DNS black list lookup returned more than one A record, Exim ignored
3795 all but the first. It now scans all returned addresses if a particular IP
3796 value is being sought. In this situation, the contents of the
3797 $dnslist_value variable are a list of all the addresses, separated by a
3800 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3801 to accept a host name and all its aliases if the forward lookup for any of
3802 them yielded the IP address of the incoming connection. Now it accepts only
3803 those names whose forward lookup yields the correct IP address. Any other
3804 names are discarded. This closes a loophole whereby a rogue DNS
3805 administrator could create reverse DNS records to break through a
3806 wildcarded host restriction in an ACL.
3808 37. If a user filter or a system filter that ran in a subprocess used any of
3809 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3810 the wrong values were passed to the pipe command ($thisaddress had the
3811 value of $0, $0 had the value of $1, etc). This bug was introduced by
3812 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3814 38. Improved the line breaking for long SMTP error messages from ACLs.
3815 Previously, if there was no break point between 40 and 75 characters, Exim
3816 left the rest of the message alone. Two changes have been made: (a) I've
3817 reduced the minimum length to 35 characters; (b) if it can't find a break
3818 point between 35 and 75 characters, it looks ahead and uses the first one
3819 that it finds. This may give the occasional overlong line, but at least the
3820 remaining text gets split now.
3822 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3823 file descriptors might be low, and that setting 1000 would always raise it.
3824 It turns out that in some environments, the limit is already over 1000 and
3825 that lowering it causes trouble. So now Exim takes care not to decrease it.
3827 40. When delivering a message, the value of $return_path is set to $sender_
3828 address at the start of routing (routers may change the value). By an
3829 oversight, this default was not being set up when an address was tested by
3830 -bt or -bv, which affected the outcome if any router or filter referred to
3833 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3834 log only when "message" or "log_message" are set. However, if one of the
3835 conditions was an address verification, or a call to a nested ACL, the
3836 messages generated by the underlying test were being passed through. This
3837 no longer happens. The underlying message is available in $acl_verify_
3838 message for both "message" and "log_message" expansions, so it can be
3839 passed through if needed.
3841 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3842 new expansion $bh_ to give the encoded byte string without charset
3843 translation. Translation happens only if iconv() is available; HAVE_ICONV
3844 indicates this at build time. HEADERS_CHARSET gives the charset to
3845 translate to; headers_charset can change it in the configuration, and
3846 "headers charset" can change it in an individual filter file.
3848 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3849 that creates RFC 2047 encoded "words" labels them as that charset instead
3850 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3851 expansion operator; (ii) when Exim creates a From: line for a local
3852 message; (iii) when a header line is rewritten to include a "phrase" part.
3854 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3855 buggy, causing it to skip the first lines of messages whose message ID
3856 ended in 'D'. This would not have bitten before Exim release 4.14, because
3857 message IDs were unlikely to end in 'D' before then. The effect was to have
3858 incorrect size information for certain domains.
3860 45. #include "config.h" was missing at the start of the crypt16.c module. This
3861 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3864 46. If there was a timeout during a "random" callout check, Exim treated it as
3865 a failure of the random address, and carried on sending RSET and the real
3866 address. If the delay was just some slowness somewhere, the response to the
3867 original RCPT would be taken as a response to RSET and so on, causing
3868 mayhem of various kinds.
3870 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3871 when I implemented it. It didn't allow for the fact that some option values
3872 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3873 the right test for positive values.
3875 48. Domain names in DNS records are case-independent. Exim always looks them up
3876 in lower case. Some resolvers return domain names in exactly the case they
3877 appear in the zone file, that is, they may contain uppercase letters. Not
3878 all resolvers do this - some return always lower case. Exim was treating a
3879 change of case by a resolver as a change of domain, similar to a widening
3880 of a domain abbreviation. This triggered its re-routing code and so it was
3881 trying to route what was effectively the same domain again. This normally
3882 caused routing to fail (because the router wouldn't handle the domain
3883 twice). Now Exim checks for this case specially, and just changes the
3884 casing of the domain that it ultimately uses when it transmits the message
3887 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3890 50. If a filter generated a file delivery with a non-absolute name (possible if
3891 no home directory exists for the router), the forbid_file option was not
3894 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3895 the existing equality matching.
3897 52. Exim was using ints instead of ino_t variables in some places where it was
3898 dealing with inode numbers.
3900 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3901 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3902 called TMPDIR, and if it finds it is different, it changes its value.
3904 54. The smtp_printf() function is now made available to local_scan() so
3905 additional output lines can be written before returning. There is also an
3906 smtp_fflush() function to enable the detection of a dropped connection.
3907 The variables smtp_input and smtp_batched_input are exported to
3910 55. Changed the default runtime configuration: the message "Unknown user"
3911 has been removed from the ACL, and instead placed on the localuser router,
3912 using the cannot_route_message feature. This means that any verification
3913 failures that generate their own messages won't get overridden. Similarly,
3914 the "Unrouteable address" message that was in the ACL for unverifiable
3915 relay addresses has also been removed.
3917 56. Added hosts_avoid_esmtp to the smtp transport.
3919 57. The exicyclog script was not checking for the esoteric option
3920 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3921 will work only if exicyclog is run under the appropriate euid.
3923 58. Following a discussion on the list, the rules by which Exim recognises line
3924 endings on incoming messages have been changed. The -dropcr and drop_cr
3925 options are now no-ops, retained only for backwards compatibility. The
3926 following line terminators are recognized: LF CRLF CR. However, special
3927 processing applies to CR:
3929 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3930 nor a local message in the state where . is a terminator.
3932 (ii) If a bare CR is encountered in a header line, an extra space is added
3933 after the line terminator so as not to end the header. The reasoning
3934 behind this is that bare CRs in header lines are most likely either
3935 to be mistakes, or people trying to play silly games.
3937 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3938 was being incorrectly given as 18 bytes larger than it should have been.
3939 This is a VOB (very old bug).
3941 60. This may never have affected anything current, but just in case it has:
3942 When the local host is found other than at the start of a list of hosts,
3943 the local host, those with the same MX, and any that follow, are discarded.
3944 When the list in question was part of a longer list of hosts, the following
3945 hosts (not currently being processed) were also being discarded. This no
3946 longer happens. I'm not sure if this situation could ever has previously
3949 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3952 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3953 header. This is something that is recommended in a new Internet Draft, and
3954 is something that is documented as being done by Sendmail. There are two
3955 possible values. For messages generated by the autoreply transport, Exim
3958 Auto-Submitted: auto-replied
3960 whereas for all other generated messages (e.g. bounces) it adds
3962 Auto-Submitted: auto-generated
3964 63. The "personal" condition in filters now includes a test for the
3965 Auto-Submitted: header. If it contains the string "auto-" the message it
3966 not considered personal.
3968 64. Added rcpt_include_affixes as a generic transport option.
3970 65. Added queue_only_override (default true).
3972 66. Added the syslog_duplication option.
3974 67. If what should have been the first header line of a message consisted of
3975 a space followed by a colon, Exim was mis-interpreting it as a header line.
3976 It isn't of course - it is syntactically invalid and should therefore be
3977 treated as the start of the message body. The misbehaviour could have
3978 caused a number of strange effects, including loss of data in subsequent
3979 header lines, and spool format errors.
3981 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3982 client host had authenticated. This control can now be exercised by an ACL
3983 for more flexibility.
3985 69. By default, callouts do not happen when testing with -bh. There is now a
3986 variant, -bhc, which does actually run the callout code, including
3987 consulting and updating the callout cache.
3989 70. Added support for saslauthd authentication, courtesy of Alexander
3992 71. If statvfs() failed on the spool or log directories while checking their
3993 size for availability, Exim confusingly gave the error "space shortage".
3994 Furthermore, in debugging mode it crashed with a floating point exception.
3995 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3996 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3997 serious problem, Exim now writes to the main and panic logs when this
3998 happens, with details of the failure. It then refuses to accept the
3999 incoming message, giving the message "spool directory problem" or "log
4000 directory problem" with a 421 code for SMTP messages.
4002 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4003 0, 1, and 2 exist, because some OS complain for execs without them (see
4004 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4005 descriptors. However, the code omitted to check that the open succeeded,
4006 causing mysterious errors if for some reason the permissions on /dev/null
4007 got screwed. Now Exim writes a message to the main and panic logs, and
4008 bombs out if it can't open /dev/null.
4010 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4011 interact so that it is all more flexible. It is supposed to remain
4012 backwards compatible. Also added extra_local_interfaces.
4014 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4015 to bomb out with an assertion failure - to the client this appears as a
4016 connection drop. This problem occurs in the part of the code that was taken
4017 from the Samba project. Fortunately, the assertion is in a very simple
4018 function, so I have fixed this by reproducing the function inline in the
4019 one place where it is called, and arranging for authentication to fail
4020 instead of killing the process with assert().
4022 75. The SPA client code was not working when the server requested OEM rather
4023 than Unicode encoding.
4025 76. Added code to make require_files with a specific uid setting more usable in
4026 the case where statting the file as root fails - usually a non-root-mounted
4027 NFS file system. When this happens and the failure is EACCES, Exim now
4028 forks a subprocess and does the per-uid checking as the relevant uid.
4030 77. Added process_log_path.
4032 78. If log_file_path was not explicitly set, a setting of check_log_space or
4033 check_log_inodes was ignored.
4035 79. If a space check for the spool or log partitions fails, the incident is now
4036 logged. Of course, in the latter case the data may get lost...
4038 80. Added the %p formatting code to string_format() so that it can be used to
4039 print addresses in debug_print(). Adjusted all the address printing in the
4040 debugging in store.c to use %p rather than %d.
4042 81. There was a concern that a line of code in smtp_in.c could overflow a
4043 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4044 initially expressed, the concern was not well-founded, because trailing
4045 spaces are removed early. However, if the trailing spaces were followed by
4046 a NULL, they did not get removed, so the overflow was possible. Two fixes
4049 (a) I re-wrote the offending code in a cleaner fashion.
4050 (b) If an incoming SMTP command contains a NULL character, it is rejected
4053 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4054 runs initgroups(), so that if the Exim user is in any additional groups,
4055 they will be used during message reception.
4061 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob