1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.368 2006/07/06 14:19:50 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.
82 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
83 other effects) broke the use of negated acl sub-conditions.
85 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
88 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
89 "Deny" causes Exim to reject the incoming connection with a 554 error.
90 Unfortunately, if there is a major crisis, such as a disk failure,
91 tcp-wrappers gives "deny", whereas what one would like would be some
92 kind of temporary error. A kludge has been added to help with this.
93 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
94 554 error is used if errno is still zero or contains ENOENT (which occurs
95 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
98 PH/03 Add -lutil to the default FreeBSD LIBS setting.
100 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
101 errors. Otherwise a message that provokes a temporary error (when other
102 messages do not) can cause a whole host to time out.
104 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
105 addresses were routed directly to files or pipes from a redirect router.
106 File deliveries just didn't batch; pipe deliveries might have suffered
109 PH/06 A failure to get a lock for a hints database would erroneously always say
110 "Failed to get write lock", even when it was really a read lock.
112 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
113 of 0600. This has been changed to use the value of the lockfile_mode
114 option (which defaults to 0600).
116 PH/08 Applied small patch from the Sieve maintainer.
118 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
119 folder from quota calculations, a direct delivery into this folder messed
120 up the contents of the maildirsize file. This was because the regex was
121 used only to exclude .Trash (or whatever) when the size of the mailbox
122 was calculated. There was no check that a delivery was happening into an
123 excluded directory. This bug has been fixed by ignoring all quota
124 processing for deliveries into excluded directories.
126 PH/10 Added the maildirfolder_create_regex option to appendfile.
132 PH/01 The code for finding all the local interface addresses on a FreeBSD
133 system running IPv6 was broken. This may well have applied to all BSD
134 systems, as well as to others that have similar system calls. The broken
135 code found IPv4 interfaces correctly, but gave incorrect values for the
136 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
137 that it would not match correctly against @[] and not recognize the IPv6
140 PH/02 The ipliteral router was not recognizing addresses of the form user@
141 [ipv6:....] because it didn't know about the "ipv6:" prefix.
143 PH/03 Added disable_ipv6.
145 PH/04 Changed $reply_address to use the raw form of the headers instead of the
146 decoded form, because it is most often used to construct To: headers
147 lines in autoreplies, and the decoded form may well be syntactically
148 invalid. However, $reply_address has leading white space removed, and all
149 newlines turned into spaces so that the autoreply transport does not
152 PH/05 If group was specified without a user on a router, and no group or user
153 was specified on a transport, the group from the router was ignored.
155 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
156 for visible compile-time settings that can be used to change these
157 numbers, for those that want even more. Backwards compatibility with old
158 spool files has been maintained. However, going back to a previous Exim
159 release will lost any variables that are in spool files.
161 PH/07 Two small changes when running in the test harness: increase delay when
162 passing a TCP/IP connection to a new process, in case the original
163 process has to generate a bounce, and remove special handling of
164 127.0.0.2 (sic), which is no longer necessary.
166 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
167 be the same on different OS.
169 PH/09 Moved a debug statement in filter processing to avoid a race problem when
172 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
173 whether --show-vars was specified or not
175 JJ/02 exipick: Added support for new ACL variable spool format introduced
178 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
179 syntactically invalid From: or Reply-to: line, and a filter used this to
180 generate an autoreply, and therefore failed to obtain an address for the
181 autoreply, Exim could try to deliver to a non-existent relative file
182 name, causing unrelated and misleading errors. What now happens is that
183 it logs this as a hard delivery error, but does not attempt to create a
186 PH/11 The exinext utility has a -C option for testing purposes, but although
187 the given file was scanned by exinext itself; it wasn't being passed on
188 when Exim was called.
190 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
191 an end-of-file indication when reading a command response.
193 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
194 compiled. In many other places in Exim, IPv6 addresses are always
195 recognized, so I have changed this. It also means that IPv4 domain
196 literals of the form [IPV4:n.n.n.n] are now always recognized.
198 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
199 used if the router is not running as root, for example, when verifying at
200 ACL time, or when using -bh. The debugging output from this situation was
201 non-existent - all you got was a failure to exec. I have made two
204 (a) Failures to set uid/gid, the current directory, or a process leader
205 in a subprocess such as that created by queryprogram now generate
206 suitable debugging ouput when -d is set.
208 (b) The queryprogram router detects when it is not running as root,
209 outputs suitable debugging information if -d is set, and then runs
210 the subprocess without attempting to change uid/gid.
212 PH/15 Minor change to Makefile for building test_host (undocumented testing
215 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
216 additional section of a DNS packet that returns MX or SRV records.
217 Instead, it always explicitly searches for A/AAAA records. This avoids
218 major problems that occur when a DNS server includes only records of one
219 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
220 fixed another bug: if SRV records were looked up and the corresponding
221 address records were *not* found in the additional section, the port
222 values from the SRV records were lost.
224 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
225 using the correct key (the original address) when searching the retry
226 rules in order to find which one to use for generating the retry hint.
228 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
229 adding the default one. Similarly, if it contains a Reply-To: header, the
230 errors_reply_to option, if set, is not used.
232 PH/19 When calculating a retry time, Exim used to measure the "time since
233 failure" by looking at the "first failed" field in the retry record. Now
234 it does not use this if it is later than than the arrival time of the
235 message. Instead it uses the arrival time. This makes for better
236 behaviour in cases where some deliveries succeed, thus re-setting the
237 "first failed" field. An example is a quota failure for a huge message
238 when small messages continue to be delivered. Without this change, the
239 "time since failure" will always be short, possible causing more frequent
240 delivery attempts for the huge message than are intended.
241 [Note: This change was subsequently modified - see PH/04 for 4.62.]
243 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
244 $1, $2, $3) because the numerical variables can be reset during some
245 expansion items (e.g. "match"), thereby losing the authentication data.
247 PH/21 Make -bV show the size of off_t variables so that the test suite can
248 decide whether to run tests for quotas > 2G.
250 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
251 mailbox_size, and mailbox_filecount in the appendfile transport. If a
252 filecount value is greater than 2G or if a quota value is greater than 2G
253 on a system where the size of off_t is not greater than 4, a panic error
256 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
257 never match. The debug and -bh output now contains an explicit error
258 message indicating a malformed IPv4 address or mask.
260 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
261 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
264 PH/25 Do not write to syslog when running in the test harness. The only
265 occasion when this arises is a failure to open the main or panic logs
266 (for which there is an explicit test).
268 PH/26 Added the /no_tell option to "control=freeze".
270 PH/27 If a host name lookup failed very early in a connection, for example, if
271 the IP address matched host_lookup and the reverse lookup yielded a name
272 that did not have a forward lookup, an error message of the form "no IP
273 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
274 could be logged. Now it outputs the IP address instead of "NULL".
276 PH/28 An enabling patch from MH: add new function child_open_exim2() which
277 allows the sender and the authenticated sender to be set when
278 submitting a message from within Exim. Since child_open_exim() is
279 documented for local_scan(), the new function should be too.
281 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
282 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
283 results in an empty string is now treated as unset.
285 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
287 PH/31 Added sender_verify_fail logging option.
289 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
290 needed by RFC 822 but not by RFC 2822 was commented out. I have now
291 tidied the source and removed it altogether.
293 PH/33 When a queue run was abandoned because the load average was too high, a
294 log line was always written; now it is written only if the queue_run log
295 selector is set. In addition, the log line for abandonment now contains
296 information about the queue run such as the pid. This is always present
297 in "start" and "stop" lines but was omitted from the "abandon" line.
299 PH/34 Omit spaces between a header name and the colon in the error message that
300 is given when verify = headers_syntax fails (if there are lots of them,
301 the message gets confusing).
303 PH/35 Change the default for dns_check_names_pattern to allow slashes within
304 names, as there are now some PTR records that contain slashes. This check
305 is only to protect against broken name servers that fall over on strange
306 characters, so the fact that it applies to all lookups doesn't matter.
308 PH/36 Now that the new test suite is complete, we can remove some of the
309 special code in Exim that was needed for the old test suite. For example,
310 sorting DNS records because real resolvers return them in an arbitrary
311 order. The new test suite's fake resolver always returns records in the
314 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
315 bounces) except when queue_only is set, to avoid logging races between
316 the different processes.
318 PH/38 Panic-die if .include specifies a non-absolute path.
320 PH/39 A tweak to the "H" retry rule from its user.
322 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
323 a label. They prevented compilation on older perls.
325 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
326 a warning to be raised on newish perls.
328 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
329 on queue. Changes to match documented behaviour of showing count of
330 messages matching specified criteria.
332 PH/40 Changed the default ident timeout from 30s to 5s.
334 PH/41 Added support for the use of login_cap features, on those BSD systems
335 that have them, for controlling the resources used by pipe deliveries.
337 PH/42 The content-scanning code uses fopen() to create files in which to put
338 message data. Previously it was not paying any attention to the mode of
339 the files. Exim runs with umask(0) because the rest of the code creates
340 files with open(), and sets the required mode explicitly. Thus, these
341 files were ending up world-writeable. This was not a big issue, because,
342 being within the spool directory, they were not world-accessible. I have
343 created a function called modefopen, which takes an additional mode
344 argument. It sets umask(777), creates the file, chmods it to the required
345 mode, then resets the umask. All the relevant calls to fopen() in the
346 content scanning code have been changed to use this function.
348 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
349 to 24 hours. This avoids potential overflow problems when processing G
350 and H retry rules. I suspect nobody ever tinkers with this value.
352 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
354 PH/45 When the plaintext authenticator is running as a client, the server's
355 challenges are checked to ensure they are valid base64 strings. By
356 default, the authentication attempt is cancelled if an invalid string is
357 received. Setting client_ignore_invalid_base64 true ignores these errors.
358 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
359 they are received. Thus, the responses can be made to depend on the
360 challenges. If an invalid string is ignored, an empty string is placed in
363 PH/46 Messages that are created by the autoreply transport now contains a
364 References: header, in accordance with RFCs 2822 and 3834.
366 PH/47 Added authenticated_sender_force to the smtp transport.
368 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
370 PH/49 Installed latest patch from the Sieve maintainer.
372 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
373 was also set, the appendfile transport was unnecessarily scanning a
374 directory of message files (e.g. for maildir delivery) to find the count
375 of files (along with the size), even though it did not need this
376 information. It now does the scan only if it needs to find either the
377 size of the count of files.
379 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
381 PH/52 Two bugs concerned with error handling when the smtp transport is
384 (i) Exim was not creating retry information for temporary errors given
385 for individual recipients after the DATA command when the smtp transport
386 was used in LMTP mode. This meant that they could be retried too
387 frequently, and not timed out correctly.
389 (ii) Exim was setting the flag that allows error details to be returned
390 for LMTP errors on RCPT commands, but not for LMTP errors for individual
391 recipients that were returned after the DATA command.
393 PH/53 This is related to PH/52, but is more general: for any failing address,
394 when detailed error information was permitted to be returned to the
395 sender, but the error was temporary, then after the final timeout, only
396 "retry timeout exceeded" was returned. Now it returns the full error as
397 well as "retry timeout exceeded".
399 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
400 do this, and (what is worse) MTAs that accept it.
402 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
403 will now be deprecated.
405 PH/56 New os.c-cygwin from the Cygwin maintainer.
407 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
408 formats (previously only available in exim formats via -bpr, -bpru,
409 and -bpra. Now also available in native and exiqgrep formats)
411 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
412 with very large, slow to parse queues
414 JJ/08 exipick: added ! as generic prefix to negate any criteria format
416 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
418 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
419 responses to authentication challenges, though it was showing the
420 challenges; (ii) I've removed the CR characters from the debug output for
423 PH/58 Allow for the insertion of a newline as well as a space when a string
424 is turned into more than one encoded-word during RFC 2047 encoding. The
425 Sieve code now uses this.
427 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
428 data_4xx, lost_connection, tls_required.
430 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
431 message was being sent as an SMTP response.
433 PH/61 Add -l and -k options to exicyclog.
435 PH/62 When verifying, if an address was redirected to one new address, so that
436 verification continued, and the new address failed or deferred after
437 having set something in $address_data, the value of $address_data was not
438 passed back to the ACL. This was different to the case when no
439 redirection occurred. The value is now passed back in both cases.
441 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
442 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
443 use login_cap.h, so on its own it isn't the distinguishing feature. The
444 new name refers directly to the setclassresources() function.
446 PH/65 Added configuration files for NetBSD3.
448 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
450 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
451 is preferred over IPv4.
453 PH/68 The bounce_return_message and bounce_return_body options were not being
454 honoured for bounces generated during the reception of non-SMTP messages.
455 In particular, this applied to messages rejected by the ACL. This bug has
456 been fixed. However, if bounce_return_message is true and bounce_return_
457 body is false, the headers that are returned for a non-SMTP message
458 include only those that have been read before the error was detected.
459 (In the case of an ACL rejection, they have all been read.)
461 PH/69 The HTML version of the specification is now built in a directory called
462 spec_html instead of spec.html, because the latter looks like a path with
463 a MIME-type, and this confuses some software.
465 PH/70 Catch two compiler warnings in sieve.c.
467 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
468 function verify_get_ident() calls ip_connect() to connect a socket, but
469 if the "connect()" function timed out, ip_connect() used to close the
470 socket. However, verify_get_ident() also closes the socket later, and in
471 between Exim writes to the log, which may get opened at this point. When
472 the socket was closed in ip_connect(), the log could get the same file
473 descriptor number as the socket. This naturally causes chaos. The fix is
474 not to close the socket in ip_connect(); the socket should be closed by
475 the function that creates it. There was only one place in the code where
476 this was missing, in the iplookup router, which I don't think anybody now
477 uses, but I've fixed it anyway.
479 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
480 well as to direct DNS lookups. Otherwise the handling of names in host
481 lists is inconsistent and therefore confusing.
487 PH/01 Two changes to the default runtime configuration:
489 (1) Move the checks for relay_from_hosts and authenticated clients from
490 after to before the (commented out) DNS black list checks.
492 (2) Add control=submission to the relay_from_hosts and authenticated
493 clients checks, on the grounds that messages accepted by these
494 statements are most likely to be submissions.
496 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
498 (1) Generate an error if the third argument for the ${prvs expansion is
501 (2) Treat a missing third argument of ${prvscheck as if it were an empty
504 (3) Reset the variables that are obtained from the first argument of
505 ${prvscheck and used in the second argument before leaving the code,
506 because their memory is reclaimed, so using them afterwards may do
509 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
510 one (it's much easier than Tom thought :-).
512 (5) Because of (4), we can now allow for the use of $prvscheck_result
513 inside the third argument.
515 PH/03 For some reason, the default setting of PATH when running a command from
516 a pipe transport was just "/usr/bin". I have changed it to
519 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
520 anything to be listed in the output from -bV.
522 PH/05 When a filter generated an autoreply, the entire To: header line was
523 quoted in the delivery log line, like this:
525 => >A.N.Other <ano@some.domain> <original@ddress> ...
527 This has been changed so that it extracts the operative address. There
528 may be more than one such address. If so, they are comma-separated, like
531 => >ano@some.domain,ona@other.domain <original@ddress> ...
533 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
534 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
535 not being looked up in the rDNS to get a host name, Exim was showing the
536 IP address twice in Received: lines, even though the IP addresses were
537 identical. For example:
539 Received: from [1.2.3.4] (helo=[1.2.3.4])
541 However, if the real host name was known, it was omitting the HELO data
542 if it matched the actual IP address. This has been tidied up so that it
543 doesn't show the same IP address twice.
545 PH/07 When both +timestamp and +memory debugging was on, the value given by
546 $tod_xxx expansions could be wrong, because the tod_stamp() function was
547 called by the debug printing, thereby overwriting the timestamp buffer.
548 Debugging no longer uses the tod_stamp() function when +timestamp is set.
550 PH/08 When the original message was included in an autoreply transport, it
551 always said "this is a copy of the message, including all the headers",
552 even if body_only or headers_only was set. It now gives an appropriate
555 PH/09 Applied a patch from the Sieve maintainer which:
557 o fixes some comments
558 o adds the (disabled) notify extension core
559 o adds some debug output for the result of if/elsif tests
560 o points to the current vacation draft in the documentation
561 and documents the missing references header update
565 o fixes a bug in processing the envelope test (when testing
566 multiple envelope elements, the last element determinted the
569 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
570 Electronic Mail") by including:
572 Auto-submitted: auto-generated
574 in the messages that it generates (bounce messages and others, such as
575 warnings). In the case of bounce messages for non-SMTP mesages, there was
576 also a typo: it was using "Auto_submitted" (underscore instead of
577 hyphen). Since every message generated by Exim is necessarily in response
578 to another message, thes have all been changed to:
580 Auto-Submitted: auto-replied
582 in accordance with these statements in the RFC:
584 The auto-replied keyword:
586 - SHOULD be used on messages sent in direct response to another
587 message by an automatic process,
589 - MUST NOT be used on manually-generated messages,
591 - MAY be used on Delivery Status Notifications (DSNs) and Message
592 Disposition Notifications (MDNs),
594 - MUST NOT be used on messages generated by automatic or periodic
595 processes, except for messages which are automatic responses to
598 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
599 to the default Received: header definition.
601 PH/12 Added log selector acl_warn_skipped (default on).
603 PH/13 After a successful wildlsearch lookup, discard the values of numeric
604 variables because (a) they are in the wrong storage pool and (b) even if
605 they were copied, it wouldn't work properly because of the caching.
607 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
608 checking when decoding. Apparently there are clients that generate
609 overlong encoded strings. Why am I not surprised?
611 PH/15 If the first argument of "${if match_address" was not empty, but did not
612 contain an "@" character, Exim crashed. Now it writes a panic log message
613 and treats the condition as false.
615 PH/16 In autoreply, treat an empty string for "once" the same as unset.
617 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
618 extension "envelope-auth". The code is finished and in agreement with
619 other implementations, but there is no documentation so far and in fact,
620 nobody wrote the draft yet. This extension is currently #undef'ed, thus
621 not changing the active code.
623 Print executed "if" and "elsif" statements when debugging is used. This
624 helps a great deal to understand what a filter does.
626 Document more things not specified clearly in RFC3028. I had all this
627 sorted out, when out of a sudden new issues came to my mind. Oops."
629 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
632 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
633 canonical form (as documented). However, after a host name lookup from
634 the IP address, check_host() was doing a simple string comparison with
635 addresses acquired from the DNS when checking that the found name did
636 have the original IP as one of its addresses. Since any found IPv6
637 addresses are likely to be in abbreviated form, the comparison could
638 fail. Luckily, there already exists a function for doing the comparison
639 by converting both addresses to binary, so now that is used instead of
642 PH/20 There was another similar case to PH/19, when a complete host name was
643 given in a host list; looking up its IP address could give an abbreviated
644 form, whereas the current host's name might or might not be abbreviated.
645 The same fix has been applied.
651 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
652 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
655 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
656 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
658 PH/03 Typo: missing ".o" in src/pcre/Makefile.
660 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
661 header line, restrict the check to what is listed in RFCs 2369 and 2929.
662 Also, for "Auto-Submitted", treat anything other than "no" as
663 non-personal, in accordance with RFC 3834. (Previously it treated
664 anything starting "auto-" as non-personal.)
666 TF/01 The control=submission/name=... option had a problem with syntax
667 errors if the name included a slash character. The /name= option
668 now slurps the rest of the string, so it can include any characters
669 but it must come last in the list of options (after /sender_retain
672 PH/05 Some modifications to the interface to the fake nameserver for the new
680 TK/01 Added the "success_on_redirect" address verification option. See
681 NewStuff for rationale and an example.
683 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
685 PH/02 Patch to exigrep to allow it to work on syslog lines.
687 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
688 fread() to skip over the body file's header line, because in Cygwin the
689 header line is locked and is inaccessible.
691 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
692 co-exist for some time) to make it clear that it is the Exim ID that is
693 referenced, not the Message-ID: header line.
695 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
696 string_format() function, because snprintf() does not exist on all
699 PH/06 The use of forbid_filter_existstest now also locks out the use of the
700 ${stat: expansion item.
702 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
703 protocol synchronization error", to keep the pedants happy.
705 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
706 well as for IRIX systems, when gcc is being used. See the host.c source
709 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
711 PH/10 Named domain lists were not working if used in a queue_smtp_domains
714 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
715 transport and to the smtp transport in LMTP mode.
717 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
719 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
720 run a filter in a subprocess. This could lead to confusion in subsequent
721 lookups in the parent process. There should also be a search_tidyup() at
722 the end of the subprocess.
724 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
725 only if the host matched helo_try_verify_hosts, which caused the
726 verification to occur when the EHLO/HELO command was issued. The ACL just
727 tested the remembered result. Now, if a previous verification attempt has
728 not happened, "verify = helo" does it there and then.
730 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
732 TK/03 Fix log output including CR from clamd.
734 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
735 exist provoked a memory error which could cause a segfault.
737 PH/15 Installed PCRE 6.2
739 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
741 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
742 of the problem. Specifically, suggested +O2 rather than +O1 for the
745 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
747 PH/20 If a delivery was routed to a non-standard port by means of an SRV
748 record, the port was not correctly logged when the outgoing_port log
749 selector was set (it logged the transort's default port).
751 PH/21 Added support for host-specific ports to manualroute, queryprogram,
752 fallback_hosts, and "hosts" in the smtp transport.
754 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
755 host errors such as "Connection refused".
757 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
758 authentication with radiusclient 0.4.9:
760 - Error returned from rc_read_config was caught wrongly
761 - Username/password not passed on to radius server due to wrong length.
763 The presumption is that some radiusclient API changes for 4.51/PH/17
764 were not taken care of correctly. The code is still untested by me (my
765 Linux distribution still has 0.3.2 of radiusclient), but it was
766 contributed by a Radius user.
768 PH/24 When doing a callout, the value of $domain wasn't set correctly when
769 expanding the "port" option of the smtp transport.
771 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
772 while reading a MIME header. Thanks to Tom Hughes for a patch.
774 PH/24 Include config.h inside local_scan.h so that configuration settings are
777 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
778 that in an ACL for RCPT (for example), you can examine exactly what was
781 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
782 commands, but it was not correctly comparing the address with the actual
783 client host address. Thus, it would show the EHLO address in Received:
784 header lines when this was not necessary.
786 PH/27 Added the % operator to ${eval:}.
788 PH/28 Exim tries to create and chdir to its spool directory when it starts;
789 it should be ignoring failures (because with -C, for example, it has lost
790 privilege). It wasn't ignoring creation failures other than "already
793 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
796 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
797 because an input line was too long, either on its own, or by virtue of
798 too many continuations, the temporary file was not being removed, and the
799 return code was incorrect.
801 PH/31 Missing "BOOL" in function definition in filtertest.c.
803 PH/32 Applied Sieve patches from the maintainer.
805 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
807 PH/33 Added "verify = not_blind".
809 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
810 Local/Makefile (with some defaults set). These are used in built scripts
811 such as exicyclog, but they have never been used in the exim_install
812 script (though there are many overriding facilities there). I have
813 arranged that the exim_install script now takes note of these two
816 PH/35 Installed configuration files for Dragonfly.
818 PH/36 When a locally submitted message by a trusted user did not contain a
819 From: header, and the sender address was obtained from -f or from an SMTP
820 MAIL command, and the trusted user did not use -F to supply a sender
821 name, $originator_name was incorrectly used when constructing a From:
822 header. Furthermore, $originator_name was used for submission mode
823 messages from external hosts without From: headers in a similar way,
824 which is clearly wrong.
826 PH/37 Added control=suppress_local_fixups.
828 PH/38 When log_selector = +received_sender was set, and the addition of the
829 sender made the log line's construction buffer exactly full, or one byte
830 less than full, an overflow happened when the terminating "\n" was
833 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
834 when the result of a list match is failure because a DNS lookup failed.
836 PH/40 RM_COMMAND is now used in the building process.
838 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
839 the "build-* directories that it finds.
841 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
842 address in a domain literal was a prefix of an interface address.
844 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
845 when verifying a sender address, unless rewrite_headers is false.
847 PH/44 Wrote a long comment about why errors_to addresses are verified as
848 recipients, not senders.
850 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
851 the ratelimit ACL was added.
853 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
855 PH/46 Added extra information about PostgreSQL errors to the error string.
857 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
858 avoiding the need to install special zones in a real server. This is
859 backwards compatible; if it can't find the fake resolver, it drops back.
860 Thus, both old and new test suites can be run.
862 TF/02 Added util/ratelimit.pl
864 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
865 clock is set back in time.
867 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
868 Candler <B.Candler@pobox.com>.
870 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
871 OK for addresses that are the result of redirections.
873 PH/48 A number of further additions for the benefit of the new test suite,
874 including a fake gethostbyname() that interfaces to the fake DNS resolver
877 TF/06 The fix for widen_domains has also been applied to qualify_single and
878 search_parents which are the other dnslookup options that can cause
881 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
884 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
886 TF/07 Exim produced the error message "an SRV record indicated no SMTP
887 service" if it encountered an MX record with an empty target hostname.
888 The message is now "an MX or SRV record indicated no SMTP service".
890 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
891 if the DNS of the sending site is misconfigured. This is quite a
892 common situation. This change restores the behaviour of treating a
893 helo verification defer as a failure.
895 PH/51 If self=fail was set on a router, the bounce message did not include the
896 actual error message.
902 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
904 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
905 command itself ended in error, the underlying message about the transport
906 filter timeout was being overwritten with the pipe command error. Now the
907 underlying error message should be appended to the second error message.
909 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
911 PH/02 Reduce the amount of output that "make" produces by default. Full output
912 can still be requested.
914 PH/03 The warning log line about a condition test deferring for a "warn" verb
915 was being output only once per connection, rather than after each
916 occurrence (because it was using the same function as for successful
917 "warn" verbs). This seems wrong, so I have changed it.
919 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
920 it should not have, which might have caused a crash in the right
921 circumstances, but probably never did.
923 PH/04 Installed a modified version of Tony Finch's patch to make submission
924 mode fix the return path as well as the Sender: header line, and to
925 add a /name= option so that you can make the user's friendly name appear
928 TF/03 Added the control = fakedefer ACL modifier.
930 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
931 Mark Lowes for thorough testing.
933 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
935 TK/03 Merged latest SRS patch from Miles Wilton.
937 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
938 with the definition in sysexits.h (which is #included earlier).
939 Fortunately, Exim does not actually use EX_OK. The code used to try to
940 preserve the sysexits.h value, by assumimg that macro definitions were
941 scanned for macro replacements. I have been disabused of this notion,
942 so now the code just undefines EX_OK before #including unistd.h.
944 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
945 in the smtp transport. When a block could not be written in a single
946 write() function, the timeout was being re-applied to each part-write.
947 This seems wrong - if the receiver was accepting one byte at a time it
948 would take for ever. The timeout is now adjusted when this happens. It
949 doesn't have to be particularly precise.
951 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
952 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
954 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
955 without a domain if the check to <postmaster@domain> fails.
957 SC/01 Eximstats: added -xls and the ability to specify output files
958 (patch written by Frank Heydlauf).
960 SC/02 Eximstats: use FileHandles for outputing results.
962 SC/03 Eximstats: allow any combination of xls, txt, and html output.
964 SC/04 Eximstats: fixed display of large numbers with -nvr option
966 SC/05 Eximstats: fixed merging of reports with empty tables.
968 SC/06 Eximstats: added the -include_original_destination flag
970 SC/07 Eximstats: removed tabs and trailing whitespace.
972 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
974 TK/06 MBOX spool code: Add real "From " MBOX separator line
975 so the .eml file is really in mbox format (even though
976 most programs do not really care). Patch from Alex Miller.
978 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
979 The latter is generated from $received_to and is only set if the
980 message has one envelope recipient. SA can use these headers,
981 obviously out-of-the-box. Patch from Alex Miller.
983 PH/08 The ${def test on a variable was returning false if the variable's
984 value was "0", contrary to what the specification has always said!
985 The result should be true unless the variable is empty.
987 PH/09 The syntax error of a character other than { following "${if
988 def:variable_name" (after optional whitespace) was not being diagnosed.
989 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
990 accidental colon was present, for example, could give incorrect results.
992 PH/10 Tidied the code in a number of places where the st_size field of a stat()
993 result is used (not including appendfile, where other changes are about
996 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
997 This involved changing a lot of size variables from int to off_t. It
998 should work with maildirs and everything.
1000 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1001 spamd dying while we are connected to it.
1003 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1006 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1007 function for BDB 4.3.
1009 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1011 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1012 the sources for PCRE have been re-arranged and more files are now
1015 PH/15 The code I had for printing potentially long long variables in PH/11
1016 above was not the best (it lost precision). The length of off_t variables
1017 is now inspected at build time, and an appropriate printing format (%ld
1018 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1019 to be "long long int" or "long int". This is needed for the internal
1020 formatting function string_vformat().
1022 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1023 the configuration file to be ":syslog", then the script "guesses" where
1024 the logs files are, rather than using the compiled in default. In our
1025 case the guess is not the same as the compiled default, so the script
1026 suddenly stopped working when I started to use syslog. The patch checks
1027 to see if log_file_path is "". If so, it attempts to read it from exim
1028 with no configuration file to get the compiled in version, before it
1029 falls back to the previous guessing code."
1031 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1032 implementing BATV in an Exim configuration. See NewStuff for the gory
1035 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1036 Makefile that are specific to HP-UX.
1038 PH/18 If the "use_postmaster" option was set for a recipient callout together
1039 with the "random" option, the postmaster address was used as the MAIL
1040 FROM address for the random test, but not for the subsequent recipient
1041 test. It is now used for both.
1043 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1044 patch removes a few documentation additions to RFC 3028, because the
1045 latest draft now contains them. It adds the new en;ascii-case comparator
1046 and a new error check for 8bit text in MIME parts. Comparator and
1047 require names are now matched exactly. I enabled the subaddress
1048 extension, but it is not well tested yet (read: it works for me)."
1050 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1051 rework some of the code of TK/09 above to avoid the hardwired use of
1052 "%lld" and "long long". Replaced the call to snprintf() with a call to
1055 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1056 records point to non-existent hosts", "retry timeout exceeded", and
1057 "retry time not reached for any host after a long failure period".
1059 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1060 experimental DomainKeys support:
1062 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1063 (2) On an error such as an illegally used "control", the wrong name for
1064 the control was given.
1066 These problems did NOT occur unless DomainKeys support was compiled.
1068 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1070 PH/24 Added ${if match_ip condition.
1072 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1073 databases so that it will be absolutely obvious if a crash occurs in the
1074 DB library. This is a regular occurrence (often caused by mis-matched
1077 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1078 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1079 picked up on a user's system that detects such things. There doesn't seem
1080 to be a gcc warning option for this - only an attribute that has to be
1081 put on the function's prototype. It seems that in Fedora Core 4 they have
1082 set this on a number of new functions. No doubt there will be more in due
1085 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1086 specify a transport. However, if an address that was verified by such a
1087 router was the subject of a callout, Exim crashed because it tried to
1088 read the rcpt_include_affixes from the non-existent transport. Now it
1089 just assumes that the setting of that option is false. This bug was
1090 introduced by 4.51/PH/31.
1092 PH/28 Changed -d+all to exclude +memory, because that information is very
1093 rarely of interest, but it makes the output a lot bigger. People tend to
1094 do -d+all out of habit.
1096 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1097 code in os-type was giving problems when libc.so lives in lib64, like on
1100 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1101 aren't the modern standard, and it seems that some systems' include files
1102 don't always have them. Exim was already checking for some of the newer
1103 ones like T_AAAA, and defining it itself. I've added checks for all the
1104 record types that Exim uses.
1106 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1107 not automatically generating a new one, as it is supposed to. This
1108 prevented TLS from working. If the file did exist, but contained invalid
1109 data, a new version was generated, as expected. It was only the case of a
1110 non-existent file that was broken.
1112 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1113 with a change in libdomainkeys > 0.64.
1115 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1116 from DNS. If the selector record carries the flag, it now has
1117 precedence over the domain-wide flag.
1119 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1121 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1122 the use of a port name that isn't defined in /etc/services) occurred, the
1123 message was deferred as in a normal delivery, and thus remained on the
1124 spool, instead of being failed because of the mua_wrapper setting. This
1125 is now fixed, and I tidied up some of the mua_wrapper messages at the
1128 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1129 the messages in a hash of arrays rather than using individual hashes.
1130 This is a bit cleaner and results in dramatic memory savings, albeit
1131 at a slight CPU cost.
1133 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1134 as requested by Marc Sherman.
1136 SC/10 Eximstats: added histograms for user specified patterns as requested
1139 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1141 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1142 fopen() in the content-scanning modules that did not already have it.
1148 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1149 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1151 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1153 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1155 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1156 to test Sieve filters that use "vacation".
1158 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1159 that changes the way the GnuTLS parameters are stored in the cache file.
1160 The new format can be generated externally. For backward compatibility,
1161 if the data in the cache doesn't make sense, Exim assumes it has read an
1162 old-format file, and it generates new data and writes a new file. This
1163 means that you can't go back to an older release without removing the
1166 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1167 work if there are any delivery delays because "one_time" forces the
1168 parent to be marked "delivered", so its unseen clone is never tried
1169 again. For this reason, Exim now forbids the simultaneous setting of
1172 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1173 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1174 entry if you want to know the details. The fix, however, neglected to
1175 consider the case when local delivery batching is involved. The test for
1176 "previously delivered" was not happening when checking to see if an
1177 address could be batched with a previous (undelivered) one; under
1178 certain circumstances this could lead to multiple deliveries to the same
1181 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1182 in its include files, and this causes problems building Exim.
1184 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1185 header_syntax) but Exim was just ignoring anything given after a slash.
1186 In particular, this caused confusion with an attempt to use "verify =
1187 reverse_host_lookup/defer_ok". An error is now given when options are
1188 supplied for verify items that do not have them. (Maybe reverse_host_
1189 lookup should have a defer_ok option, but that's a different point.)
1191 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1192 defined by RFC 821) to 2048, because there were problems with some AUTH
1193 commands, and RFC 1869 says the size should be increased for extended
1194 SMTP commands that take arguments.
1196 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1199 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1200 "unknown" error; now it says that the functionality isn't in the binary.
1202 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1203 an address' error message when a string expansion fails (syntax or
1204 whatever). Otherwise the password may appear in the log. Following change
1205 PH/42 below, there is no longer a chance of it appearing in a bounce
1208 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1210 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1211 of its addresses was ever tried. (Bugzilla bug #2.)
1213 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1214 the result incorrectly in the debug output. (It correctly added a newline
1215 to what was transported.)
1217 TF/01 Added $received_time.
1219 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1220 commented out examples of how to interface to a virus scanner and to
1221 SpamAssassin. Also added commented examples of av_scanner and
1222 spamd_address settings.
1224 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1225 and controls are allowed in which ACLs. There were a couple of minor
1226 errors. Some of the entries in the conditions table (which is a table of
1227 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1228 negation of where the condition IS allowed.
1230 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1232 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1233 header file does not have a version number, so I've had to invent a new
1234 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1235 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1236 radiusclient), but it was contributed by a Radius user.
1238 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1239 files or directories, for OpenSSL.
1241 PH/19 When an Exim process that is running as root has to create an Exim log
1242 file, it does so in a subprocess that runs as exim:exim so as to get the
1243 ownership right at creation (otherwise, other Exim processes might see
1244 the file with the wrong ownership). There was no test for failure of this
1245 fork() call, which would lead to the process getting stuck as it waited
1246 for a non-existent subprocess. Forks do occasionally fail when resources
1247 run out. I reviewed all the other calls to fork(); they all seem to check
1250 PH/20 When checking for unexpected SMTP input at connect time (before writing
1251 the banner), Exim was not dealing correctly with a non-positive return
1252 from the read() function. If the client had disconnected by this time,
1253 the result was a log entry for a synchronization error with an empty
1254 string after "input=" when read() returned zero. If read() returned -1
1255 (an event I could not check), uninitialized data bytes were printed.
1256 There were reports of junk text (parts of files, etc) appearing after
1259 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1261 PH/22 Added support for macro redefinition, and (re)definition in between
1262 driver and ACL definitions.
1264 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1265 forgetting to use the resulting value; it was using the unexpanded value.
1267 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1268 hadn't been configured. The fix is from Juergen Kreileder, who
1269 understands it better than I do:
1271 "Here's what I see happening with three configured cyrus_sasl
1272 authenticators configured (plain, login, cram-md5):
1274 On startup auth_cyrus_sasl_init() gets called for each of these.
1275 This means three calls to sasl_listmech() without a specified mech_list.
1276 => SASL tests which mechs of all available mechs actually work
1277 => three warnings about OTP not working
1278 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1280 With the patch, sasl_listmech() also gets called three times. But now
1281 SASL's mech_list option is set to the server_mech specified in the the
1282 authenticator. Or in other words, the answer from sasl_listmech()
1283 gets limited to just the mech you're testing for (which is different
1285 => the return list contains just 'plain' or 'login', 'cram-md5' or
1286 nothing depending on the value of ob->server_mech.
1288 I've just tested the patch: Authentication still works fine,
1289 unavailable mechs specified in the exim configuration are still
1290 caught, and the auth.log warnings about OTP are gone."
1292 PH/25 When debugging is enabled, the contents of the command line are added
1293 to the debugging output, even when log_selector=+arguments is not
1296 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1297 answer is "GNU", and only if the return is "GNU/something" is the answer
1300 PH/27 $acl_verify_message is now set immediately after the failure of a
1301 verification in an ACL, and so is available in subsequent modifiers. In
1302 particular, the message can be preserved by coding like this:
1304 warn !verify = sender
1305 set acl_m0 = $acl_verify_message
1307 Previously, $acl_verify_message was set only while expanding "message"
1308 and "log_message" when a very denied access.
1310 PH/28 Modified OS/os.c-Linux with
1312 -#ifndef OS_LOAD_AVERAGE
1313 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1315 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1316 nomenclature these days.)
1318 PH/29 Installed patch from the Sieve maintainer that adds the options
1319 sieve_useraddress and sieve_subaddress to the redirect router.
1321 PH/30 In these circumstances:
1322 . Two addresses routed to the same list of hosts;
1323 . First host does not offer TLS;
1324 . First host accepts first address;
1325 . First host gives temporary error to second address;
1326 . Second host offers TLS and a TLS session is established;
1327 . Second host accepts second address.
1328 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1329 and peerdn, if requested) that were in fact used only for the second
1332 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1333 attention to any local part prefix or suffix that was matched by the
1334 router that accepted the address. It now behaves in the same way as it
1335 does for delivery: the affixes are removed from the local part unless
1336 rcpt_include_affixes is set on the transport.
1338 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1339 timeout during the DATA phase of an incoming message.
1341 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1342 applied a patch sanctioned by the Sieve maintainer.
1344 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1345 the uid or gid is negative. A case of a negative gid caused this to be
1346 noticed. The fix allows for either to be negative.
1348 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1349 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1352 PH/36 The change PH/12 above was broken. Fixed it.
1354 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1355 the grounds that routing the same address twice would always produce the
1356 same answer. This might have been true once, but it is certainly no
1357 longer true now. Routing a child address may depend on the previous
1358 routing that produced that child. Some complicated redirection strategies
1359 went wrong when messages had multiple recipients, and made Exim's
1360 behaviour dependent on the order in which the addresses were given.
1362 I have moved the duplicate checking until after the routing is complete.
1363 Exim scans the addresses that are assigned to local and remote
1364 transports, and removes any duplicates. This means that more work will be
1365 done, as duplicates will always all be routed, but duplicates are
1366 presumably rare, so I don't expect this is of any significance.
1368 For deliveries to pipes, files, and autoreplies, the duplicate checking
1369 still happens during the routing process, since they are not going to be
1372 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1373 It corrects a timeout issue with spamd. This is Ian's comment: "The
1374 background is that sometimes spamd either never reads data from a
1375 connection it has accepted, or it never writes response data. The exiscan
1376 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1377 blindly assumes that writes won't block so it may never time out."
1379 PH/39 Allow G after quota size as well as K and M.
1381 PH/40 The value set for $authenticated_id in an authenticator may not contain
1382 binary zeroes or newlines because the value is written to log lines and
1383 to spool files. There was no check on this. Now the value is run through
1384 the string_printing() function so that such characters are converted to
1385 printable escape sequences.
1387 PH/41 $message_linecount is a new variable that contains the total number of
1388 lines in the message. Compare $body_linecount, which is the count for the
1391 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1392 bounce and delay warning messages, except in certain special cases, which
1395 (a) An SMTP error message from a remote host;
1396 (b) A message specified in a :fail: redirection;
1397 (c) A message specified in a "fail" command in a system filter;
1398 (d) A message specified in a FAIL return from the queryprogram router;
1399 (e) A message specified by the cannot_route_message router option.
1401 In these cases only, Exim does include the error details in bounce and
1402 warning messages. There are also a few cases where bland messages such
1403 as "unrouteable address" or "local delivery error" are given.
1405 PH/43 $value is now also set for the "else" part of a ${run expansion.
1407 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1408 being worked on, but at least Exim now implements the latest version to
1411 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1412 process to complete was treated as a delivery failure, a timeout while
1413 writing the message to the pipe was logged, but erroneously treated as a
1414 successful delivery. Such timeouts include transport filter timeouts. For
1415 consistency with the overall process timeout, these timeouts are now
1416 treated as errors, giving rise to delivery failures by default. However,
1417 there is now a new Boolean option for the pipe transport called
1418 timeout_defer, which, if set TRUE, converts the failures into defers for
1419 both kinds of timeout. A transport filter timeout is now identified in
1422 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1423 systems where "make" and "gmake" are different, calling "gmake" at top
1424 level broke things. I've arranged for the value of $(MAKE) to be passed
1425 from the Makefile to this script so that it can call the same version of
1429 A note about Exim versions 4.44 and 4.50
1430 ----------------------------------------
1432 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1433 changes of various kinds. As a consequence, a big documentation update was
1434 needed. This delayed the release for rather longer than seemed good, especially
1435 in the light of a couple of (minor) security issues. Therefore, the changes
1436 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1437 release. So 4.44 and 4.50 are in effect two different branches that both start
1440 I have left the 4.50 change log unchanged; it contains all the changes since
1441 4.43. The change log for 4.44 is below; many of its items are identical to
1442 those for 4.50. This seems to be the most sensible way to preserve the
1443 historical information.
1449 1. Minor wording change to the doc/README.SIEVE file.
1451 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1452 computation of the current number of files was incorrect.
1454 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1455 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1458 4. Give more explanation in the error message when the command for a transport
1459 filter fails to execute.
1461 5. There are several places where Exim runs a non-Exim command in a
1462 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1463 was being done only for the command run by the queryprogram router. It is
1464 now done for all such subprocesses. The other cases are: ${run, transport
1465 filters, and the commands run by the lmtp and pipe transports.
1467 6. Added CONFIGURE_GROUP build-time option.
1469 7. Some older OS have a limit of 256 on the maximum number of file
1470 descriptors. Exim was using setrlimit() to set 1000 as a large value
1471 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1472 systems. I've change it so that if it can't get 1000, it tries for 256.
1474 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1475 was an oversight, and furthermore, ever since the addition of extra
1476 controls (e.g. 4.43/32), the checks on when to allow different forms of
1477 "control" were broken. There should now be diagnostics for all cases when a
1478 control that does not make sense is encountered.
1480 9. Added the /retain_sender option to "control=submission".
1482 10. $recipients is now available in the predata ACL (oversight).
1484 11. Tidy the search cache before the fork to do a delivery from a message
1485 received from the command line. Otherwise the child will trigger a lookup
1486 failure and thereby defer the delivery if it tries to use (for example) a
1487 cached ldap connection that the parent has called unbind on.
1489 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1490 of $address_data from the recipient verification was clobbered by the
1491 sender verification.
1493 13. The value of address_data from a sender verification is now available in
1494 $sender_address_data in subsequent conditions in the ACL statement.
1496 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1498 15. Added a new option "connect=<time>" to callout options, to set a different
1501 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1502 was its contents. (It was OK if the option was not defined at all.)
1504 17. A "Completed" log line is now written for messages that are removed from
1505 the spool by the -Mrm option.
1507 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1508 information about exactly what failed.
1510 19. Added -dd to debug only the daemon process.
1512 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1513 handles timeouts, both on the server side and network timeouts. Renamed the
1514 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1516 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1517 It is now set to "smtps".
1519 22. $host_address is now set to the target address during the checking of
1520 ignore_target_hosts.
1522 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1523 being passed; this would have caused $sender_host_name to have been used if
1524 matching the list had actually called for a host name (not very likely,
1525 since this list is usually IP addresses). A host name is now passed as
1528 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1529 code that specifies a non-restarting handler (typically sigaction() in
1530 modern systems) in an attempt to fix a rare and obscure crash bug.
1532 25. Narrowed the window for a race in the daemon that could cause it to ignore
1533 SIGCHLD signals. This is not a major problem, because they are used only to
1534 wake it up if nothing else does.
1536 26. A malformed maildirsize file could cause Exim to calculate negative values
1537 for the mailbox size or file count. Odd effects could occur as a result.
1538 The maildirsize information is now recalculated if the size or filecount
1541 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1542 support for a long time. Removed HAVE_SYS_VFS_H.
1544 28. Installed the latest version of exipick from John Jetmore.
1546 29. In an address list, if the pattern was not a regular expression, an empty
1547 subject address (from a bounce message) matched only if the pattern was an
1548 empty string. Non-empty patterns were not even tested. This was the wrong
1549 because it is perfectly reasonable to use an empty address as part of a
1550 database query. An empty address is now tested by patterns that are
1551 lookups. However, all the other forms of pattern expect the subject to
1552 contain a local part and a domain, and therefore, for them, an empty
1553 address still always fails if the pattern is not itself empty.
1555 30. Exim went into a mad DNS loop when attempting to do a callout where the
1556 host was specified on an smtp transport, and looking it up yielded more
1557 than one IP address.
1559 31. Re-factored the code for checking spool and log partition space into a
1560 function that finds that data and another that does the check. The former
1561 is then used to implement four new variables: $spool_space, $log_space,
1562 $spool_inodes, and $log_inodes.
1564 32. The RFC2047 encoding function was originally intended for short strings
1565 such as real names; it was not keeping to the 75-character limit for
1566 encoded words that the RFC imposes. It now respects the limit, and
1567 generates multiple encoded words if necessary. To be on the safe side, I
1568 have increased the buffer size for the ${rfc2047: expansion operator from
1571 33. It is now permitted to omit both strings after an "if" condition; if the
1572 condition is true, the result is "true". As before, when the second string
1573 is omitted, a false condition yields an empty string. This makes it less
1574 cumbersome to write custom ACL and router conditions.
1576 34. Failure to deliver a bounce message always caused it to be frozen, even if
1577 there was an errors_to setting on the router. The errors_to setting is now
1580 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1581 canonical form (fully expanded) before being placed in
1582 $sender_host_address.
1584 36. The table in the code that translates DNS record types into text (T_A to
1585 "A" for instance) was missing entries for NS and CNAME. It is just possible
1586 that this could have caused confusion if both these types were looked up
1587 for the same domain, because the text type is used as part of Exim's
1588 per-process caching. But the chance of anyone hitting this buglet seems
1591 37. The dnsdb lookup has been extended in a number of ways.
1593 (1) There is a new type, "zns", which walks up the domain tree until it
1594 finds some nameserver records. It should be used with care.
1596 (2) There is a new type, "mxh", which is like "mx" except that it returns
1597 just the host names, not the priorities.
1599 (3) It is now possible to give a list of domains (or IP addresses) to be
1600 looked up. The behaviour when one of the lookups defers can be
1601 controlled by a keyword.
1603 (4) It is now possible to specify the separator character for use when
1604 multiple records are returned.
1606 38. The dnslists ACL condition has been extended: it is now possible to supply
1607 a list of IP addresses and/or domains to be looked up in a particular DNS
1610 39. Added log_selector=+queue_time_overall.
1612 40. When running the queue in the test harness, wait just a tad after forking a
1613 delivery process, to get repeatability of debugging output.
1615 41. Include certificate and key file names in error message when GnuTLS fails
1616 to set them up, because the GnuTLS error message doesn't include the name
1617 of the failing file when there is a problem reading it.
1619 42. Allow both -bf and -bF in the same test run.
1621 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1623 44. The "Exiscan patch" is now merged into the mainline Exim source.
1625 45. Sometimes the final signoff response after QUIT could fail to get
1626 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1627 < 0 before doing a fflush(). This bug looks as though it goes back to the
1628 introduction of TLS in release 3.20, but "sometimes" must have been rare
1629 because the tests only now provoked it.
1631 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1632 (this can affect the format of dates).
1634 47. exim_tidydb, when checking for the continued existence of a message for
1635 which it has found a message-specific retry record, was not finding
1636 messages that were in split spool directories. Consequently, it was
1637 deleting retry records that should have stayed in existence.
1639 48. Steve fixed some bugs in eximstats.
1641 49. The SPA authentication driver was not abandoning authentication and moving
1642 on to the next authenticator when an expansion was forced to fail,
1643 contradicting the general specification for all authenticators. Instead it
1644 was generating a temporary error. It now behaves as specified.
1646 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1647 (the order specifies the preference for clients). The order is now AES256,
1648 AES128, 3DES, ARCFOUR128.
1650 51. Small patch to Sieve code - explicitly set From: when generating an
1653 52. Exim crashed if a remote delivery caused a very long error message to be
1654 recorded - for instance if somebody sent an entire SpamAssassin report back
1655 as a large number of 550 error lines. This bug was coincidentally fixed by
1656 increasing the size of one of Exim's internal buffers (big_buffer) that
1657 happened as part of the Exiscan merge. However, to be on the safe side, I
1658 have made the code more robust (and fixed the comments that describe what
1661 53. Now that there can be additional text after "Completed" in log lines (if
1662 the queue_time_overall log selector is set), a one-byte patch to exigrep
1663 was needed to allow it to recognize "Completed" as not the last thing in
1666 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1667 patch that reportedly fixes this has been added. I am not expert enough to
1668 create a test for it. This is what the patch creator wrote:
1670 "I found a little strange behaviour of ldap code when working with
1671 Windows 2003 AD Domain, where users was placed in more than one
1672 Organization Units. When I tried to give exim partial DN, the exit code
1673 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1674 But simultaneously result of request was absolutely normal ldap result,
1675 so I produce this patch..."
1677 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1678 REFERENCE, so I have modified the code to exclude the patch when that macro
1681 55. Some experimental protocols are using DNS PTR records for new purposes. The
1682 keys for these records are domain names, not reversed IP addresses. The
1683 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1684 leaves it alone. Component reversal etc. now happens only for IP addresses.
1687 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1689 57. Double the size of the debug message buffer (to 2048) so that more of very
1690 long debug lines gets shown.
1692 58. The exicyclog utility now does better if the number of log files to keep
1693 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1695 59. Two changes related to the smtp_active_hostname option:
1697 (1) $smtp_active_hostname is now available as a variable.
1698 (2) The default for smtp_banner uses $smtp_active_hostname instead
1699 of $primary_hostname.
1701 60. The host_aton() function is supposed to be passed a string that is known
1702 to be a valid IP address. However, in the case of IPv6 addresses, it was
1703 not checking this. This is a hostage to fortune. Exim now panics and dies
1704 if the condition is not met. A case was found where this could be provoked
1705 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1706 components; fortuitously, this particular loophole had already been fixed
1707 by change 4.50/55 above.
1709 If there are any other similar loopholes, the new check in host_aton()
1710 itself should stop them being exploited. The report I received stated that
1711 data on the command line could provoke the exploit when Exim was running as
1712 exim, but did not say which command line option was involved. All I could
1713 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1714 running as the user.
1717 61. There was a buffer overflow vulnerability in the SPA authentication code
1718 (which came originally from the Samba project). I have added a test to the
1719 spa_base64_to_bits() function which I hope fixes it.
1722 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1723 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1725 63. The daemon start-up calls getloadavg() while still root for those OS that
1726 need the first call to be done as root, but it missed one case: when
1727 deliver_queue_load_max is set with deliver_drop_privilege. This is
1728 necessary for the benefit of the queue runner, because there is no re-exec
1729 when deliver_drop_privilege is set.
1731 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1732 This has been fixed.
1734 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1735 was in use, was not putting the data itself into the right store pool;
1736 consequently, it could be overwritten for a subsequent message in the same
1737 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1740 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1742 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1743 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1744 the function were treating the return as a boolean value, which happened to
1745 work because 0=false and not-0=true, but is not correct code.
1747 68. The host_aton() function was not handling scoped IPv6 addresses (those
1748 with, for example, "%eth0" on the end) correctly.
1750 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1751 negated items (that is, ~something) in unsigned ints. Some compilers
1752 apparently mutter when there is no cast.
1754 70. If an address verification called from an ACL failed, and did not produce a
1755 user-specific message (i.e. there was only a "system" message), nothing was
1756 put in $acl_verify_message. In this situation, it now puts the system
1759 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1760 session; change 4.31/43 added the unwanted input to the log line - except
1761 that it did not do this in the start of session case. It now does.
1763 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1764 This is wrong and can cause the other end to generate a synchronization
1765 error if it is another Exim or anything else that does the synchronization
1766 check. A QUIT command is no longer sent after a timeout.
1768 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1769 during host lookups.
1771 74. The defer_ok option of callout verification was not working if it was used
1772 when verifying addresses in header lines, that is, for this case:
1774 verify = header_sender/callout=defer_ok
1776 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1777 those file descriptors could be used for SMTP connections. If anything
1778 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1779 could be sent to the SMTP client, causing chaos. The daemon now opens
1780 stdin, stdout, and stderr to /dev/null when it puts itself into the
1783 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1784 log by default. The user can override this with suitable Perl magic.
1786 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1787 the log message when discard triggers, was not working for the DATA ACL or
1788 for the non-SMTP ACL.
1790 78. Error message wording change in sieve.c.
1792 79. If smtp_accept_max_per_host was set, the number of connections could be
1793 restricted to fewer than expected, because the daemon was trying to set up
1794 a new connection before checking whether the processes handling previous
1795 connections had finished. The check for completed processes is now done
1796 earlier. On busy systems, this bug wouldn't be noticed because something
1797 else would have woken the daemon, and it would have reaped the completed
1800 80. If a message was submitted locally by a user whose login name contained one
1801 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1802 It caused a spool format error. I have fixed the spool reading code. A
1803 related problem was that the "from" clause in the Received: line became
1804 illegal because of the space(s). It is now covered by ${quote_local_part.
1806 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1809 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1811 83. Patch from Sieve maintainer to fix unterminated string problem in
1812 "vacation" handling.
1814 84. Some minor changes to the Linux configuration files to help with other
1815 OS variants using glibc.
1817 85. One more patch for Sieve to update vacation handling to latest spec.
1820 ----------------------------------------------------
1821 See the note above about the 4.44 and 4.50 releases.
1822 ----------------------------------------------------
1828 1. Change 4.43/35 introduced a bug that caused file counts to be
1829 incorrectly computed when quota_filecount was set in an appendfile
1832 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1833 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1836 3. Give more explanation in the error message when the command for a transport
1837 filter fails to execute.
1839 4. There are several places where Exim runs a non-Exim command in a
1840 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1841 was being done only for the command run by the queryprogram router. It is
1842 now done for all such subprocesses. The other cases are: ${run, transport
1843 filters, and the commands run by the lmtp and pipe transports.
1845 5. Some older OS have a limit of 256 on the maximum number of file
1846 descriptors. Exim was using setrlimit() to set 1000 as a large value
1847 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1848 systems. I've change it so that if it can't get 1000, it tries for 256.
1850 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1851 was an oversight, and furthermore, ever since the addition of extra
1852 controls (e.g. 4.43/32), the checks on when to allow different forms of
1853 "control" were broken. There should now be diagnostics for all cases when a
1854 control that does not make sense is encountered.
1856 7. $recipients is now available in the predata ACL (oversight).
1858 8. Tidy the search cache before the fork to do a delivery from a message
1859 received from the command line. Otherwise the child will trigger a lookup
1860 failure and thereby defer the delivery if it tries to use (for example) a
1861 cached ldap connection that the parent has called unbind on.
1863 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1864 of $address_data from the recipient verification was clobbered by the
1865 sender verification.
1867 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1868 was its contents. (It was OK if the option was not defined at all.)
1870 11. A "Completed" log line is now written for messages that are removed from
1871 the spool by the -Mrm option.
1873 12. $host_address is now set to the target address during the checking of
1874 ignore_target_hosts.
1876 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1877 being passed; this would have caused $sender_host_name to have been used if
1878 matching the list had actually called for a host name (not very likely,
1879 since this list is usually IP addresses). A host name is now passed as
1882 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1883 code that specifies a non-restarting handler (typically sigaction() in
1884 modern systems) in an attempt to fix a rare and obscure crash bug.
1886 15. Narrowed the window for a race in the daemon that could cause it to ignore
1887 SIGCHLD signals. This is not a major problem, because they are used only to
1888 wake it up if nothing else does.
1890 16. A malformed maildirsize file could cause Exim to calculate negative values
1891 for the mailbox size or file count. Odd effects could occur as a result.
1892 The maildirsize information is now recalculated if the size or filecount
1895 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1896 support for a long time. Removed HAVE_SYS_VFS_H.
1898 18. Updated exipick to current release from John Jetmore.
1900 19. Allow an empty sender to be matched against a lookup in an address list.
1901 Previously the only cases considered were a regular expression, or an
1904 20. Exim went into a mad DNS lookup loop when doing a callout where the
1905 host was specified on the transport, if the DNS lookup yielded more than
1908 21. The RFC2047 encoding function was originally intended for short strings
1909 such as real names; it was not keeping to the 75-character limit for
1910 encoded words that the RFC imposes. It now respects the limit, and
1911 generates multiple encoded words if necessary. To be on the safe side, I
1912 have increased the buffer size for the ${rfc2047: expansion operator from
1915 22. Failure to deliver a bounce message always caused it to be frozen, even if
1916 there was an errors_to setting on the router. The errors_to setting is now
1919 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1920 canonical form (fully expanded) before being placed in
1921 $sender_host_address.
1923 24. Updated eximstats to version 1.33
1925 25. Include certificate and key file names in error message when GnuTLS fails
1926 to set them up, because the GnuTLS error message doesn't include the name
1927 of the failing file when there is a problem reading it.
1929 26. Expand error message when OpenSSL has problems setting up cert/key files.
1932 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1933 (this can affect the format of dates).
1935 28. exim_tidydb, when checking for the continued existence of a message for
1936 which it has found a message-specific retry record, was not finding
1937 messages that were in split spool directories. Consequently, it was
1938 deleting retry records that should have stayed in existence.
1940 29. eximstats updated to version 1.35
1941 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1942 1.35 - bugfix such that pie charts by volume are generated correctly
1944 30. The SPA authentication driver was not abandoning authentication and moving
1945 on to the next authenticator when an expansion was forced to fail,
1946 contradicting the general specification for all authenticators. Instead it
1947 was generating a temporary error. It now behaves as specified.
1949 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1950 (the order specifies the preference for clients). The order is now AES256,
1951 AES128, 3DES, ARCFOUR128.
1953 31. Small patch to Sieve code - explicitly set From: when generating an
1956 32. Exim crashed if a remote delivery caused a very long error message to be
1957 recorded - for instance if somebody sent an entire SpamAssassin report back
1958 as a large number of 550 error lines. This bug was coincidentally fixed by
1959 increasing the size of one of Exim's internal buffers (big_buffer) that
1960 happened as part of the Exiscan merge. However, to be on the safe side, I
1961 have made the code more robust (and fixed the comments that describe what
1964 33. Some experimental protocols are using DNS PTR records for new purposes. The
1965 keys for these records are domain names, not reversed IP addresses. The
1966 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1967 leaves it alone. Component reversal etc. now happens only for IP addresses.
1970 34. The host_aton() function is supposed to be passed a string that is known
1971 to be a valid IP address. However, in the case of IPv6 addresses, it was
1972 not checking this. This is a hostage to fortune. Exim now panics and dies
1973 if the condition is not met. A case was found where this could be provoked
1974 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1975 components; fortuitously, this particular loophole had already been fixed
1976 by change 4.50/55 or 4.44/33 above.
1978 If there are any other similar loopholes, the new check in host_aton()
1979 itself should stop them being exploited. The report I received stated that
1980 data on the command line could provoke the exploit when Exim was running as
1981 exim, but did not say which command line option was involved. All I could
1982 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1983 running as the user.
1986 35. There was a buffer overflow vulnerability in the SPA authentication code
1987 (which came originally from the Samba project). I have added a test to the
1988 spa_base64_to_bits() function which I hope fixes it.
1991 36. The daemon start-up calls getloadavg() while still root for those OS that
1992 need the first call to be done as root, but it missed one case: when
1993 deliver_queue_load_max is set with deliver_drop_privilege. This is
1994 necessary for the benefit of the queue runner, because there is no re-exec
1995 when deliver_drop_privilege is set.
1997 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1998 was in use, was not putting the data itself into the right store pool;
1999 consequently, it could be overwritten for a subsequent message in the same
2000 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2003 38. Sometimes the final signoff response after QUIT could fail to get
2004 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2005 < 0 before doing a fflush(). This bug looks as though it goes back to the
2006 introduction of TLS in release 3.20, but "sometimes" must have been rare
2007 because the tests only now provoked it.
2013 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2014 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2015 Now it returns an integer. A number of calls were still expecting a T/F
2016 return. Fortuitously, in all cases, the tests worked in OK situations,
2017 which is the norm. However, things would have gone wrong on any write
2018 failures on the smtp file descriptor. This function is used when sending
2019 messages over SMTP and also when doing verify callouts.
2021 2. When Exim is called to do synchronous delivery of a locally submitted
2022 message (the -odf or -odi options), it no longer closes stderr before doing
2025 3. Implemented the mua_wrapper option.
2027 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2029 5. Implemented the functions header_remove(), header_testname(),
2030 header_add_at_position(), and receive_remove_recipient(), and exported them
2033 6. If an ACL "warn" statement specified the addition of headers, Exim already
2034 inserted X-ACL-Warn: at the start if there was no header name. However, it
2035 was not making this test for the second and subsequent header lines if
2036 there were newlines in the string. This meant that an invalid header could
2037 be inserted if Exim was badly configured.
2039 7. Allow an ACL "warn" statement to add header lines at the start or after all
2040 the Received: headers, as well as at the end.
2042 8. Added the rcpt_4xx retry error code.
2044 9. Added postmaster_mailfrom=xxx to callout verification option.
2046 10. Added mailfrom=xxxx to the callout verification option, for verify=
2049 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2050 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2052 12. Inserted some casts to stop certain compilers warning when using pointer
2053 differences as field lengths or precisions in printf-type calls (mostly
2054 affecting debugging statements).
2056 13. Added optional readline() support for -be (dynamically loaded).
2058 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2059 same clock tick as a message's arrival, so that its received time was the
2060 same as the "first fail" time on the retry record, and that message
2061 remained on the queue past the ultimate address timeout, every queue runner
2062 would try a delivery (because it was past the ultimate address timeout) but
2063 after another failure, the ultimate address timeout, which should have then
2064 bounced the address, did not kick in. This was a "< instead of <=" error;
2065 in most cases the first failure would have been in the next clock tick
2066 after the received time, and all would be well.
2068 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2069 being recognized when the domain list was tested by the match_domain
2070 condition in an expansion string.
2072 16. Added the ${str2b64: operator.
2074 17. Exim was always calling setrlimit() to set a large limit for the number of
2075 processes, without checking whether the existing limit was already
2076 adequate. (It did check for the limit on file descriptors.) Furthermore,
2077 errors from getrlimit() and setrlimit() were being ignored. Now they are
2078 logged to the main and panic logs, but Exim does carry on, to try to do its
2079 job under whatever limits there are.
2081 18. Imported PCRE 5.0.
2083 19. Trivial typo in log message " temporarily refused connection" (the leading
2086 20. If the log selector return_path_on_delivery was set and an address was
2087 redirected to /dev/null, the delivery process crashed because it assumed
2088 that a return path would always be set for a "successful" delivery. In this
2089 case, the whole delivery is bypassed as an optimization, and therefore no
2092 21. Internal re-arrangement: the function for sending a challenge and reading
2093 a response while authentication was assuming a zero-terminated challenge
2094 string. It's now changed to take a pointer and a length, to allow for
2095 binary data in such strings.
2097 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2099 23. Exim was not respecting finduser_retries when seeking the login of the
2100 uid under which it was called; it was always trying 10 times. (The default
2101 setting of finduser_retries is zero.) Also, it was sleeping after the final
2102 failure, which is pointless.
2104 24. Implemented tls_on_connect_ports.
2106 25. Implemented acl_smtp_predata.
2108 26. If the domain in control=submission is set empty, Exim assumes that the
2109 authenticated id is a complete email address when it generates From: or
2110 Sender: header lines.
2112 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2113 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2114 chown and chgrp in /bin and hostname in /usr/bin.
2116 28. Exim was keeping the "process log" file open after each use, just as it
2117 does for the main log. This opens the possibility of it remaining open for
2118 long periods when the USR1 signal hits a daemon. Occasional processlog
2119 errors were reported, that could have been caused by this. Anyway, it seems
2120 much more sensible not to leave this file open at all, so that is what now
2123 29. The long-running daemon process does not normally write to the log once it
2124 has entered its main loop, and it closes the log before doing so. This is
2125 so that log files can straightforwardly be renamed and moved. However,
2126 there are a couple of unusual error situations where the daemon does write
2127 log entries, and I had neglected to close the log afterwards.
2129 30. The text of an SMTP error response that was received during a remote
2130 delivery was being truncated at 512 bytes. This is too short for some of
2131 the long messages that one sometimes sees. I've increased the limit to
2134 31. It is now possible to make retry rules that apply only when a message has a
2135 specific sender, in particular, an empty sender.
2137 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2138 it possible to be selective about when SMTP synchronization is enforced.
2140 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2142 32. Implemented hosts_connection_nolog.
2144 33. Added an ACL for QUIT.
2146 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2149 35. Added mailbox_size and mailbox_filecount to appendfile.
2151 36. Added control = no_multiline_responses to ACLs.
2153 37. There was a bug in the logic of the code that waits for the clock to tick
2154 in the case where the clock went backwards by a substantial amount such
2155 that the microsecond fraction of "now" was more than the microsecond
2156 fraction of "then" (but the whole seconds number was less).
2158 38. Added support for the libradius Radius client library this is found on
2159 FreeBSD (previously only the radiusclient library was supported).
2165 1. When certain lookups returned multiple values in the form name=value, the
2166 quoting of the values was not always being done properly. Specifically:
2167 (a) If the value started with a double quote, but contained no whitespace,
2169 (b) If the value contained whitespace other than a space character (i.e.
2170 tabs or newlines or carriage returns) it was not quoted.
2171 This fix has been applied to the mysql and pgsql lookups by writing a
2172 separate quoting function and calling it from the lookup code. The fix
2173 should probably also be applied to nisplus, ibase and oracle lookups, but
2174 since I cannot test any of those, I have not disturbed their existing code.
2176 2. A hit in the callout cache for a specific address caused a log line with no
2177 reason for rejecting RCPT. Now it says "Previous (cached) callout
2178 verification failure".
2180 3. There was an off-by-one bug in the queryprogram router. An over-long
2181 return line was truncated at 256 instead of 255 characters, thereby
2182 overflowing its buffer with the terminating zero. As well as fixing this, I
2183 have increased the buffer size to 1024 (and made a note to document this).
2185 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2186 when Exim is waiting for an SMTP response from a remote server, Exim
2187 restarts its select() call on the socket, thereby resetting its timeout.
2188 This is not a problem when such interrupts are rare. Somebody set up a cron
2189 job to run exiwhat every 2 minutes, which is less than the normal select()
2190 timeout (5 or 10 minutes). This meant that the select() timeout never
2191 kicked in because it was always reset. I have fixed this by comparing the
2192 time when an interrupt arrives with the time at the start of the first call
2193 to select(). If more time than the timeout has elapsed, the interrupt is
2194 treated as a timeout.
2196 5. Some internal re-factoring in preparation for the addition of Sieve
2197 extensions (by MH). In particular, the "personal" test is moved to a
2198 separate function, and given an option for scanning Cc: and Bcc: (which is
2199 not set for Exim filters).
2201 6. When Exim created an email address using the login of the caller as the
2202 local part (e.g. when creating a From: or Sender: header line), it was not
2203 quoting the local part when it contained special characters such as @.
2205 7. Installed new OpenBSD configuration files.
2207 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2208 try to make them clearer.
2210 9. Callout options, other than the timeout value, were being ignored when
2211 verifying sender addresses in header lines. For example, when using
2213 verify = header_sender/callout=no_cache
2215 the cache was (incorrectly) being used.
2217 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2218 only the Cygwin environment.
2220 11. When return_path_on_delivery was set as a log selector, if different remote
2221 addresses in the same message used different return paths and parallel
2222 remote delivery occurred, the wrong values would sometimes be logged.
2223 (Whenever a remote delivery process finished, the return path value from
2224 the most recently started remote delivery process was logged.)
2226 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2227 lines when AUTH and/or TLS are in use. This is the "received protocol"
2228 field. Exim used to use "asmtp" for authenticated SMTP, without any
2229 indication (in the protocol name) for TLS use. Now it follows the RFC and
2230 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2231 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2232 names appear in log lines as well as in Received: header lines.
2234 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2235 extensions, and comparison tests, and to fix some bugs.
2237 14. Changes to the "personal" filter test:
2239 (1) The test was buggy in that it was just doing the equivalent of
2240 "contains" tests on header lines. For example, if a user's address was
2241 anne@some.where, the "personal" test would incorrectly be true for
2243 To: susanne@some.where
2245 This test is now done by extracting each address from the header in turn,
2246 and checking the entire address. Other tests that are part of "personal"
2247 are now done using regular expressions (for example, to check local parts
2248 of addresses in From: header lines).
2250 (2) The list of non-personal local parts in From: addresses has been
2251 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2252 taken from the Sieve specification recommendations.
2254 (3) If the message contains any header line starting with "List-" it is
2255 treated as non-personal.
2257 (4) The test for "circular" in the Subject: header line has been removed
2258 because it now seems ill-conceived.
2260 15. Minor typos in src/EDITME comments corrected.
2262 16. Installed latest exipick from John Jetmore.
2264 17. If headers_add on a router specified a text string that was too long for
2265 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2266 of string_sprintf() is now avoided.
2268 18. $message_body_size was not set (it was always zero) when running the DATA
2269 ACL and the local_scan() function.
2271 19. For the "mail" command in an Exim filter, no default was being set for
2272 the once_repeat time, causing a random time value to be used if "once" was
2273 specified. (If the value happened to be <= 0, no repeat happened.) The
2274 default is now 0s, meaning "never repeat". The "vacation" command was OK
2275 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2276 (I found it when inspecting the code).
2278 20. There is now an overall timeout for performing a callout verification. It
2279 defaults to 4 times the callout timeout, which applies to individual SMTP
2280 commands during the callout. The overall timeout applies when there is more
2281 than one host that can be tried. The timeout is checked before trying the
2282 next host. This prevents very long delays if there are a large number of
2283 hosts and all are timing out (e.g. when the network connections are timing
2284 out). The value of the overall timeout can be changed by specifying an
2285 additional sub-option for "callout", called "maxwait". For example:
2287 verify = sender/callout=5s,maxwait=20s
2289 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2290 to the end before writing, but this should make it even safer).
2292 22. Exim was forgetting that it had advertised PIPELINING for the second and
2293 subsequent messages on an SMTP connection. It was also not resetting its
2294 memory on STARTTLS and an internal HELO.
2296 23. When Exim logs an SMTP synchronization error within a session, it now
2297 records whether PIPELINING has been advertised or not.
2299 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2300 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2301 rather than long int.
2303 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2305 26. Added the never_mail option to autoreply.
2311 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2312 crash if the getsockname() call failed; this can happen if a connection is
2313 closed very soon after it is established. The problem was simply in the
2314 order in which certain operations were done, causing Exim to try to write
2315 to the SMTP stream before it had set up the file descriptor. The bug has
2316 been fixed by making things happen in the correct order.
2322 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2323 before the connection was closed, thus losing the rejection response.
2325 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2326 some early Solaris releases, but causes trouble in current releases where
2327 socklen_t is defined.
2329 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2332 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2335 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2336 that releasing the top of it at the end releases what was used for sub-
2337 expansions (unless the block got too big). However, discard this block if
2338 the first thing is a variable or header, so that we can use its block when
2339 it is dynamic (useful for very large $message_headers, for example).
2341 6. Lookups now cache *every* query, not just the most recent. A new, separate
2342 store pool is used for this. It can be recovered when all lookup caches are
2343 flushed. Lookups now release memory at the end of their result strings.
2344 This has involved some general refactoring of the lookup sources.
2346 7. Some code has been added to the store_xxx() functions to reduce the amount
2347 of flapping under certain conditions.
2349 8. log_incoming_interface used to affect only the <= reception log lines. Now
2350 it causes the local interface and port to be added to several more SMTP log
2351 lines, for example "SMTP connection from", and rejection lines.
2353 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2355 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2357 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2358 could be overwritten at the end of the current message (or the start of a
2359 new message if it was set in a HELO ACL). The value is now preserved for
2360 the duration of the SMTP connection.
2362 12. If a transport had a headers_rewrite setting, and a matching header line
2363 contained an unqualified address, that address was qualified, even if it
2364 did not match any rewriting rules. The underlying bug was that the values
2365 of the flags that permit the existence of unqualified sender and recipient
2366 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2367 non-local messages, and by -bnq for local messages) were not being
2368 preserved with the message after it was received.
2370 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2371 "next input=" as part of the text comprising the host identity instead of
2372 the correct text. The code was using the same buffer for two different
2373 strings. However, depending on which order the printing function evaluated
2374 its arguments, the bug did not always show up. Under Linux, for example, my
2375 test suite worked just fine.
2377 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2378 doesn't work with some older versions of Perl. It has been changed to "my",
2379 which in any case is probably the better facility to use.
2381 15. A really picky compiler found some instances of statements for creating
2382 error messages that either had too many or two few arguments for the format
2385 16. The size of the buffer for calls to the DNS resolver has been increased
2386 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2387 for addresses that have a lot of PTR records. This alleviates a problem; it
2388 does not fully solve it.
2390 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2391 buffer now truncates the list and logs the incident, which is the same
2392 action as happens when Exim is looking up a host name and its aliases.
2393 Previously in this situation something unpredictable would happen;
2394 sometimes it was "internal error: store_reset failed".
2396 18. If a server dropped the connection unexpectedly when an Exim client was
2397 using GnuTLS and trying to read a response, the client delivery process
2398 crashed while trying to generate an error log message.
2400 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2401 string, for example:
2403 warn message = H1: something\nH2: something
2405 the text was added as a single header line from Exim's point of view
2406 though it ended up OK in the delivered message. However, searching for the
2407 second and subsequent header lines using $h_h2: did not work. This has been
2408 fixed. Similarly, if a system filter added multiple headers in this way,
2409 the routers could not see them.
2411 20. Expanded the error message when iplsearch is called with an invalid key to
2412 suggest using net-iplsearch in a host list.
2414 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2415 ACLs are no longer actually imposed (and a message to that effect is
2418 22. If a "gecos" field in a passwd entry contained escaped characters, in
2419 particular, if it contained a \" sequence, Exim got it wrong when building
2420 a From: or a Sender: header from that name. A second bug also caused
2421 incorrect handling when an unquoted " was present following a character
2422 that needed quoting.
2424 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2425 was not being matched caselessly.
2427 24. Arranged for all hyphens in the exim.8 source to be escaped with
2430 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2431 properly. Recipient callout cache records were still being keyed to include
2432 the sender, even when use_sender was set false. This led to far more
2433 callouts that were necessary. The sender is no longer included in the key
2434 when use_sender is false.
2436 26. Added "control = submission" modifier to ACLs.
2438 27. Added the ${base62d: operator to decode base 62 numbers.
2440 28. dnsdb lookups can now access SRV records.
2442 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2443 the configuration file.
2445 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2446 (-v) mode. This makes the output for a verbose queue run more intelligible.
2448 31. Added a use_postmaster feature to recipient callouts.
2450 32. Added the $body_zerocount variable, containing the number of binary zero
2451 bytes in the message body.
2453 33. The time of last modification of the "new" subdirectory is now used as the
2454 "mailbox time last read" when there is a quota error for a maildir
2457 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2459 35. Added +ignore_unknown as a special item in host lists.
2461 36. Code for decoding IPv6 addresses in host lists is now included, even if
2462 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2463 address was recognized as an IP address, but was then not correctly decoded
2464 into binary, causing unexpected and incorrect effects when compared with
2471 1. Very minor rewording of debugging text in manualroute to say "list of
2472 hosts" instead of "hostlist".
2474 2. If verify=header_syntax was set, and a header line with an unqualified
2475 address (no domain) and a large number of spaces between the end of the
2476 name and the colon was received, the reception process suffered a buffer
2477 overflow, and (when I tested it) crashed. This was caused by some obsolete
2478 code that should have been removed. The fix is to remove it!
2480 3. When running in the test harness, delay a bit after writing a bounce
2481 message to get a bit more predictability in the log output.
2483 4. Added a call to search_tidyup() just before forking a reception process. In
2484 theory, someone could use a lookup in the expansion of smtp_accept_max_
2485 per_host which, without the tidyup, could leave open a database connection.
2487 5. Added the variables $recipient_data and $sender_data which get set from a
2488 lookup success in an ACL "recipients" or "senders" condition, or a router
2489 "senders" option, similar to $domain_data and $local_part_data.
2491 6. Moved the writing of debug_print from before to after the "senders" test
2494 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2495 problems for message scanning, either using a data ACL, or using
2496 local_scan() because the Received: header was not generated till after they
2497 were called (in order to set the time as the time of reception completion).
2498 I have revised the way this works. The header is now generated after the
2499 body is received, but before the ACL or local_scan() are called. After they
2500 are run, the timestamp in the header is updated.
2506 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2507 before starting a queue runner without re-exec. This happened only when
2508 deliver_drop_privilege was set or when the Exim user was set to root. The
2509 effect of the bug was that timeouts during subsequent deliveries caused
2510 crashes instead of being properly handled. The handler is now left at its
2511 default (and expected) setting.
2513 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2514 message, again when deliver_drop_privilege is set or Exim is run as root.
2515 The bug described in (1) was not present in this case, but the tidying up
2516 of the other signals was missing. I have made the two cases consistent.
2518 3. The ignore_target_hosts setting on a manualroute router was being ignored
2519 for hosts that were looked up using the /MX notation.
2521 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2524 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2525 operated on the sender address. After changing the $sender_address to <>
2526 for the sender address verify, Exim was re-instated it as the original
2527 (before rewriting) address, but remembering that it had rewritten it, so it
2528 wasn't rewriting it again. This bug also had the effect of breaking the
2529 sender address verification caching when the sender address was rewritten.
2531 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2532 This has been changed so that if the ip literal address matches
2533 ignore_target_hosts, the router declines.
2535 7. Added expansion conditions match_domain, match_address, and match_local_
2536 part (NOT match_host).
2538 8. The placeholder for the Received: header didn't have a length field set.
2540 9. Added code to Exim itself and to exim_lock to test for a specific race
2541 condition that could lead to file corruption when using MBX delivery. The
2542 issue is with the lockfile that is created in /tmp. If this file is removed
2543 after a process has opened it but before that process has acquired a lock,
2544 there is the potential for a second process to recreate the file and also
2545 acquire a lock. This could lead to two Exim processes writing to the file
2546 at the same time. The added code performs the same test as UW imapd; it
2547 checks after acquiring the lock that its file descriptor still refers to
2548 the same named file.
2550 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2551 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2554 11. Added the smtp_active_hostname option. If used, this will typically be made
2555 to depend on the incoming interface address. Because $interface_address is
2556 not set up until the daemon has forked a reception process, error responses
2557 that can happen earlier (such as "too many connections") no longer contain
2560 12. If an expansion in a condition on a "warn" statement fails because a lookup
2561 defers, the "warn" statement is abandoned, and the next ACL statement is
2562 processed. Previously this caused the whole ACL to be aborted.
2564 13. Added the iplsearch lookup type.
2566 14. Added ident_timeout as a log selector.
2568 15. Added tls_certificate_verified as a log selector.
2570 16. Added a global option tls_require_ciphers (compare the smtp transport
2571 option of the same name). This controls incoming TLS connections.
2573 17. I finally figured out how to make tls_require_ciphers do a similar thing
2574 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2575 before starting the TLS session.
2577 18. Tabs are now shown as \t in -bP output.
2579 19. If the log selector return_path_on_delivery was set, Exim crashed when
2580 bouncing a message because it had too many Received: header lines.
2582 20. If two routers both had headers_remove settings, and the first one included
2583 a superfluous trailing colon, the final name in the first list and the
2584 first name in the second list were incorrectly joined into one item (with a
2585 colon in the middle).
2591 1. Added -C and -D options to the exinext utility, mainly to make it easier
2592 to include in the automated testing, but these could be helpful when
2593 multiple configurations are in use.
2595 2. The exinext utility was not formatting the output nicely when there was
2596 an alternate port involved in the retry record key, nor when there was a
2597 message id as well (for retries that were specific to a specific message
2598 and a specific host). It was also confused by IPv6 addresses, because of
2599 the additional colons they contain. I have fixed the IPv4 problem, and
2600 patched it up to do a reasonable job for IPv6.
2602 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2603 delivery, the log line now contains "pipelined" if PIPELINING was used.
2605 4. An SMTP transport process used to panic and die if the bind() call to set
2606 an explicit outgoing interface failed. This has been changed; it is now
2607 treated in the same way as a connect() failure.
2609 5. A reference to $sender_host_name in the part of a conditional expansion
2610 that was being skipped was still causing a DNS lookup. This no longer
2613 6. The def: expansion condition was not recognizing references to header lines
2614 that used bh_ and bheader_.
2616 7. Added the _cache feature to named lists.
2618 8. The code for checking quota_filecount in the appendfile transport was
2619 allowing one more file than it should have been.
2621 9. For compatibility with Sendmail, the command line option
2629 and sets the incoming protocol and host name (for trusted callers). The
2630 host name and its colon can be omitted when only the protocol is to be set.
2631 Note the Exim already has two private options, -pd and -ps, that refer to
2632 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2633 "s", but I don't think that's a major issue.
2635 10. A number of refactoring changes to the code, none of which should affect
2638 (a) The number of logging options was getting close to filling up the
2639 32-bit word that was used as a bit map. I have split them into two classes:
2640 those that are passed in the argument to log_write(), and those that are
2641 only ever tested independently outside of that function. These are now in
2642 separate 32-bit words, so there is plenty of room for expansion again.
2643 There is no change in the user interface or the logging behaviour.
2645 (b) When building, for example, log lines, the code previously used a
2646 macro that called string_cat() twice, in order to add two strings. This is
2647 not really sufficiently general. Furthermore, there was one instance where
2648 it was actually wrong because one of the argument was used twice, and in
2649 one call a function was used. (As it happened, calling the function twice
2650 did not affect the overall behaviour.) The macro has been replaced by a
2651 function that can join an arbitrary number of extra strings onto a growing
2654 (c) The code for expansion conditions now uses a table and a binary chop
2655 instead of a serial search (which was left over from when there were very
2656 few conditions). Also, it now recognizes conditions like "pam" even when
2657 the relevant support is not compiled in: a suitably worded error message is
2658 given if an attempt is made to use such a condition.
2660 11. Added ${time_interval:xxxxx}.
2662 12. A bug was causing one of the ddress fields not to be passed back correctly
2663 from remote delivery subprocesses. The field in question was not being
2664 subsequently used, so this caused to problems in practice.
2666 13. Added new log selectors queue_time and deliver_time.
2668 14. Might have fixed a bug in maildirsizefile handling that threw up
2669 "unexpected character" debug warnings, and recalculated the data
2670 unnecessarily. In any case, I expanded the warning message to give more
2673 15. Added the message "Restricted characters in address" to the statements in
2674 the default ACL that block characters like @ and % in local parts.
2676 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2677 Three changes have been made:
2679 (a) There was a serious bug; a negative response to MAIL caused the whole
2680 recipient domain to be cached as invalid, thereby blocking all messages
2681 to all local parts at the same domain, from all senders. This bug has
2682 been fixed. The domain is no longer cached after a negative response to
2683 MAIL if the sender used is not empty.
2685 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2688 (c) A new callout option, "use_sender" has been added for people who want
2689 the modified behaviour.
2695 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2698 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2699 indeed breaks things for older releases.
2701 3. Added additional logging to the case where there is a problem reading data
2702 from a filter that is running in a subprocess using a pipe, in order to
2703 try to track down a specific problem.
2705 4. Testing facility fudge: when running in the test harness and attempting
2706 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2707 getting "No route to host". Convert this to a timeout.
2709 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2712 6. Some OS don't have socklen_t but use size_t instead. This affects the
2713 fifth argument of getsockopt() amongst other things. This is now
2714 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2715 can be set for individual OS. I have set it for SunOS5, OSF1, and
2716 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2717 some earlier ones do not.
2719 7. Change 4.30/15 was not doing the test caselessly.
2721 8. The standard form for an IPv6 address literal was being rejected by address
2722 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2723 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2724 this, as well as the form without the "IPv6" on the front (but only when
2725 address literals are enabled, of course).
2727 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2729 10. Exim crashed if a message with an empty sender address specified by -f
2730 encountered a router with an errors_to setting. This could be provoked only
2731 by a command such as
2735 where an empty string was supplied; "<>" did not hit this bug.
2737 11. Installed PCRE release 4.5.
2739 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2740 remained set. It is now erased.
2742 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2743 times from message ids (which are base 36 rather than the normal 62).
2745 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2746 were being counted as actual protocol errors, and logged if the log
2747 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2748 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2749 and DATA following a set of rejected RCPTs do not count as protocol errors.
2750 In other words, Exim assumes they were pipelined, though this may not
2751 actually be the case. Of course, in all cases the client gets an
2752 appropriate error code.
2754 15. If a lookup fails in an ACL condition, a message about the failure may
2755 be available; it is used if testing the ACL cannot continue, because most
2756 such messages specify what the cause of the deferral is. However, some
2757 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2758 that caused an old message to be retained and used if a later statement
2759 caused a defer, replacing the real cause of the deferral.
2761 16. If an IP address had so many PTR records that the DNS lookup buffer
2762 was not large enough to hold them, Exim could crash while trying to process
2763 the truncated data. It now detects and logs this case.
2765 17. Further to 4.21/58, another change has been made: if (and only if) the
2766 first line of a message (the first header line) ends with CRLF, a bare LF
2767 in a subsequent header line has a space inserted after it, so as not to
2768 terminate the header.
2770 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2771 unnecessarily, used atoi() instead of strtol(), and didn't check the
2772 termination when getting file sizes from file names by regex.
2774 19. Completely re-implemented the support for maildirsize files, in the light
2775 of a number of problems with the previous contributed implementation
2776 (4.30/29). In particular:
2778 . If the quota is zero, the maildirsize file is maintained, but no quota is
2781 . If the maildir directory does not exist, it is created before any attempt
2782 to write a maildirsize file.
2784 . The quota value in the file is just a cache; if the quota is changed in
2785 the transport, the new value overrides.
2787 . A regular expression is available for excluding directories from the
2790 20. The autoreply transport checks the characters in options that define the
2791 message's headers; it allows continued headers, but it was checking with
2792 isspace() after an embedded newline instead of explicitly looking for a
2795 21. If all the "regular" hosts to which an address was routed had passed their
2796 expiry times, and had not reached their retry times, the address was
2797 bounced, even if fallback hosts were defined. Now Exim should go on to try
2800 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2801 equivalent code in the SMTP transport. Some hosts send humungous responses
2802 to HELO/EHLO, more than 1024 it seems.
2804 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2805 gives compiler warnings in some environments. I've now done it "properly",
2808 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2809 (because of problems with the built-in one) was declared to return uschar *
2810 instead of char *, causing compiler failure.
2812 25. Fixed a file descriptor leak when processing alias/forward files.
2814 26. Fixed a minor format string issue in dbfn.c.
2816 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2818 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2819 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2822 29. When a sender address is verified, it is cached, to save repeating the test
2823 when there is more than one recipient in a message. However, when the
2824 verification involves a callout, it is possible for different callout
2825 options to be set for different recipients. It is too complicated to keep
2826 track of this in the cache, so now Exim always runs a verification when a
2827 callout is required, relying on the callout cache for the optimization.
2828 The overhead is duplication of the address routing, but this should not be
2831 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2832 to be verified with callout=postmaster, and the main callout worked but the
2833 postmaster check failed, the verification correctly failed. However, if a
2834 subsequent RCPT command asked for sender verification *without* the
2835 postmaster check, incorrect caching caused this verification also to fail,
2838 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2839 it was not caching the DNS options (qualify_single, search_parents) that
2840 were used when the lookup failed. A subsequent lookup with different
2841 options therefore always gave the same answer, though there were cases
2842 where it should not have. (Example: a "domains = !$mx_any" option on a
2843 dnslookup router: the "domains" option is always processed without any
2844 widening, but the router might have qualify_single set.) Now Exim uses the
2845 cached value only when the same options are set.
2847 32. Added John Jetmore's "exipick" utility to the distribution.
2849 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2850 than a timeout (e.g. a certificate is required, and is not provided), an
2851 Exim server now closes the connection immediately. Previously it waited for
2852 the client to close - but if the client is SSL, it seems that they each
2853 wait for each other, leading to a delay before one of them times out.
2855 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2856 maintained 0.8.x compatibility because I don't think many are using it, and
2857 it is clearly obsolete.
2859 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2862 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2863 client certificate was expired. A simple patch fixes this, though I don't
2864 understand the full logic of why the verify callback is called multiple
2867 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2868 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2869 which causes problems with some clients (such as the Certicom SSL Plus
2870 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2871 disables the coutermeasure allowing Eudora to connect."
2873 38. Exim was not checking that a write() to a log file succeeded. This could
2874 lead to Bad Things if a log got too big, in particular if it hit a file
2875 size limit. Exim now panics and dies if it cannot write to a log file, just
2876 as it does if it cannot open a log file.
2878 39. Modified OS/Makefile-Linux so that it now contains
2880 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2882 The two -D definitions ensure that Exim is compiled with large file
2883 support, which makes it possible to handle log files that are bigger than
2886 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2887 instance) a domain was checked against a named list that involved a lookup,
2888 causing $domain_data to be set, then another domain was checked against the
2889 same list, then the first domain was re-checked, the value of $domain_data
2890 after the final check could be wrong. In particular, if the second check
2891 failed, it could be set empty. This bug probably also applied to
2894 41. The strip_trailing_dot option was not being applied to the address given
2895 with the -f command-line option.
2897 42. The code for reading a message's header from the spool was incrementing
2898 $received_count, but never initializing it. This meant that the value was
2899 incorrect (doubled) while delivering a message in the same process in which
2900 it was received. In the most common configuration of Exim, this never
2901 happens - a fresh exec is done - but it can happen when
2902 deliver_drop_privilege is set.
2904 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2905 it now includes up to 150 characters of the unexpected data in the log
2908 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2909 and building data strings. The size of both of these buffers was 10 000
2910 bytes - far larger than anybody would *ever* want, thought I. Needless to
2911 say, somebody hit the limit. I have increased the maximum line length to
2912 20 000 and the maximum data length of concatenated lines to 100 000. I have
2913 also fixed two bugs, because there was no checking on these buffers. Tsk,
2914 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2915 buffer is too small.
2917 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2918 lsearch lookups. Now it does.
2920 46. When parsing a route_list item in a manualroute router, a fixed-length
2921 buffer was used for the list of hosts. I made this 1024 bytes long,
2922 thinking that nobody would ever have a list of hosts that long. Wrong.
2923 Somebody had a whole pile of complicated expansion conditions, and the
2924 string was silently truncated, leading to an expansion error. It turns out
2925 that it is easier to change to an unlimited length (owing to other changes
2926 that have happened since this code was originally written) than to build
2927 structure for giving a limitation error. The length of the item that
2928 expands into the list of hosts is now unlimited.
2930 47. The lsearch lookup could not handle data where the length of text line was
2931 more than 4095 characters. Such lines were truncated, leading to shortened
2932 data being returned. It should now handle lines of any length.
2934 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2935 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2938 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2939 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2940 no longer happens in comment lines. A list of replacements is now placed
2941 at the head of all of the source files, except those whose only change is
2942 to replace PERL_COMMAND in the very first #! line.
2944 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2945 messages on the queue, with a bottom-up merge sort, using code contributed
2946 by Michael Haardt. This should make operations like -bp somewhat faster on
2947 large queues. It won't affect queue runners, except when queue_run_in_order
2950 51. Installed eximstats 1.31 in the distribution.
2952 52. Added support for SRV lookups to the dnslookup router.
2954 53. If an ACL referred to $message_body or $message_body_end, the value was not
2955 reset for any messages that followed in the same SMTP session.
2957 54. The store-handling optimization for building very long strings was not
2958 differentiating between the different store pools. I don't think this
2959 actually made any difference in practice, but I've tidied it.
2961 55. While running the routers to verify a sender address, $sender_address
2962 was still set to the sender address. This is wrong, because when routing to
2963 send a bounce to the sender, it would be empty. Therefore, I have changed
2964 it so that, while verifying a sender address, $sender_address is set to <>.
2965 (There is no change to what happens when verifying a recipient address.)
2967 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2968 target A or AAAA records (if not already returned) without resetting the
2969 qualify_single or search_parents options of the DNS resolver. These are
2970 inappropriate in this case because the targets of MX and SRV records must
2971 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2972 target that, when qualified, matched something in the local domain. These
2973 two options are now turned off when doing these lookups.
2975 57. It seems that at least some releases of Reiserfs (which does not have the
2976 concept of a fixed number of inodes) returns zero and not -1 for the
2977 number of available inodes. This interacted badly with check_spool_inodes,
2978 which assumed that -1 was the "no such thing" setting. What I have done is
2979 to check that the total number of inodes is greater than zero before doing
2980 the test of how many are available.
2982 58. When a "warn" ACL statement has a log_message modifier, the message is
2983 remembered, and not repeated. This is to avoid a lot of repetition when a
2984 message has many recipients that cause the same warning to be written.
2985 Howewer, Exim was preserving the list of already written lines for an
2986 entire SMTP session, which doesn't seem right. The memory is now reset if a
2987 new message is started.
2989 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2990 debugging output unless log_rewrite was also set.
2992 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2993 of (int)(handle) when we know that handle contains (void *)(-1).
2995 61. The Exim daemon panic-logs an error return when it closes the incoming
2996 connection. However "connection reset by peer" seems to be common, and
2997 isn't really an error worthy of noting specially, so that particular error
3000 62. When Exim is trying to find all the local interfaces, it used to panic and
3001 die if the ioctl to get the interface flags failed. However, it seems that
3002 on at least one OS (Solaris 9) it is possible to have an interface that is
3003 included in the list of interfaces, but for which you get a failure error
3004 for this call. This happens when the interface is not "plumbed" into a
3005 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3006 failure of the "get flags" call assumes that the interface is down.
3008 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3009 makes life easier for people who are doing arithmetic on fields extracted
3010 from dates, where you often get leading zeros that should not be
3011 interpreted as octal.
3013 64. Added qualify_domain to the redirect router, to override the global
3016 65. If a pathologically long header line contained very many addresses (the
3017 report of this problem mentioned 10 000) and each of them was rewritten,
3018 Exim could use up a very large amount of memory. (It kept on making new
3019 copies of the header line as it rewrote, and never released the old ones.)
3020 At the expense of a bit more processing, the header rewriting function has
3021 been changed so that it no longer eats memory in this way.
3023 66. The generation of the Received: header has been moved from the time that a
3024 message starts to be received, to the time that it finishes. The timestamp
3025 in the Received: header should now be very close to that of the <= log
3026 line. There are two side-effects of this change:
3028 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3029 logged header lines no longer include the local Received: line, because
3030 it has not yet been created. The same applies to a copy of the message
3031 that is returned to a non-SMTP sender when a message is rejected.
3033 (b) When a filter file is tested using -bf, no additional Received: header
3034 is added to the test message. After some thought, I decided that this
3037 This change does not affect the value of $received_for. It is still set
3038 after address rewriting, but before local_scan() is called.
3040 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3042 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3043 gave an unhelpful panic error message, and a defer error. I have managed to
3044 change this behaviour so that it now rejects any supplied certificate,
3045 which seems right, as the list of acceptable certificates is empty.
3047 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3048 gave an unhelpful defer error. I have not managed to make this reject any
3049 supplied certificates, but the error message it gives is "no certificate
3050 supplied", which is not helpful.
3052 70. exigrep's output now also includes lines that are not associated with any
3053 message, but which match the given pattern. Implemented by a patch from
3054 Martin Sluka, which also tidied up the Perl a bit.
3056 71. Recipient callout verification, like sender verification, was using <> in
3057 the MAIL FROM command. This isn't really the right thing, since the actual
3058 sender may affect whether the remote host accepts the recipient or not. I
3059 have changed it to use the actual sender in the callout; this means that
3060 the cache record is now keyed on a recipient/sender pair, not just the
3061 recipient address. There doesn't seem to be a real danger of callout loops,
3062 since a callout by the remote host to check the sender would use <>.
3063 [SEE ABOVE: changed after hitting problems.]
3065 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3066 temporary errors. However, in the case of such a code being given after
3067 the end of a data transmission (i.e. after ".") Exim was failing to write
3068 a retry record for the message. (Yes, there was some broken host that was
3069 actually sending 8xx at this point.)
3071 73. An unknown lookup type in a host list could cause Exim to panic-die when
3072 the list was checked. (An example that provoked this was putting <; in the
3073 middle of a list instead of at the start.) If this happened during a DATA
3074 ACL check, a -D file could be left lying around. This kind of configuration
3075 error no longer causes Exim to die; instead it causes a defer errror. The
3076 incident is still logged to the main and panic logs.
3078 74. Buglet left over from Exim 3 conversion. The message "too many messages
3079 in one connection" was written to the rejectlog but not the mainlog, except
3080 when address rewriting (yes!) was being logged.
3082 75. Added write_rejectlog option.
3084 76. When a system filter was run not as root (that is, when system_filter_user
3085 was set), the values of the $n variables were not being returned to the
3086 main process; thus, they were not subsequently available in the $sn
3089 77. Added +return_path_on_delivery log selector.
3091 78. A connection timeout was being treated differently from recipients deferred
3092 when testing hosts_max_try with a message that was older than the host's
3093 retry timeout. (The host should not be counted, thus allowing all hosts to
3094 be tried at least once before bouncing.) This may have been the cause of an
3095 occasionally reported bug whereby a message would remain on the queue
3096 longer than the retry timeout, but would be bounced if a delivery was
3097 forced. I say "may" because I never totally pinned down the problem;
3098 setting up timeout/retry tests is difficult. See also the next item.
3100 79. The ultimate address timeout was not being applied to errors that involved
3101 a combination of host plus message (for example, a timeout on a MAIL
3102 command). When an address resolved to a number of possible hosts, and they
3103 were not all tried for each delivery (e.g. because of hosts_max_try), a
3104 message could remain on the queue longer than the retry timeout.
3106 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3109 81. Fixed an obscure SMTP outgoing bug which required at least the following
3110 conditions: (a) there was another message waiting for the same server;
3111 (b) the server returned 5xx to all RCPT commands in the first message so
3112 that the message was not completed; (c) the server dropped the connection
3113 or gave a negative response to the RSET that Exim sends to abort the
3114 transaction. The observed case was a dropped connection after DATA that had
3115 been sent in pipelining mode. That is, the server had advertised PIPELINING
3116 but was not implementing it correctly. The effect of the bug was incorrect
3117 behaviour, such as trying another host, and this could lead to a crash.
3123 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3124 and daemon.c were passed as pointers to ints; they should have been
3125 pointers to socklen_t variables (which are typically unsigned ints).
3127 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3130 3. Fixed a really odd bug that affected only the testing scheme; patching a
3131 certain fixed string in the binary changed the value of another string that
3132 happened to be identical to the end of the original first string.
3134 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3135 name", it returns that address as the IP address. On some operating
3136 systems (e.g. Solaris), it also passes back the IP address string as the
3137 "host name". However, on others (e.g. Linux), it passes back an empty
3138 string. Exim wasn't checking for this, and was changing the host name to an
3139 empty string, assuming it had been canonicized.
3141 5. Although rare, it is permitted to have more than one PTR record for a given
3142 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3143 all the names associated with an address, because they do in Solaris.
3144 However, it seems that they do not in Linux for data that comes from the
3145 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3146 I found this out when I moved to a new Linux workstation and tried to run
3147 the Exim test suite.
3149 To get round this problem I have changed the code so that it now does its
3150 own call to the DNS to look up PTR records when searching for a host name.
3151 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3152 addresses that are only in /etc/hosts are still found.
3154 This behaviour is, however, controlled by an option called host_lookup_
3155 order, which defaults to "bydns:byaddr". If people want to use the other
3156 order, or indeed, just use one or the other means of lookup, they can
3157 specify it in this variable.
3159 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3160 some operating systems, this comes back from gethostbyaddr() as an empty
3161 string, and this is what Exim used to test for. However, it seems that in
3162 other systems, "." is yielded. Exim now tests for this case too.
3164 7. The values of check_spool_space and check_log_space are now held internally
3165 as a number of kilobytes instead of an absolute number of bytes. If a
3166 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3167 kilobyte. This means that much larger values can be stored.
3169 8. Exim monitor: an attempt to get the action menu when not actually pointing
3170 at a message produces an empty menu entitled "No message selected". This
3171 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3172 no entries in it ("Shell widget menu has zero width and/or height"). So I
3173 have added a single, blank menu entry in this case.
3175 9. Added ${quote_local_part.
3177 10. MIME decoding is now applied to the contents of Subject: header lines when
3180 11. Now that a reference to $sender_host_address automatically causes a reverse
3181 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3182 host lookup before query-style lookups in lists that might use this
3183 variable. This has therefore been abolished, and the "net-" prefix is no
3184 longer necessary for query-style lookups.
3186 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3187 have been a typo for LFLAGS, so it has been changed.
3189 13. The install script calls Exim with "-C /dev/null" in order to find the
3190 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3191 to be output. Howeve, since Exim outputs its version number before the
3192 error, it didn't break the script. It just looked ugly. I fixed this by
3193 always allowing "-C /dev/null" if the caller is root.
3195 14. Ignore overlarge ACL variable number when reading spool file - insurance
3196 against a later release with more variables having written the file.
3198 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3199 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3200 this, as well as the form without the "IPv6" on the front.
3202 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3203 OS/Makefile-Darwin file.
3205 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3206 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3208 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3209 variable. However, some OS use ints for inodes. I've added cast to long int
3210 to get rid of the compiler warning.
3212 19. I had forgotten to lock out "/../" in configuration file names when
3213 ALT_CONFIG_PREFIX was set.
3215 20. Routers used for verification do not need to specify transports. However,
3216 if such a router generated a host list, and callout was configured, Exim
3217 crashed, because it could not find a port number from the (non-existent)
3218 transport. It now assumes port 25 in this circumstance.
3220 21. Added the -t option to exigrep.
3222 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3223 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3224 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3225 them is set, LOOKUP_LSEARCH is forced.
3227 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3229 24. Added sender and host information to the "rejected by local_scan()" log
3230 line; previously there was no indication of these.
3232 25. Added .include_if_exists.
3234 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3235 Linux. It turns out that not all file systems support this. Apparently some
3236 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3237 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3238 not supported on the file descriptor, is now ignored when Exim is trying to
3239 sync a directory. This applies only to Linux.
3241 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3243 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3244 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3247 29. Added support for maildirsize files from supplied patch (modified a bit).
3249 30. The use of :fail: followed by an empty string could lead Exim to respond to
3250 sender verification failures with (e.g.):
3252 550 Verification failed for <xxx>
3253 550 Sender verify failed
3255 where the first response line was missing the '-' that indicates it is not
3256 the final line of the response.
3258 31. The loop for finding the name of the user that called Exim had a hardwired
3259 limit of 10; it now uses the value of finduser_retries, which is used for
3260 all other user lookups.
3262 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3265 33. Exim was neglecting to zero errno before one call of strtol() when
3266 expanding a string and expecting an integer value. On some systems this
3267 resulted in spurious "integer overflow" errors. Also, it was casting the
3268 result into an int without checking.
3270 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3271 did not work. The code looks as if it has *never* worked, though it appears
3272 to have been documented since at least releast 1.62. I have made it work.
3274 35. The "timeout_DNS" error in retry rules, also documented since at least
3275 1.62, also never worked. As it isn't clear exactly what this means, and
3276 clearly it isn't a major issue, I have abolished the feature by treating it
3277 as "timeout", and writing a warning to the main and panic logs.
3279 36. The display of retry rules for -brt wasn't always showing the error code
3282 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3283 timeout_connect_A, timeout_connect_MX.
3285 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3286 to the empty sender.
3288 39. The daemon was not analysing the content of -oX till after it had closed
3289 stderr and disconnected from the controlling terminal. This meant that any
3290 syntax errors were only noted on the panic log, and the return code from
3291 the command was 0. By re-arranging the code a little, I've made the
3292 decoding happen first, so such errors now appear on stderr, and the return
3293 code is 1. However, the actual setting up of the sockets still happens in
3294 the disconnected process, so errors there are still only recorded on the
3297 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3298 connections (as happens on some IP stacks) was logged at start up time as
3299 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3300 it from "IPv6 and IPv4", which means that two separate sockets are being
3303 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3304 says whether AF_INET or AF_INET6 was passed as an argument.
3306 42. Exiwhat output was messed up when time zones were included in log
3309 43. Exiwhat now gives more information about the daemon's listening ports,
3310 and whether -tls-on-connect was used.
3312 44. The "port" option of the smtp transport is now expanded.
3314 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3315 silently ignored. Now an error message is written to the main and panic
3318 46. There's a new ACL modifier called "logwrite" which writes to a log file
3319 as soon as it is encountered.
3321 47. Added $local_user_uid and $local_user_gid at routing time.
3323 48. Exim crashed when trying to verify a sender address that was being
3326 49. Exim was recognizing only a space character after ".include". It now also
3327 recognizes a tab character.
3329 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3330 extracting the relevant information from the specification. The man page no
3331 longer contains scrambled data for the -d option, and I've added a section
3332 at the front about calling Exim under different names.
3334 51. Added "extra_headers" argument to the "mail" command in filter files.
3336 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3339 53. Installed eximstats 1.29.
3341 54. Added transport_filter_timeout as a generic transport option.
3343 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3344 Cc: header lines. This was required by RFC 822, but it not required by RFC
3347 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3348 incoming messages that did not have them. Now it does so only if the
3349 message originates locally, that is, if there is no associated remote host
3350 address. When Resent- header lines are present, this applies to the Resent-
3351 lines rather than the non-Resent- lines.
3353 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3354 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3356 58. Messages for configuration errors now include the name of the main
3357 configuration file - useful now that there may be more than one file in a
3358 list (.included file names were always shown).
3360 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3361 for those rare installations that do not start the daemon as root or run it
3362 setuid root. I've cut out the call to initgroups() if the daemon is not
3365 60. The Exim user and group can now be bound into the binary as text strings
3366 that are looked up at the start of Exim's processing.
3368 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3370 62. Added $mailstore_basename variable.
3372 63. Installed patch to sieve.c from Michael Haardt.
3374 64. When Exim failed to open the panic log after failing to open the main log,
3375 the original message it was trying to log was written to stderr and debug
3376 output, but if they were not available (the usual case in production), it
3377 was lost. Now it is written to syslog before the two lines that record the
3378 failures to open the logs.
3380 65. Users' Exim filters run in subprocesses under the user's uid. It is
3381 possible for a "deliver" command or an alias in a "personal" command to
3382 provoke an address rewrite. If logging of address rewriting is configured,
3383 this fails because the process is not running as root or exim. There may be
3384 a better way of dealing with this, but for the moment (because 4.30 needs
3385 to be released), I have disabled address rewrite logging when running a
3386 filter in a non-root, non-exim process.
3392 1. The buildconfig auxiliary program wasn't quoting the value set for
3393 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3394 not defined. This bug was masked in 4.22 by the effect that was fixed in
3397 2. Some messages that were rejected after a message id was allocated were
3398 shown as "incomplete" by exigrep. It no longer does this for messages that
3399 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3401 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3402 have allow_domain_literals set, the ID did not get logged in the <= line.
3403 Domain literals are now always recognized in Message-ID: header lines.
3405 4. The first argument for a ${extract expansion item is the key name or field
3406 number. Leading and trailing spaces in this item were not being ignored,
3407 causing some misleading effects.
3409 5. When deliver_drop_privilege was set, single queue runner processes started
3410 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3411 same command in the process it spins off) were not dropping privilege.
3413 6. When the daemon running as "exim" started a queue runner, it always
3414 re-executed Exim in the spun-off process. This is a waste of effort when
3415 deliver_drop_privilege is set. The new process now just calls the
3416 queue-runner function directly.
3422 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3425 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3426 ignored. Though the use of -oP was forcing the writing of a pid file, it
3427 was always written to the default place.
3429 3. If the message "no IP address found for host xxxx" is generated during
3430 incoming verification, it is now followed by identification of the incoming
3431 connection (so you can more easily find what provoked it).
3433 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3435 5. Added some features to "harden" Exim a bit more against certain attacks:
3437 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3438 be put in Local/Makefile. This is like the never_users runtime option,
3439 but it cannot be overridden. The default setting is "root".
3441 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3442 prefix string with which any file named in a -C command line option
3445 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3446 is retained for -C and -D only if the caller of Exim is root. Without
3447 it, the exim user may also use -C and -D and retain privilege.
3449 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3450 command line option is disabled.
3452 6. Macro names set by the -D option must start with an upper case letter, just
3453 like macro names defined in the configuration file.
3455 7. Added "dereference=" facility to LDAP.
3457 8. Two instances of the typo "uknown" in the source files are fixed.
3459 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3460 the Configure-Makefile script screwed up while processing it.
3462 10. Incorporated PCRE 4.4.
3464 11. The SMTP synchronization check was not operating right at the start of an
3465 SMTP session. For example, it could not catch a HELO sent before the client
3466 waited for the greeting. There is now a check for outstanding input at the
3467 point when the greeting is written. Because of the duplex, asynchronous
3468 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3469 way, but not yet received, when the check is performed.
3471 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3472 on TCP/IP sockets, because this apparently causes some broken clients to
3475 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3476 unchanged) from the Cygwin maintainer.
3478 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3479 was supported instead of testing for mbx. Effectively a typo.
3481 15. The spa authenticator server code was not checking that the input it
3482 received was valid base64.
3484 16. The debug output line for the "set" modifier in ACLs was not showing the
3485 name of the variable that was being set.
3487 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3489 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3490 lookup on its own. Something else was needed to trigger the lookup. For
3491 example, a match in host_lookup or the need for a host name in a host list.
3492 Now, if $sender_host_name is referenced and the host name has not yet been
3493 looked up, a lookup is performed. If the lookup fails, the variable remains
3494 empty, and $host_lookup_failed is set to "1".
3496 19. Added "eqi" as a case-independent comparison operator.
3498 20. The saslauthd authentication condition could segfault if neither service
3499 nor realm was specified.
3501 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3502 error message that was logged was misleading, and incoming SMTP
3503 connections were dropped. The message is now more accurate, and temporary
3504 errors are given to SMTP connections.
3506 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3507 (except RSET) with a 421 error, until QUIT is received. However, it was
3508 failing to send a response to QUIT.
3510 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3511 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3512 ACL is now run after the helo_try_verify_hosts code.
3514 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3515 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3516 case-independent, so other case variants are also recognized). Apparently
3517 some systems use these upper case variants.
3519 25. If more than two messages were waiting for the same host, and a transport
3520 filter was specified for the transport, Exim sent two messages over the
3521 same TCP/IP connection, and then failed with "socket operation on non-
3522 socket" when it tried to send the third.
3524 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3526 27. The extern definition of crypt16() in expand.c was not being excluded when
3527 the OS had its own crypt16() function.
3529 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3530 as a preferred synonym for return_size_limit, both as an option and as an
3533 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3535 30. Changed the default configuration ACL to relax the local part checking rule
3536 for addresses that are not in any local domains. For these addresses,
3537 slashes and pipe symbols are allowed within local parts, but the sequence
3538 /../ is explicitly forbidden.
3540 31. SPA server authentication was not clearing the challenge buffer before
3543 32. log_message in a "warn" ACL statement was writing to the reject log as
3544 well as to the main log, which contradicts the documentation and doesn't
3545 seem right (because no rejection is happening). So I have stopped it.
3547 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3548 However, I am unable to do any testing of this.
3550 34. Fixed an infelicity in the appendfile transport. When checking directories
3551 for a mailbox, to see if any needed to be created, it was accidentally
3552 using path names with one or more superfluous leading slashes; tracing
3553 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3555 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3556 contents are added to the log line that is written for every discarded
3557 recipient. (Previously a log_message setting was ignored.)
3559 36. The ${quote: operator now quotes the string if it is empty.
3561 37. The install script runs exim in order to find its version number. If for
3562 some reason other than non-existence or emptiness, which it checks, it
3563 could not run './exim', it was installing it with an empty version number,
3564 i.e. as "exim-". This error state is now caught, and the installation is
3567 38. An argument was missing from the function that creates an error message
3568 when Exim fails to connect to the socket for saslauthd authentication.
3569 This could cause Exim to crash, or give a corrupted message.
3571 39. Added isip, isip4, and isip6 to ${if conditions.
3573 40. The ACL variables $acl_xx are now saved with the message, and can be
3574 accessed later in routers, transports, and filters.
3576 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3577 strings in the file are not string-expanded.
3579 42. If a MAIL command specified a SIZE value that was too large to fit into an
3580 int variable, the check against message_size_limit failed. Such values are
3581 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3582 day this will have to be increased, but I don't think I want to be around
3583 when emails are that large.
3590 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3591 iconv() is not standard in FreeBSD.
3593 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3594 IPv6 enabled. The observed symptom was a segmentation fault on return from
3595 the function os_common_find_running_interfaces() in src/os.c.
3597 3. In the check_special_case() function in daemon.c I had used "errno" as an
3598 argument name, which causes warnings on some systems. This was basically a
3599 typo, since it was named "eno" in the comments!
3601 4. The code that waits for the clock to tick (at a resolution of some fraction
3602 of a second) so as to ensure message-id uniqueness was always waiting for
3603 at least one whole tick, when it could have waited for less. [This is
3604 almost certainly not relevant at current processor speeds, where it is
3605 unlikely to ever wait at all. But we try to future-proof.]
3607 5. The function that sleeps for a time interval that includes fractions of a
3608 second contained a race. It did not block SIGALRM between setting the
3609 timer, and suspending (a couple of lines later). If the interval was short
3610 and the sigsuspend() was delayed until after it had expired, the suspension
3611 never ended. On busy systems this could lead to processes getting stuck for
3614 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3615 process, before it forks any delivery processes. The open lookup caching
3616 mechanism meant that the open file or database connection was passed into
3617 the delivery process. The problem was that delivery processes always tidy
3618 up cached lookup data. This could cause a problem for the next delivery
3619 process started by the queue runner, because the external queue runner
3620 process does not know about the closure. So the next delivery process
3621 still has data in the lookup cache. In the case of a file lookup, there was
3622 no problem because closing a file descriptor in a subprocess doesn't affect
3623 the parent. However, if the lookup was caching a connection to a database,
3624 the connection was closed, and the second delivery process was likely to
3625 see errors such as "PGSQL: query failed: server closed the connection
3626 unexpectedly". The problem has been fixed by closing all cached lookups
3627 in a queue runner before running a delivery process.
3629 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3630 seem to have the "const" qualifier which it has on other OS. I've
3633 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3634 *of the same type* (client or server) with the same public name that an
3635 error should be diagnosed.
3637 9. When Exim looked up a host name for an IP address, but failed to find the
3638 original IP address when looking up the host name (a safety check), it
3639 output the message "<ip address> does not match any IP for NULL", which was
3640 confusing, to say the least. The bug was that the host name should have
3641 appeared instead of "NULL".
3643 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3644 user that is built into the binary, and the -C or -D options is used, root
3645 privilege is dropped before the configuration file is read. In addition,
3646 logging is switched to stderr instead of the normal log files. If the
3647 configuration then re-defines the Exim user, the unprivileged environment
3648 is probably not what is expected, so Exim logs a panic warning message (but
3651 However, if deliver_drop_privilege is set, the unprivileged state may well
3652 be exactly what is intended, so the warning has been cut out in that case,
3653 and Exim is allowed to try to write to its normal log files.
3659 1. smtp_return_error_details was not giving details for temporary sender
3660 or receiver verification errors.
3662 2. Diagnose a configuration error if two authenticators have the same public
3665 3. Exim used not to create the message log file for a message until the first
3666 delivery attempt. This could be confusing when incoming messages were held
3667 for policy or load reasons. The message log file is now created at the time
3668 the message is received, and an initial "Received" line is written to it.
3670 4. The automatically generated man page for command line options had a minor
3671 bug that caused no ill effects; however, a more serious problem was that
3672 the procedure for building the man page automatically didn't always
3673 operate. Consequently, release 4.20 contains an out-of-date version. This
3674 shouldn't happen again.
3676 5. When building Exim with embedded Perl support, the script that builds the
3677 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3678 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3680 6. The freeze_tell option was not being used for messages that were frozen on
3681 arrival, either by an ACL or by local_scan().
3683 7. Added the smtp_incomplete_transaction log selector.
3685 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3686 was accepting AUTH without a new EHLO.
3688 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3689 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3692 10. Logging of TCP/IP connections (when configured) now happens in the main
3693 daemon process instead of the child process, so that the TCP/IP connection
3694 count is more accurate (but it can never be perfect).
3696 11. The use of "drop" in a nested ACL was not being handled correctly in the
3697 outer ACL. Now, if condition failure induced by the nested "drop" causes
3698 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3699 or "require"), the connection is dropped.
3701 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3702 that yield "discard" can now be used with an "accept" or a "discard" verb,
3703 but an error is generated for any others (because I can't see a useful way
3704 to define what should happen).
3706 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3707 are now processed in the same way as lines in the Exim configuration file.
3708 In particular, continuation lines are supported.
3710 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3712 15. Added -ti meaning -t -i.
3714 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3715 domains, and warn by logging if others are found.
3717 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3718 returned from GIFCONF when Exim is trying to find the list of interfaces on
3719 a host. The code in os.c has been modified to copy each ifreq to an aligned
3720 structure in all cases.
3722 Also, in some cases, the returned ifreq's were being copied to a 'struct
3723 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3724 means the last couple of bytes of an IPv6 address could be chopped if the
3725 ifreq contained only a normal sockaddr (14 bytes storage).
3727 18. Named domain lists were not supported in the hosts_treat_as_local option.
3728 An entry such as +xxxx was not recognized, and was treated as a literal
3731 19. Ensure that header lines added by a DATA ACL are included in the reject log
3732 if the ACL subsequently rejects the message.
3734 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3735 MD5 (which is deprecated).
3737 21. When testing a filter file using -bf, Exim was writing a message when it
3738 took the sender from a "From " line in the message, but it was not doing so
3739 when it took $return_path from a Return-Path: header line. It now does.
3741 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3742 with a valid header line field name (a series of printing characters
3743 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3745 23. Changed "disc" in the source to "disk" to conform to the documentation and
3746 the book and for uniformity.
3748 24. Ignore Sendmail's -Ooption=value command line item.
3750 25. When execve() failed while trying to run a command in a pipe transport,
3751 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3752 could be confused with a return value of 69 from the command itself. This
3753 has been changed to 127, the value the shell returns if it is asked to run
3754 a non-existent command. The wording for the related log line suggests a
3755 non-existent command as the problem.
3757 26. If received_header_text expands to an empty string, do not add a Received:
3758 header line to the message. (Well, it adds a token one on the spool, but
3759 marks it "old" so that it doesn't get used or transmitted.)
3761 27. Installed eximstats 1.28 (addition of -nt option).
3763 28. There was no check for failure on the call to getsockname() in the daemon
3764 code. This can fail if there is a shortage of resources on the system, with
3765 ENOMEM, for example. A temporary error is now given on failure.
3767 29. Contrary to the C standard, it seems that in some environments, the
3768 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3769 program. Exim now does this explicitly; it affects the formatting of
3770 timestamps using strftime().
3772 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3773 complaints. I've now initialized all the variables, to avoid this.
3775 32. Header lines added by a system filter were not being "seen" during
3776 transport-time rewrites.
3778 33. The info_callback() function passed to OpenSSL is set up with type void
3779 (*)(SSL *, int, int), as described somewhere. However, when calling the
3780 function (actually a macro) that sets it up, the type void(*)() is
3781 expected. I've put in a cast to prevent warnings from picky compilers.
3783 34. If a DNS black list lookup found a CNAME record, but there were no A
3784 records associated with the domain it pointed at, Exim crashed.
3786 35. If a DNS black list lookup returned more than one A record, Exim ignored
3787 all but the first. It now scans all returned addresses if a particular IP
3788 value is being sought. In this situation, the contents of the
3789 $dnslist_value variable are a list of all the addresses, separated by a
3792 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3793 to accept a host name and all its aliases if the forward lookup for any of
3794 them yielded the IP address of the incoming connection. Now it accepts only
3795 those names whose forward lookup yields the correct IP address. Any other
3796 names are discarded. This closes a loophole whereby a rogue DNS
3797 administrator could create reverse DNS records to break through a
3798 wildcarded host restriction in an ACL.
3800 37. If a user filter or a system filter that ran in a subprocess used any of
3801 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3802 the wrong values were passed to the pipe command ($thisaddress had the
3803 value of $0, $0 had the value of $1, etc). This bug was introduced by
3804 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3806 38. Improved the line breaking for long SMTP error messages from ACLs.
3807 Previously, if there was no break point between 40 and 75 characters, Exim
3808 left the rest of the message alone. Two changes have been made: (a) I've
3809 reduced the minimum length to 35 characters; (b) if it can't find a break
3810 point between 35 and 75 characters, it looks ahead and uses the first one
3811 that it finds. This may give the occasional overlong line, but at least the
3812 remaining text gets split now.
3814 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3815 file descriptors might be low, and that setting 1000 would always raise it.
3816 It turns out that in some environments, the limit is already over 1000 and
3817 that lowering it causes trouble. So now Exim takes care not to decrease it.
3819 40. When delivering a message, the value of $return_path is set to $sender_
3820 address at the start of routing (routers may change the value). By an
3821 oversight, this default was not being set up when an address was tested by
3822 -bt or -bv, which affected the outcome if any router or filter referred to
3825 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3826 log only when "message" or "log_message" are set. However, if one of the
3827 conditions was an address verification, or a call to a nested ACL, the
3828 messages generated by the underlying test were being passed through. This
3829 no longer happens. The underlying message is available in $acl_verify_
3830 message for both "message" and "log_message" expansions, so it can be
3831 passed through if needed.
3833 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3834 new expansion $bh_ to give the encoded byte string without charset
3835 translation. Translation happens only if iconv() is available; HAVE_ICONV
3836 indicates this at build time. HEADERS_CHARSET gives the charset to
3837 translate to; headers_charset can change it in the configuration, and
3838 "headers charset" can change it in an individual filter file.
3840 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3841 that creates RFC 2047 encoded "words" labels them as that charset instead
3842 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3843 expansion operator; (ii) when Exim creates a From: line for a local
3844 message; (iii) when a header line is rewritten to include a "phrase" part.
3846 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3847 buggy, causing it to skip the first lines of messages whose message ID
3848 ended in 'D'. This would not have bitten before Exim release 4.14, because
3849 message IDs were unlikely to end in 'D' before then. The effect was to have
3850 incorrect size information for certain domains.
3852 45. #include "config.h" was missing at the start of the crypt16.c module. This
3853 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3856 46. If there was a timeout during a "random" callout check, Exim treated it as
3857 a failure of the random address, and carried on sending RSET and the real
3858 address. If the delay was just some slowness somewhere, the response to the
3859 original RCPT would be taken as a response to RSET and so on, causing
3860 mayhem of various kinds.
3862 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3863 when I implemented it. It didn't allow for the fact that some option values
3864 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3865 the right test for positive values.
3867 48. Domain names in DNS records are case-independent. Exim always looks them up
3868 in lower case. Some resolvers return domain names in exactly the case they
3869 appear in the zone file, that is, they may contain uppercase letters. Not
3870 all resolvers do this - some return always lower case. Exim was treating a
3871 change of case by a resolver as a change of domain, similar to a widening
3872 of a domain abbreviation. This triggered its re-routing code and so it was
3873 trying to route what was effectively the same domain again. This normally
3874 caused routing to fail (because the router wouldn't handle the domain
3875 twice). Now Exim checks for this case specially, and just changes the
3876 casing of the domain that it ultimately uses when it transmits the message
3879 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3882 50. If a filter generated a file delivery with a non-absolute name (possible if
3883 no home directory exists for the router), the forbid_file option was not
3886 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3887 the existing equality matching.
3889 52. Exim was using ints instead of ino_t variables in some places where it was
3890 dealing with inode numbers.
3892 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3893 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3894 called TMPDIR, and if it finds it is different, it changes its value.
3896 54. The smtp_printf() function is now made available to local_scan() so
3897 additional output lines can be written before returning. There is also an
3898 smtp_fflush() function to enable the detection of a dropped connection.
3899 The variables smtp_input and smtp_batched_input are exported to
3902 55. Changed the default runtime configuration: the message "Unknown user"
3903 has been removed from the ACL, and instead placed on the localuser router,
3904 using the cannot_route_message feature. This means that any verification
3905 failures that generate their own messages won't get overridden. Similarly,
3906 the "Unrouteable address" message that was in the ACL for unverifiable
3907 relay addresses has also been removed.
3909 56. Added hosts_avoid_esmtp to the smtp transport.
3911 57. The exicyclog script was not checking for the esoteric option
3912 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3913 will work only if exicyclog is run under the appropriate euid.
3915 58. Following a discussion on the list, the rules by which Exim recognises line
3916 endings on incoming messages have been changed. The -dropcr and drop_cr
3917 options are now no-ops, retained only for backwards compatibility. The
3918 following line terminators are recognized: LF CRLF CR. However, special
3919 processing applies to CR:
3921 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3922 nor a local message in the state where . is a terminator.
3924 (ii) If a bare CR is encountered in a header line, an extra space is added
3925 after the line terminator so as not to end the header. The reasoning
3926 behind this is that bare CRs in header lines are most likely either
3927 to be mistakes, or people trying to play silly games.
3929 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3930 was being incorrectly given as 18 bytes larger than it should have been.
3931 This is a VOB (very old bug).
3933 60. This may never have affected anything current, but just in case it has:
3934 When the local host is found other than at the start of a list of hosts,
3935 the local host, those with the same MX, and any that follow, are discarded.
3936 When the list in question was part of a longer list of hosts, the following
3937 hosts (not currently being processed) were also being discarded. This no
3938 longer happens. I'm not sure if this situation could ever has previously
3941 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3944 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3945 header. This is something that is recommended in a new Internet Draft, and
3946 is something that is documented as being done by Sendmail. There are two
3947 possible values. For messages generated by the autoreply transport, Exim
3950 Auto-Submitted: auto-replied
3952 whereas for all other generated messages (e.g. bounces) it adds
3954 Auto-Submitted: auto-generated
3956 63. The "personal" condition in filters now includes a test for the
3957 Auto-Submitted: header. If it contains the string "auto-" the message it
3958 not considered personal.
3960 64. Added rcpt_include_affixes as a generic transport option.
3962 65. Added queue_only_override (default true).
3964 66. Added the syslog_duplication option.
3966 67. If what should have been the first header line of a message consisted of
3967 a space followed by a colon, Exim was mis-interpreting it as a header line.
3968 It isn't of course - it is syntactically invalid and should therefore be
3969 treated as the start of the message body. The misbehaviour could have
3970 caused a number of strange effects, including loss of data in subsequent
3971 header lines, and spool format errors.
3973 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3974 client host had authenticated. This control can now be exercised by an ACL
3975 for more flexibility.
3977 69. By default, callouts do not happen when testing with -bh. There is now a
3978 variant, -bhc, which does actually run the callout code, including
3979 consulting and updating the callout cache.
3981 70. Added support for saslauthd authentication, courtesy of Alexander
3984 71. If statvfs() failed on the spool or log directories while checking their
3985 size for availability, Exim confusingly gave the error "space shortage".
3986 Furthermore, in debugging mode it crashed with a floating point exception.
3987 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3988 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3989 serious problem, Exim now writes to the main and panic logs when this
3990 happens, with details of the failure. It then refuses to accept the
3991 incoming message, giving the message "spool directory problem" or "log
3992 directory problem" with a 421 code for SMTP messages.
3994 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3995 0, 1, and 2 exist, because some OS complain for execs without them (see
3996 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3997 descriptors. However, the code omitted to check that the open succeeded,
3998 causing mysterious errors if for some reason the permissions on /dev/null
3999 got screwed. Now Exim writes a message to the main and panic logs, and
4000 bombs out if it can't open /dev/null.
4002 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4003 interact so that it is all more flexible. It is supposed to remain
4004 backwards compatible. Also added extra_local_interfaces.
4006 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4007 to bomb out with an assertion failure - to the client this appears as a
4008 connection drop. This problem occurs in the part of the code that was taken
4009 from the Samba project. Fortunately, the assertion is in a very simple
4010 function, so I have fixed this by reproducing the function inline in the
4011 one place where it is called, and arranging for authentication to fail
4012 instead of killing the process with assert().
4014 75. The SPA client code was not working when the server requested OEM rather
4015 than Unicode encoding.
4017 76. Added code to make require_files with a specific uid setting more usable in
4018 the case where statting the file as root fails - usually a non-root-mounted
4019 NFS file system. When this happens and the failure is EACCES, Exim now
4020 forks a subprocess and does the per-uid checking as the relevant uid.
4022 77. Added process_log_path.
4024 78. If log_file_path was not explicitly set, a setting of check_log_space or
4025 check_log_inodes was ignored.
4027 79. If a space check for the spool or log partitions fails, the incident is now
4028 logged. Of course, in the latter case the data may get lost...
4030 80. Added the %p formatting code to string_format() so that it can be used to
4031 print addresses in debug_print(). Adjusted all the address printing in the
4032 debugging in store.c to use %p rather than %d.
4034 81. There was a concern that a line of code in smtp_in.c could overflow a
4035 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4036 initially expressed, the concern was not well-founded, because trailing
4037 spaces are removed early. However, if the trailing spaces were followed by
4038 a NULL, they did not get removed, so the overflow was possible. Two fixes
4041 (a) I re-wrote the offending code in a cleaner fashion.
4042 (b) If an incoming SMTP command contains a NULL character, it is rejected
4045 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4046 runs initgroups(), so that if the Exim user is in any additional groups,
4047 they will be used during message reception.
4053 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob