1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.365 2006/06/30 15:36:08 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.)
74 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
75 other effects) broke the use of negated acl sub-conditions.
77 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
80 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
81 "Deny" causes Exim to reject the incoming connection with a 554 error.
82 Unfortunately, if there is a major crisis, such as a disk failure,
83 tcp-wrappers gives "deny", whereas what one would like would be some
84 kind of temporary error. A kludge has been added to help with this.
85 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
86 554 error is used if errno is still zero or contains ENOENT (which occurs
87 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
90 PH/03 Add -lutil to the default FreeBSD LIBS setting.
92 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
93 errors. Otherwise a message that provokes a temporary error (when other
94 messages do not) can cause a whole host to time out.
96 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
97 addresses were routed directly to files or pipes from a redirect router.
98 File deliveries just didn't batch; pipe deliveries might have suffered
101 PH/06 A failure to get a lock for a hints database would erroneously always say
102 "Failed to get write lock", even when it was really a read lock.
104 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
105 of 0600. This has been changed to use the value of the lockfile_mode
106 option (which defaults to 0600).
108 PH/08 Applied small patch from the Sieve maintainer.
110 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
111 folder from quota calculations, a direct delivery into this folder messed
112 up the contents of the maildirsize file. This was because the regex was
113 used only to exclude .Trash (or whatever) when the size of the mailbox
114 was calculated. There was no check that a delivery was happening into an
115 excluded directory. This bug has been fixed by ignoring all quota
116 processing for deliveries into excluded directories.
118 PH/10 Added the maildirfolder_create_regex option to appendfile.
124 PH/01 The code for finding all the local interface addresses on a FreeBSD
125 system running IPv6 was broken. This may well have applied to all BSD
126 systems, as well as to others that have similar system calls. The broken
127 code found IPv4 interfaces correctly, but gave incorrect values for the
128 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
129 that it would not match correctly against @[] and not recognize the IPv6
132 PH/02 The ipliteral router was not recognizing addresses of the form user@
133 [ipv6:....] because it didn't know about the "ipv6:" prefix.
135 PH/03 Added disable_ipv6.
137 PH/04 Changed $reply_address to use the raw form of the headers instead of the
138 decoded form, because it is most often used to construct To: headers
139 lines in autoreplies, and the decoded form may well be syntactically
140 invalid. However, $reply_address has leading white space removed, and all
141 newlines turned into spaces so that the autoreply transport does not
144 PH/05 If group was specified without a user on a router, and no group or user
145 was specified on a transport, the group from the router was ignored.
147 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
148 for visible compile-time settings that can be used to change these
149 numbers, for those that want even more. Backwards compatibility with old
150 spool files has been maintained. However, going back to a previous Exim
151 release will lost any variables that are in spool files.
153 PH/07 Two small changes when running in the test harness: increase delay when
154 passing a TCP/IP connection to a new process, in case the original
155 process has to generate a bounce, and remove special handling of
156 127.0.0.2 (sic), which is no longer necessary.
158 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
159 be the same on different OS.
161 PH/09 Moved a debug statement in filter processing to avoid a race problem when
164 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
165 whether --show-vars was specified or not
167 JJ/02 exipick: Added support for new ACL variable spool format introduced
170 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
171 syntactically invalid From: or Reply-to: line, and a filter used this to
172 generate an autoreply, and therefore failed to obtain an address for the
173 autoreply, Exim could try to deliver to a non-existent relative file
174 name, causing unrelated and misleading errors. What now happens is that
175 it logs this as a hard delivery error, but does not attempt to create a
178 PH/11 The exinext utility has a -C option for testing purposes, but although
179 the given file was scanned by exinext itself; it wasn't being passed on
180 when Exim was called.
182 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
183 an end-of-file indication when reading a command response.
185 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
186 compiled. In many other places in Exim, IPv6 addresses are always
187 recognized, so I have changed this. It also means that IPv4 domain
188 literals of the form [IPV4:n.n.n.n] are now always recognized.
190 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
191 used if the router is not running as root, for example, when verifying at
192 ACL time, or when using -bh. The debugging output from this situation was
193 non-existent - all you got was a failure to exec. I have made two
196 (a) Failures to set uid/gid, the current directory, or a process leader
197 in a subprocess such as that created by queryprogram now generate
198 suitable debugging ouput when -d is set.
200 (b) The queryprogram router detects when it is not running as root,
201 outputs suitable debugging information if -d is set, and then runs
202 the subprocess without attempting to change uid/gid.
204 PH/15 Minor change to Makefile for building test_host (undocumented testing
207 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
208 additional section of a DNS packet that returns MX or SRV records.
209 Instead, it always explicitly searches for A/AAAA records. This avoids
210 major problems that occur when a DNS server includes only records of one
211 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
212 fixed another bug: if SRV records were looked up and the corresponding
213 address records were *not* found in the additional section, the port
214 values from the SRV records were lost.
216 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
217 using the correct key (the original address) when searching the retry
218 rules in order to find which one to use for generating the retry hint.
220 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
221 adding the default one. Similarly, if it contains a Reply-To: header, the
222 errors_reply_to option, if set, is not used.
224 PH/19 When calculating a retry time, Exim used to measure the "time since
225 failure" by looking at the "first failed" field in the retry record. Now
226 it does not use this if it is later than than the arrival time of the
227 message. Instead it uses the arrival time. This makes for better
228 behaviour in cases where some deliveries succeed, thus re-setting the
229 "first failed" field. An example is a quota failure for a huge message
230 when small messages continue to be delivered. Without this change, the
231 "time since failure" will always be short, possible causing more frequent
232 delivery attempts for the huge message than are intended.
233 [Note: This change was subsequently modified - see PH/04 for 4.62.]
235 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
236 $1, $2, $3) because the numerical variables can be reset during some
237 expansion items (e.g. "match"), thereby losing the authentication data.
239 PH/21 Make -bV show the size of off_t variables so that the test suite can
240 decide whether to run tests for quotas > 2G.
242 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
243 mailbox_size, and mailbox_filecount in the appendfile transport. If a
244 filecount value is greater than 2G or if a quota value is greater than 2G
245 on a system where the size of off_t is not greater than 4, a panic error
248 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
249 never match. The debug and -bh output now contains an explicit error
250 message indicating a malformed IPv4 address or mask.
252 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
253 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
256 PH/25 Do not write to syslog when running in the test harness. The only
257 occasion when this arises is a failure to open the main or panic logs
258 (for which there is an explicit test).
260 PH/26 Added the /no_tell option to "control=freeze".
262 PH/27 If a host name lookup failed very early in a connection, for example, if
263 the IP address matched host_lookup and the reverse lookup yielded a name
264 that did not have a forward lookup, an error message of the form "no IP
265 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
266 could be logged. Now it outputs the IP address instead of "NULL".
268 PH/28 An enabling patch from MH: add new function child_open_exim2() which
269 allows the sender and the authenticated sender to be set when
270 submitting a message from within Exim. Since child_open_exim() is
271 documented for local_scan(), the new function should be too.
273 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
274 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
275 results in an empty string is now treated as unset.
277 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
279 PH/31 Added sender_verify_fail logging option.
281 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
282 needed by RFC 822 but not by RFC 2822 was commented out. I have now
283 tidied the source and removed it altogether.
285 PH/33 When a queue run was abandoned because the load average was too high, a
286 log line was always written; now it is written only if the queue_run log
287 selector is set. In addition, the log line for abandonment now contains
288 information about the queue run such as the pid. This is always present
289 in "start" and "stop" lines but was omitted from the "abandon" line.
291 PH/34 Omit spaces between a header name and the colon in the error message that
292 is given when verify = headers_syntax fails (if there are lots of them,
293 the message gets confusing).
295 PH/35 Change the default for dns_check_names_pattern to allow slashes within
296 names, as there are now some PTR records that contain slashes. This check
297 is only to protect against broken name servers that fall over on strange
298 characters, so the fact that it applies to all lookups doesn't matter.
300 PH/36 Now that the new test suite is complete, we can remove some of the
301 special code in Exim that was needed for the old test suite. For example,
302 sorting DNS records because real resolvers return them in an arbitrary
303 order. The new test suite's fake resolver always returns records in the
306 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
307 bounces) except when queue_only is set, to avoid logging races between
308 the different processes.
310 PH/38 Panic-die if .include specifies a non-absolute path.
312 PH/39 A tweak to the "H" retry rule from its user.
314 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
315 a label. They prevented compilation on older perls.
317 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
318 a warning to be raised on newish perls.
320 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
321 on queue. Changes to match documented behaviour of showing count of
322 messages matching specified criteria.
324 PH/40 Changed the default ident timeout from 30s to 5s.
326 PH/41 Added support for the use of login_cap features, on those BSD systems
327 that have them, for controlling the resources used by pipe deliveries.
329 PH/42 The content-scanning code uses fopen() to create files in which to put
330 message data. Previously it was not paying any attention to the mode of
331 the files. Exim runs with umask(0) because the rest of the code creates
332 files with open(), and sets the required mode explicitly. Thus, these
333 files were ending up world-writeable. This was not a big issue, because,
334 being within the spool directory, they were not world-accessible. I have
335 created a function called modefopen, which takes an additional mode
336 argument. It sets umask(777), creates the file, chmods it to the required
337 mode, then resets the umask. All the relevant calls to fopen() in the
338 content scanning code have been changed to use this function.
340 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
341 to 24 hours. This avoids potential overflow problems when processing G
342 and H retry rules. I suspect nobody ever tinkers with this value.
344 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
346 PH/45 When the plaintext authenticator is running as a client, the server's
347 challenges are checked to ensure they are valid base64 strings. By
348 default, the authentication attempt is cancelled if an invalid string is
349 received. Setting client_ignore_invalid_base64 true ignores these errors.
350 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
351 they are received. Thus, the responses can be made to depend on the
352 challenges. If an invalid string is ignored, an empty string is placed in
355 PH/46 Messages that are created by the autoreply transport now contains a
356 References: header, in accordance with RFCs 2822 and 3834.
358 PH/47 Added authenticated_sender_force to the smtp transport.
360 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
362 PH/49 Installed latest patch from the Sieve maintainer.
364 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
365 was also set, the appendfile transport was unnecessarily scanning a
366 directory of message files (e.g. for maildir delivery) to find the count
367 of files (along with the size), even though it did not need this
368 information. It now does the scan only if it needs to find either the
369 size of the count of files.
371 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
373 PH/52 Two bugs concerned with error handling when the smtp transport is
376 (i) Exim was not creating retry information for temporary errors given
377 for individual recipients after the DATA command when the smtp transport
378 was used in LMTP mode. This meant that they could be retried too
379 frequently, and not timed out correctly.
381 (ii) Exim was setting the flag that allows error details to be returned
382 for LMTP errors on RCPT commands, but not for LMTP errors for individual
383 recipients that were returned after the DATA command.
385 PH/53 This is related to PH/52, but is more general: for any failing address,
386 when detailed error information was permitted to be returned to the
387 sender, but the error was temporary, then after the final timeout, only
388 "retry timeout exceeded" was returned. Now it returns the full error as
389 well as "retry timeout exceeded".
391 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
392 do this, and (what is worse) MTAs that accept it.
394 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
395 will now be deprecated.
397 PH/56 New os.c-cygwin from the Cygwin maintainer.
399 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
400 formats (previously only available in exim formats via -bpr, -bpru,
401 and -bpra. Now also available in native and exiqgrep formats)
403 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
404 with very large, slow to parse queues
406 JJ/08 exipick: added ! as generic prefix to negate any criteria format
408 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
410 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
411 responses to authentication challenges, though it was showing the
412 challenges; (ii) I've removed the CR characters from the debug output for
415 PH/58 Allow for the insertion of a newline as well as a space when a string
416 is turned into more than one encoded-word during RFC 2047 encoding. The
417 Sieve code now uses this.
419 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
420 data_4xx, lost_connection, tls_required.
422 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
423 message was being sent as an SMTP response.
425 PH/61 Add -l and -k options to exicyclog.
427 PH/62 When verifying, if an address was redirected to one new address, so that
428 verification continued, and the new address failed or deferred after
429 having set something in $address_data, the value of $address_data was not
430 passed back to the ACL. This was different to the case when no
431 redirection occurred. The value is now passed back in both cases.
433 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
434 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
435 use login_cap.h, so on its own it isn't the distinguishing feature. The
436 new name refers directly to the setclassresources() function.
438 PH/65 Added configuration files for NetBSD3.
440 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
442 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
443 is preferred over IPv4.
445 PH/68 The bounce_return_message and bounce_return_body options were not being
446 honoured for bounces generated during the reception of non-SMTP messages.
447 In particular, this applied to messages rejected by the ACL. This bug has
448 been fixed. However, if bounce_return_message is true and bounce_return_
449 body is false, the headers that are returned for a non-SMTP message
450 include only those that have been read before the error was detected.
451 (In the case of an ACL rejection, they have all been read.)
453 PH/69 The HTML version of the specification is now built in a directory called
454 spec_html instead of spec.html, because the latter looks like a path with
455 a MIME-type, and this confuses some software.
457 PH/70 Catch two compiler warnings in sieve.c.
459 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
460 function verify_get_ident() calls ip_connect() to connect a socket, but
461 if the "connect()" function timed out, ip_connect() used to close the
462 socket. However, verify_get_ident() also closes the socket later, and in
463 between Exim writes to the log, which may get opened at this point. When
464 the socket was closed in ip_connect(), the log could get the same file
465 descriptor number as the socket. This naturally causes chaos. The fix is
466 not to close the socket in ip_connect(); the socket should be closed by
467 the function that creates it. There was only one place in the code where
468 this was missing, in the iplookup router, which I don't think anybody now
469 uses, but I've fixed it anyway.
471 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
472 well as to direct DNS lookups. Otherwise the handling of names in host
473 lists is inconsistent and therefore confusing.
479 PH/01 Two changes to the default runtime configuration:
481 (1) Move the checks for relay_from_hosts and authenticated clients from
482 after to before the (commented out) DNS black list checks.
484 (2) Add control=submission to the relay_from_hosts and authenticated
485 clients checks, on the grounds that messages accepted by these
486 statements are most likely to be submissions.
488 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
490 (1) Generate an error if the third argument for the ${prvs expansion is
493 (2) Treat a missing third argument of ${prvscheck as if it were an empty
496 (3) Reset the variables that are obtained from the first argument of
497 ${prvscheck and used in the second argument before leaving the code,
498 because their memory is reclaimed, so using them afterwards may do
501 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
502 one (it's much easier than Tom thought :-).
504 (5) Because of (4), we can now allow for the use of $prvscheck_result
505 inside the third argument.
507 PH/03 For some reason, the default setting of PATH when running a command from
508 a pipe transport was just "/usr/bin". I have changed it to
511 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
512 anything to be listed in the output from -bV.
514 PH/05 When a filter generated an autoreply, the entire To: header line was
515 quoted in the delivery log line, like this:
517 => >A.N.Other <ano@some.domain> <original@ddress> ...
519 This has been changed so that it extracts the operative address. There
520 may be more than one such address. If so, they are comma-separated, like
523 => >ano@some.domain,ona@other.domain <original@ddress> ...
525 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
526 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
527 not being looked up in the rDNS to get a host name, Exim was showing the
528 IP address twice in Received: lines, even though the IP addresses were
529 identical. For example:
531 Received: from [1.2.3.4] (helo=[1.2.3.4])
533 However, if the real host name was known, it was omitting the HELO data
534 if it matched the actual IP address. This has been tidied up so that it
535 doesn't show the same IP address twice.
537 PH/07 When both +timestamp and +memory debugging was on, the value given by
538 $tod_xxx expansions could be wrong, because the tod_stamp() function was
539 called by the debug printing, thereby overwriting the timestamp buffer.
540 Debugging no longer uses the tod_stamp() function when +timestamp is set.
542 PH/08 When the original message was included in an autoreply transport, it
543 always said "this is a copy of the message, including all the headers",
544 even if body_only or headers_only was set. It now gives an appropriate
547 PH/09 Applied a patch from the Sieve maintainer which:
549 o fixes some comments
550 o adds the (disabled) notify extension core
551 o adds some debug output for the result of if/elsif tests
552 o points to the current vacation draft in the documentation
553 and documents the missing references header update
557 o fixes a bug in processing the envelope test (when testing
558 multiple envelope elements, the last element determinted the
561 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
562 Electronic Mail") by including:
564 Auto-submitted: auto-generated
566 in the messages that it generates (bounce messages and others, such as
567 warnings). In the case of bounce messages for non-SMTP mesages, there was
568 also a typo: it was using "Auto_submitted" (underscore instead of
569 hyphen). Since every message generated by Exim is necessarily in response
570 to another message, thes have all been changed to:
572 Auto-Submitted: auto-replied
574 in accordance with these statements in the RFC:
576 The auto-replied keyword:
578 - SHOULD be used on messages sent in direct response to another
579 message by an automatic process,
581 - MUST NOT be used on manually-generated messages,
583 - MAY be used on Delivery Status Notifications (DSNs) and Message
584 Disposition Notifications (MDNs),
586 - MUST NOT be used on messages generated by automatic or periodic
587 processes, except for messages which are automatic responses to
590 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
591 to the default Received: header definition.
593 PH/12 Added log selector acl_warn_skipped (default on).
595 PH/13 After a successful wildlsearch lookup, discard the values of numeric
596 variables because (a) they are in the wrong storage pool and (b) even if
597 they were copied, it wouldn't work properly because of the caching.
599 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
600 checking when decoding. Apparently there are clients that generate
601 overlong encoded strings. Why am I not surprised?
603 PH/15 If the first argument of "${if match_address" was not empty, but did not
604 contain an "@" character, Exim crashed. Now it writes a panic log message
605 and treats the condition as false.
607 PH/16 In autoreply, treat an empty string for "once" the same as unset.
609 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
610 extension "envelope-auth". The code is finished and in agreement with
611 other implementations, but there is no documentation so far and in fact,
612 nobody wrote the draft yet. This extension is currently #undef'ed, thus
613 not changing the active code.
615 Print executed "if" and "elsif" statements when debugging is used. This
616 helps a great deal to understand what a filter does.
618 Document more things not specified clearly in RFC3028. I had all this
619 sorted out, when out of a sudden new issues came to my mind. Oops."
621 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
624 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
625 canonical form (as documented). However, after a host name lookup from
626 the IP address, check_host() was doing a simple string comparison with
627 addresses acquired from the DNS when checking that the found name did
628 have the original IP as one of its addresses. Since any found IPv6
629 addresses are likely to be in abbreviated form, the comparison could
630 fail. Luckily, there already exists a function for doing the comparison
631 by converting both addresses to binary, so now that is used instead of
634 PH/20 There was another similar case to PH/19, when a complete host name was
635 given in a host list; looking up its IP address could give an abbreviated
636 form, whereas the current host's name might or might not be abbreviated.
637 The same fix has been applied.
643 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
644 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
647 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
648 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
650 PH/03 Typo: missing ".o" in src/pcre/Makefile.
652 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
653 header line, restrict the check to what is listed in RFCs 2369 and 2929.
654 Also, for "Auto-Submitted", treat anything other than "no" as
655 non-personal, in accordance with RFC 3834. (Previously it treated
656 anything starting "auto-" as non-personal.)
658 TF/01 The control=submission/name=... option had a problem with syntax
659 errors if the name included a slash character. The /name= option
660 now slurps the rest of the string, so it can include any characters
661 but it must come last in the list of options (after /sender_retain
664 PH/05 Some modifications to the interface to the fake nameserver for the new
672 TK/01 Added the "success_on_redirect" address verification option. See
673 NewStuff for rationale and an example.
675 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
677 PH/02 Patch to exigrep to allow it to work on syslog lines.
679 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
680 fread() to skip over the body file's header line, because in Cygwin the
681 header line is locked and is inaccessible.
683 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
684 co-exist for some time) to make it clear that it is the Exim ID that is
685 referenced, not the Message-ID: header line.
687 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
688 string_format() function, because snprintf() does not exist on all
691 PH/06 The use of forbid_filter_existstest now also locks out the use of the
692 ${stat: expansion item.
694 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
695 protocol synchronization error", to keep the pedants happy.
697 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
698 well as for IRIX systems, when gcc is being used. See the host.c source
701 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
703 PH/10 Named domain lists were not working if used in a queue_smtp_domains
706 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
707 transport and to the smtp transport in LMTP mode.
709 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
711 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
712 run a filter in a subprocess. This could lead to confusion in subsequent
713 lookups in the parent process. There should also be a search_tidyup() at
714 the end of the subprocess.
716 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
717 only if the host matched helo_try_verify_hosts, which caused the
718 verification to occur when the EHLO/HELO command was issued. The ACL just
719 tested the remembered result. Now, if a previous verification attempt has
720 not happened, "verify = helo" does it there and then.
722 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
724 TK/03 Fix log output including CR from clamd.
726 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
727 exist provoked a memory error which could cause a segfault.
729 PH/15 Installed PCRE 6.2
731 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
733 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
734 of the problem. Specifically, suggested +O2 rather than +O1 for the
737 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
739 PH/20 If a delivery was routed to a non-standard port by means of an SRV
740 record, the port was not correctly logged when the outgoing_port log
741 selector was set (it logged the transort's default port).
743 PH/21 Added support for host-specific ports to manualroute, queryprogram,
744 fallback_hosts, and "hosts" in the smtp transport.
746 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
747 host errors such as "Connection refused".
749 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
750 authentication with radiusclient 0.4.9:
752 - Error returned from rc_read_config was caught wrongly
753 - Username/password not passed on to radius server due to wrong length.
755 The presumption is that some radiusclient API changes for 4.51/PH/17
756 were not taken care of correctly. The code is still untested by me (my
757 Linux distribution still has 0.3.2 of radiusclient), but it was
758 contributed by a Radius user.
760 PH/24 When doing a callout, the value of $domain wasn't set correctly when
761 expanding the "port" option of the smtp transport.
763 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
764 while reading a MIME header. Thanks to Tom Hughes for a patch.
766 PH/24 Include config.h inside local_scan.h so that configuration settings are
769 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
770 that in an ACL for RCPT (for example), you can examine exactly what was
773 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
774 commands, but it was not correctly comparing the address with the actual
775 client host address. Thus, it would show the EHLO address in Received:
776 header lines when this was not necessary.
778 PH/27 Added the % operator to ${eval:}.
780 PH/28 Exim tries to create and chdir to its spool directory when it starts;
781 it should be ignoring failures (because with -C, for example, it has lost
782 privilege). It wasn't ignoring creation failures other than "already
785 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
788 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
789 because an input line was too long, either on its own, or by virtue of
790 too many continuations, the temporary file was not being removed, and the
791 return code was incorrect.
793 PH/31 Missing "BOOL" in function definition in filtertest.c.
795 PH/32 Applied Sieve patches from the maintainer.
797 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
799 PH/33 Added "verify = not_blind".
801 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
802 Local/Makefile (with some defaults set). These are used in built scripts
803 such as exicyclog, but they have never been used in the exim_install
804 script (though there are many overriding facilities there). I have
805 arranged that the exim_install script now takes note of these two
808 PH/35 Installed configuration files for Dragonfly.
810 PH/36 When a locally submitted message by a trusted user did not contain a
811 From: header, and the sender address was obtained from -f or from an SMTP
812 MAIL command, and the trusted user did not use -F to supply a sender
813 name, $originator_name was incorrectly used when constructing a From:
814 header. Furthermore, $originator_name was used for submission mode
815 messages from external hosts without From: headers in a similar way,
816 which is clearly wrong.
818 PH/37 Added control=suppress_local_fixups.
820 PH/38 When log_selector = +received_sender was set, and the addition of the
821 sender made the log line's construction buffer exactly full, or one byte
822 less than full, an overflow happened when the terminating "\n" was
825 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
826 when the result of a list match is failure because a DNS lookup failed.
828 PH/40 RM_COMMAND is now used in the building process.
830 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
831 the "build-* directories that it finds.
833 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
834 address in a domain literal was a prefix of an interface address.
836 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
837 when verifying a sender address, unless rewrite_headers is false.
839 PH/44 Wrote a long comment about why errors_to addresses are verified as
840 recipients, not senders.
842 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
843 the ratelimit ACL was added.
845 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
847 PH/46 Added extra information about PostgreSQL errors to the error string.
849 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
850 avoiding the need to install special zones in a real server. This is
851 backwards compatible; if it can't find the fake resolver, it drops back.
852 Thus, both old and new test suites can be run.
854 TF/02 Added util/ratelimit.pl
856 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
857 clock is set back in time.
859 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
860 Candler <B.Candler@pobox.com>.
862 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
863 OK for addresses that are the result of redirections.
865 PH/48 A number of further additions for the benefit of the new test suite,
866 including a fake gethostbyname() that interfaces to the fake DNS resolver
869 TF/06 The fix for widen_domains has also been applied to qualify_single and
870 search_parents which are the other dnslookup options that can cause
873 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
876 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
878 TF/07 Exim produced the error message "an SRV record indicated no SMTP
879 service" if it encountered an MX record with an empty target hostname.
880 The message is now "an MX or SRV record indicated no SMTP service".
882 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
883 if the DNS of the sending site is misconfigured. This is quite a
884 common situation. This change restores the behaviour of treating a
885 helo verification defer as a failure.
887 PH/51 If self=fail was set on a router, the bounce message did not include the
888 actual error message.
894 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
896 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
897 command itself ended in error, the underlying message about the transport
898 filter timeout was being overwritten with the pipe command error. Now the
899 underlying error message should be appended to the second error message.
901 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
903 PH/02 Reduce the amount of output that "make" produces by default. Full output
904 can still be requested.
906 PH/03 The warning log line about a condition test deferring for a "warn" verb
907 was being output only once per connection, rather than after each
908 occurrence (because it was using the same function as for successful
909 "warn" verbs). This seems wrong, so I have changed it.
911 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
912 it should not have, which might have caused a crash in the right
913 circumstances, but probably never did.
915 PH/04 Installed a modified version of Tony Finch's patch to make submission
916 mode fix the return path as well as the Sender: header line, and to
917 add a /name= option so that you can make the user's friendly name appear
920 TF/03 Added the control = fakedefer ACL modifier.
922 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
923 Mark Lowes for thorough testing.
925 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
927 TK/03 Merged latest SRS patch from Miles Wilton.
929 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
930 with the definition in sysexits.h (which is #included earlier).
931 Fortunately, Exim does not actually use EX_OK. The code used to try to
932 preserve the sysexits.h value, by assumimg that macro definitions were
933 scanned for macro replacements. I have been disabused of this notion,
934 so now the code just undefines EX_OK before #including unistd.h.
936 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
937 in the smtp transport. When a block could not be written in a single
938 write() function, the timeout was being re-applied to each part-write.
939 This seems wrong - if the receiver was accepting one byte at a time it
940 would take for ever. The timeout is now adjusted when this happens. It
941 doesn't have to be particularly precise.
943 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
944 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
946 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
947 without a domain if the check to <postmaster@domain> fails.
949 SC/01 Eximstats: added -xls and the ability to specify output files
950 (patch written by Frank Heydlauf).
952 SC/02 Eximstats: use FileHandles for outputing results.
954 SC/03 Eximstats: allow any combination of xls, txt, and html output.
956 SC/04 Eximstats: fixed display of large numbers with -nvr option
958 SC/05 Eximstats: fixed merging of reports with empty tables.
960 SC/06 Eximstats: added the -include_original_destination flag
962 SC/07 Eximstats: removed tabs and trailing whitespace.
964 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
966 TK/06 MBOX spool code: Add real "From " MBOX separator line
967 so the .eml file is really in mbox format (even though
968 most programs do not really care). Patch from Alex Miller.
970 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
971 The latter is generated from $received_to and is only set if the
972 message has one envelope recipient. SA can use these headers,
973 obviously out-of-the-box. Patch from Alex Miller.
975 PH/08 The ${def test on a variable was returning false if the variable's
976 value was "0", contrary to what the specification has always said!
977 The result should be true unless the variable is empty.
979 PH/09 The syntax error of a character other than { following "${if
980 def:variable_name" (after optional whitespace) was not being diagnosed.
981 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
982 accidental colon was present, for example, could give incorrect results.
984 PH/10 Tidied the code in a number of places where the st_size field of a stat()
985 result is used (not including appendfile, where other changes are about
988 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
989 This involved changing a lot of size variables from int to off_t. It
990 should work with maildirs and everything.
992 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
993 spamd dying while we are connected to it.
995 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
998 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
999 function for BDB 4.3.
1001 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1003 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1004 the sources for PCRE have been re-arranged and more files are now
1007 PH/15 The code I had for printing potentially long long variables in PH/11
1008 above was not the best (it lost precision). The length of off_t variables
1009 is now inspected at build time, and an appropriate printing format (%ld
1010 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1011 to be "long long int" or "long int". This is needed for the internal
1012 formatting function string_vformat().
1014 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1015 the configuration file to be ":syslog", then the script "guesses" where
1016 the logs files are, rather than using the compiled in default. In our
1017 case the guess is not the same as the compiled default, so the script
1018 suddenly stopped working when I started to use syslog. The patch checks
1019 to see if log_file_path is "". If so, it attempts to read it from exim
1020 with no configuration file to get the compiled in version, before it
1021 falls back to the previous guessing code."
1023 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1024 implementing BATV in an Exim configuration. See NewStuff for the gory
1027 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1028 Makefile that are specific to HP-UX.
1030 PH/18 If the "use_postmaster" option was set for a recipient callout together
1031 with the "random" option, the postmaster address was used as the MAIL
1032 FROM address for the random test, but not for the subsequent recipient
1033 test. It is now used for both.
1035 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1036 patch removes a few documentation additions to RFC 3028, because the
1037 latest draft now contains them. It adds the new en;ascii-case comparator
1038 and a new error check for 8bit text in MIME parts. Comparator and
1039 require names are now matched exactly. I enabled the subaddress
1040 extension, but it is not well tested yet (read: it works for me)."
1042 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1043 rework some of the code of TK/09 above to avoid the hardwired use of
1044 "%lld" and "long long". Replaced the call to snprintf() with a call to
1047 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1048 records point to non-existent hosts", "retry timeout exceeded", and
1049 "retry time not reached for any host after a long failure period".
1051 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1052 experimental DomainKeys support:
1054 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1055 (2) On an error such as an illegally used "control", the wrong name for
1056 the control was given.
1058 These problems did NOT occur unless DomainKeys support was compiled.
1060 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1062 PH/24 Added ${if match_ip condition.
1064 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1065 databases so that it will be absolutely obvious if a crash occurs in the
1066 DB library. This is a regular occurrence (often caused by mis-matched
1069 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1070 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1071 picked up on a user's system that detects such things. There doesn't seem
1072 to be a gcc warning option for this - only an attribute that has to be
1073 put on the function's prototype. It seems that in Fedora Core 4 they have
1074 set this on a number of new functions. No doubt there will be more in due
1077 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1078 specify a transport. However, if an address that was verified by such a
1079 router was the subject of a callout, Exim crashed because it tried to
1080 read the rcpt_include_affixes from the non-existent transport. Now it
1081 just assumes that the setting of that option is false. This bug was
1082 introduced by 4.51/PH/31.
1084 PH/28 Changed -d+all to exclude +memory, because that information is very
1085 rarely of interest, but it makes the output a lot bigger. People tend to
1086 do -d+all out of habit.
1088 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1089 code in os-type was giving problems when libc.so lives in lib64, like on
1092 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1093 aren't the modern standard, and it seems that some systems' include files
1094 don't always have them. Exim was already checking for some of the newer
1095 ones like T_AAAA, and defining it itself. I've added checks for all the
1096 record types that Exim uses.
1098 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1099 not automatically generating a new one, as it is supposed to. This
1100 prevented TLS from working. If the file did exist, but contained invalid
1101 data, a new version was generated, as expected. It was only the case of a
1102 non-existent file that was broken.
1104 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1105 with a change in libdomainkeys > 0.64.
1107 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1108 from DNS. If the selector record carries the flag, it now has
1109 precedence over the domain-wide flag.
1111 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1113 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1114 the use of a port name that isn't defined in /etc/services) occurred, the
1115 message was deferred as in a normal delivery, and thus remained on the
1116 spool, instead of being failed because of the mua_wrapper setting. This
1117 is now fixed, and I tidied up some of the mua_wrapper messages at the
1120 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1121 the messages in a hash of arrays rather than using individual hashes.
1122 This is a bit cleaner and results in dramatic memory savings, albeit
1123 at a slight CPU cost.
1125 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1126 as requested by Marc Sherman.
1128 SC/10 Eximstats: added histograms for user specified patterns as requested
1131 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1133 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1134 fopen() in the content-scanning modules that did not already have it.
1140 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1141 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1143 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1145 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1147 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1148 to test Sieve filters that use "vacation".
1150 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1151 that changes the way the GnuTLS parameters are stored in the cache file.
1152 The new format can be generated externally. For backward compatibility,
1153 if the data in the cache doesn't make sense, Exim assumes it has read an
1154 old-format file, and it generates new data and writes a new file. This
1155 means that you can't go back to an older release without removing the
1158 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1159 work if there are any delivery delays because "one_time" forces the
1160 parent to be marked "delivered", so its unseen clone is never tried
1161 again. For this reason, Exim now forbids the simultaneous setting of
1164 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1165 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1166 entry if you want to know the details. The fix, however, neglected to
1167 consider the case when local delivery batching is involved. The test for
1168 "previously delivered" was not happening when checking to see if an
1169 address could be batched with a previous (undelivered) one; under
1170 certain circumstances this could lead to multiple deliveries to the same
1173 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1174 in its include files, and this causes problems building Exim.
1176 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1177 header_syntax) but Exim was just ignoring anything given after a slash.
1178 In particular, this caused confusion with an attempt to use "verify =
1179 reverse_host_lookup/defer_ok". An error is now given when options are
1180 supplied for verify items that do not have them. (Maybe reverse_host_
1181 lookup should have a defer_ok option, but that's a different point.)
1183 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1184 defined by RFC 821) to 2048, because there were problems with some AUTH
1185 commands, and RFC 1869 says the size should be increased for extended
1186 SMTP commands that take arguments.
1188 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1191 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1192 "unknown" error; now it says that the functionality isn't in the binary.
1194 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1195 an address' error message when a string expansion fails (syntax or
1196 whatever). Otherwise the password may appear in the log. Following change
1197 PH/42 below, there is no longer a chance of it appearing in a bounce
1200 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1202 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1203 of its addresses was ever tried. (Bugzilla bug #2.)
1205 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1206 the result incorrectly in the debug output. (It correctly added a newline
1207 to what was transported.)
1209 TF/01 Added $received_time.
1211 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1212 commented out examples of how to interface to a virus scanner and to
1213 SpamAssassin. Also added commented examples of av_scanner and
1214 spamd_address settings.
1216 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1217 and controls are allowed in which ACLs. There were a couple of minor
1218 errors. Some of the entries in the conditions table (which is a table of
1219 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1220 negation of where the condition IS allowed.
1222 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1224 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1225 header file does not have a version number, so I've had to invent a new
1226 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1227 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1228 radiusclient), but it was contributed by a Radius user.
1230 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1231 files or directories, for OpenSSL.
1233 PH/19 When an Exim process that is running as root has to create an Exim log
1234 file, it does so in a subprocess that runs as exim:exim so as to get the
1235 ownership right at creation (otherwise, other Exim processes might see
1236 the file with the wrong ownership). There was no test for failure of this
1237 fork() call, which would lead to the process getting stuck as it waited
1238 for a non-existent subprocess. Forks do occasionally fail when resources
1239 run out. I reviewed all the other calls to fork(); they all seem to check
1242 PH/20 When checking for unexpected SMTP input at connect time (before writing
1243 the banner), Exim was not dealing correctly with a non-positive return
1244 from the read() function. If the client had disconnected by this time,
1245 the result was a log entry for a synchronization error with an empty
1246 string after "input=" when read() returned zero. If read() returned -1
1247 (an event I could not check), uninitialized data bytes were printed.
1248 There were reports of junk text (parts of files, etc) appearing after
1251 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1253 PH/22 Added support for macro redefinition, and (re)definition in between
1254 driver and ACL definitions.
1256 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1257 forgetting to use the resulting value; it was using the unexpanded value.
1259 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1260 hadn't been configured. The fix is from Juergen Kreileder, who
1261 understands it better than I do:
1263 "Here's what I see happening with three configured cyrus_sasl
1264 authenticators configured (plain, login, cram-md5):
1266 On startup auth_cyrus_sasl_init() gets called for each of these.
1267 This means three calls to sasl_listmech() without a specified mech_list.
1268 => SASL tests which mechs of all available mechs actually work
1269 => three warnings about OTP not working
1270 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1272 With the patch, sasl_listmech() also gets called three times. But now
1273 SASL's mech_list option is set to the server_mech specified in the the
1274 authenticator. Or in other words, the answer from sasl_listmech()
1275 gets limited to just the mech you're testing for (which is different
1277 => the return list contains just 'plain' or 'login', 'cram-md5' or
1278 nothing depending on the value of ob->server_mech.
1280 I've just tested the patch: Authentication still works fine,
1281 unavailable mechs specified in the exim configuration are still
1282 caught, and the auth.log warnings about OTP are gone."
1284 PH/25 When debugging is enabled, the contents of the command line are added
1285 to the debugging output, even when log_selector=+arguments is not
1288 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1289 answer is "GNU", and only if the return is "GNU/something" is the answer
1292 PH/27 $acl_verify_message is now set immediately after the failure of a
1293 verification in an ACL, and so is available in subsequent modifiers. In
1294 particular, the message can be preserved by coding like this:
1296 warn !verify = sender
1297 set acl_m0 = $acl_verify_message
1299 Previously, $acl_verify_message was set only while expanding "message"
1300 and "log_message" when a very denied access.
1302 PH/28 Modified OS/os.c-Linux with
1304 -#ifndef OS_LOAD_AVERAGE
1305 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1307 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1308 nomenclature these days.)
1310 PH/29 Installed patch from the Sieve maintainer that adds the options
1311 sieve_useraddress and sieve_subaddress to the redirect router.
1313 PH/30 In these circumstances:
1314 . Two addresses routed to the same list of hosts;
1315 . First host does not offer TLS;
1316 . First host accepts first address;
1317 . First host gives temporary error to second address;
1318 . Second host offers TLS and a TLS session is established;
1319 . Second host accepts second address.
1320 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1321 and peerdn, if requested) that were in fact used only for the second
1324 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1325 attention to any local part prefix or suffix that was matched by the
1326 router that accepted the address. It now behaves in the same way as it
1327 does for delivery: the affixes are removed from the local part unless
1328 rcpt_include_affixes is set on the transport.
1330 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1331 timeout during the DATA phase of an incoming message.
1333 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1334 applied a patch sanctioned by the Sieve maintainer.
1336 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1337 the uid or gid is negative. A case of a negative gid caused this to be
1338 noticed. The fix allows for either to be negative.
1340 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1341 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1344 PH/36 The change PH/12 above was broken. Fixed it.
1346 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1347 the grounds that routing the same address twice would always produce the
1348 same answer. This might have been true once, but it is certainly no
1349 longer true now. Routing a child address may depend on the previous
1350 routing that produced that child. Some complicated redirection strategies
1351 went wrong when messages had multiple recipients, and made Exim's
1352 behaviour dependent on the order in which the addresses were given.
1354 I have moved the duplicate checking until after the routing is complete.
1355 Exim scans the addresses that are assigned to local and remote
1356 transports, and removes any duplicates. This means that more work will be
1357 done, as duplicates will always all be routed, but duplicates are
1358 presumably rare, so I don't expect this is of any significance.
1360 For deliveries to pipes, files, and autoreplies, the duplicate checking
1361 still happens during the routing process, since they are not going to be
1364 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1365 It corrects a timeout issue with spamd. This is Ian's comment: "The
1366 background is that sometimes spamd either never reads data from a
1367 connection it has accepted, or it never writes response data. The exiscan
1368 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1369 blindly assumes that writes won't block so it may never time out."
1371 PH/39 Allow G after quota size as well as K and M.
1373 PH/40 The value set for $authenticated_id in an authenticator may not contain
1374 binary zeroes or newlines because the value is written to log lines and
1375 to spool files. There was no check on this. Now the value is run through
1376 the string_printing() function so that such characters are converted to
1377 printable escape sequences.
1379 PH/41 $message_linecount is a new variable that contains the total number of
1380 lines in the message. Compare $body_linecount, which is the count for the
1383 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1384 bounce and delay warning messages, except in certain special cases, which
1387 (a) An SMTP error message from a remote host;
1388 (b) A message specified in a :fail: redirection;
1389 (c) A message specified in a "fail" command in a system filter;
1390 (d) A message specified in a FAIL return from the queryprogram router;
1391 (e) A message specified by the cannot_route_message router option.
1393 In these cases only, Exim does include the error details in bounce and
1394 warning messages. There are also a few cases where bland messages such
1395 as "unrouteable address" or "local delivery error" are given.
1397 PH/43 $value is now also set for the "else" part of a ${run expansion.
1399 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1400 being worked on, but at least Exim now implements the latest version to
1403 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1404 process to complete was treated as a delivery failure, a timeout while
1405 writing the message to the pipe was logged, but erroneously treated as a
1406 successful delivery. Such timeouts include transport filter timeouts. For
1407 consistency with the overall process timeout, these timeouts are now
1408 treated as errors, giving rise to delivery failures by default. However,
1409 there is now a new Boolean option for the pipe transport called
1410 timeout_defer, which, if set TRUE, converts the failures into defers for
1411 both kinds of timeout. A transport filter timeout is now identified in
1414 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1415 systems where "make" and "gmake" are different, calling "gmake" at top
1416 level broke things. I've arranged for the value of $(MAKE) to be passed
1417 from the Makefile to this script so that it can call the same version of
1421 A note about Exim versions 4.44 and 4.50
1422 ----------------------------------------
1424 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1425 changes of various kinds. As a consequence, a big documentation update was
1426 needed. This delayed the release for rather longer than seemed good, especially
1427 in the light of a couple of (minor) security issues. Therefore, the changes
1428 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1429 release. So 4.44 and 4.50 are in effect two different branches that both start
1432 I have left the 4.50 change log unchanged; it contains all the changes since
1433 4.43. The change log for 4.44 is below; many of its items are identical to
1434 those for 4.50. This seems to be the most sensible way to preserve the
1435 historical information.
1441 1. Minor wording change to the doc/README.SIEVE file.
1443 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1444 computation of the current number of files was incorrect.
1446 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1447 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1450 4. Give more explanation in the error message when the command for a transport
1451 filter fails to execute.
1453 5. There are several places where Exim runs a non-Exim command in a
1454 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1455 was being done only for the command run by the queryprogram router. It is
1456 now done for all such subprocesses. The other cases are: ${run, transport
1457 filters, and the commands run by the lmtp and pipe transports.
1459 6. Added CONFIGURE_GROUP build-time option.
1461 7. Some older OS have a limit of 256 on the maximum number of file
1462 descriptors. Exim was using setrlimit() to set 1000 as a large value
1463 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1464 systems. I've change it so that if it can't get 1000, it tries for 256.
1466 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1467 was an oversight, and furthermore, ever since the addition of extra
1468 controls (e.g. 4.43/32), the checks on when to allow different forms of
1469 "control" were broken. There should now be diagnostics for all cases when a
1470 control that does not make sense is encountered.
1472 9. Added the /retain_sender option to "control=submission".
1474 10. $recipients is now available in the predata ACL (oversight).
1476 11. Tidy the search cache before the fork to do a delivery from a message
1477 received from the command line. Otherwise the child will trigger a lookup
1478 failure and thereby defer the delivery if it tries to use (for example) a
1479 cached ldap connection that the parent has called unbind on.
1481 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1482 of $address_data from the recipient verification was clobbered by the
1483 sender verification.
1485 13. The value of address_data from a sender verification is now available in
1486 $sender_address_data in subsequent conditions in the ACL statement.
1488 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1490 15. Added a new option "connect=<time>" to callout options, to set a different
1493 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1494 was its contents. (It was OK if the option was not defined at all.)
1496 17. A "Completed" log line is now written for messages that are removed from
1497 the spool by the -Mrm option.
1499 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1500 information about exactly what failed.
1502 19. Added -dd to debug only the daemon process.
1504 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1505 handles timeouts, both on the server side and network timeouts. Renamed the
1506 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1508 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1509 It is now set to "smtps".
1511 22. $host_address is now set to the target address during the checking of
1512 ignore_target_hosts.
1514 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1515 being passed; this would have caused $sender_host_name to have been used if
1516 matching the list had actually called for a host name (not very likely,
1517 since this list is usually IP addresses). A host name is now passed as
1520 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1521 code that specifies a non-restarting handler (typically sigaction() in
1522 modern systems) in an attempt to fix a rare and obscure crash bug.
1524 25. Narrowed the window for a race in the daemon that could cause it to ignore
1525 SIGCHLD signals. This is not a major problem, because they are used only to
1526 wake it up if nothing else does.
1528 26. A malformed maildirsize file could cause Exim to calculate negative values
1529 for the mailbox size or file count. Odd effects could occur as a result.
1530 The maildirsize information is now recalculated if the size or filecount
1533 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1534 support for a long time. Removed HAVE_SYS_VFS_H.
1536 28. Installed the latest version of exipick from John Jetmore.
1538 29. In an address list, if the pattern was not a regular expression, an empty
1539 subject address (from a bounce message) matched only if the pattern was an
1540 empty string. Non-empty patterns were not even tested. This was the wrong
1541 because it is perfectly reasonable to use an empty address as part of a
1542 database query. An empty address is now tested by patterns that are
1543 lookups. However, all the other forms of pattern expect the subject to
1544 contain a local part and a domain, and therefore, for them, an empty
1545 address still always fails if the pattern is not itself empty.
1547 30. Exim went into a mad DNS loop when attempting to do a callout where the
1548 host was specified on an smtp transport, and looking it up yielded more
1549 than one IP address.
1551 31. Re-factored the code for checking spool and log partition space into a
1552 function that finds that data and another that does the check. The former
1553 is then used to implement four new variables: $spool_space, $log_space,
1554 $spool_inodes, and $log_inodes.
1556 32. The RFC2047 encoding function was originally intended for short strings
1557 such as real names; it was not keeping to the 75-character limit for
1558 encoded words that the RFC imposes. It now respects the limit, and
1559 generates multiple encoded words if necessary. To be on the safe side, I
1560 have increased the buffer size for the ${rfc2047: expansion operator from
1563 33. It is now permitted to omit both strings after an "if" condition; if the
1564 condition is true, the result is "true". As before, when the second string
1565 is omitted, a false condition yields an empty string. This makes it less
1566 cumbersome to write custom ACL and router conditions.
1568 34. Failure to deliver a bounce message always caused it to be frozen, even if
1569 there was an errors_to setting on the router. The errors_to setting is now
1572 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1573 canonical form (fully expanded) before being placed in
1574 $sender_host_address.
1576 36. The table in the code that translates DNS record types into text (T_A to
1577 "A" for instance) was missing entries for NS and CNAME. It is just possible
1578 that this could have caused confusion if both these types were looked up
1579 for the same domain, because the text type is used as part of Exim's
1580 per-process caching. But the chance of anyone hitting this buglet seems
1583 37. The dnsdb lookup has been extended in a number of ways.
1585 (1) There is a new type, "zns", which walks up the domain tree until it
1586 finds some nameserver records. It should be used with care.
1588 (2) There is a new type, "mxh", which is like "mx" except that it returns
1589 just the host names, not the priorities.
1591 (3) It is now possible to give a list of domains (or IP addresses) to be
1592 looked up. The behaviour when one of the lookups defers can be
1593 controlled by a keyword.
1595 (4) It is now possible to specify the separator character for use when
1596 multiple records are returned.
1598 38. The dnslists ACL condition has been extended: it is now possible to supply
1599 a list of IP addresses and/or domains to be looked up in a particular DNS
1602 39. Added log_selector=+queue_time_overall.
1604 40. When running the queue in the test harness, wait just a tad after forking a
1605 delivery process, to get repeatability of debugging output.
1607 41. Include certificate and key file names in error message when GnuTLS fails
1608 to set them up, because the GnuTLS error message doesn't include the name
1609 of the failing file when there is a problem reading it.
1611 42. Allow both -bf and -bF in the same test run.
1613 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1615 44. The "Exiscan patch" is now merged into the mainline Exim source.
1617 45. Sometimes the final signoff response after QUIT could fail to get
1618 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1619 < 0 before doing a fflush(). This bug looks as though it goes back to the
1620 introduction of TLS in release 3.20, but "sometimes" must have been rare
1621 because the tests only now provoked it.
1623 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1624 (this can affect the format of dates).
1626 47. exim_tidydb, when checking for the continued existence of a message for
1627 which it has found a message-specific retry record, was not finding
1628 messages that were in split spool directories. Consequently, it was
1629 deleting retry records that should have stayed in existence.
1631 48. Steve fixed some bugs in eximstats.
1633 49. The SPA authentication driver was not abandoning authentication and moving
1634 on to the next authenticator when an expansion was forced to fail,
1635 contradicting the general specification for all authenticators. Instead it
1636 was generating a temporary error. It now behaves as specified.
1638 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1639 (the order specifies the preference for clients). The order is now AES256,
1640 AES128, 3DES, ARCFOUR128.
1642 51. Small patch to Sieve code - explicitly set From: when generating an
1645 52. Exim crashed if a remote delivery caused a very long error message to be
1646 recorded - for instance if somebody sent an entire SpamAssassin report back
1647 as a large number of 550 error lines. This bug was coincidentally fixed by
1648 increasing the size of one of Exim's internal buffers (big_buffer) that
1649 happened as part of the Exiscan merge. However, to be on the safe side, I
1650 have made the code more robust (and fixed the comments that describe what
1653 53. Now that there can be additional text after "Completed" in log lines (if
1654 the queue_time_overall log selector is set), a one-byte patch to exigrep
1655 was needed to allow it to recognize "Completed" as not the last thing in
1658 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1659 patch that reportedly fixes this has been added. I am not expert enough to
1660 create a test for it. This is what the patch creator wrote:
1662 "I found a little strange behaviour of ldap code when working with
1663 Windows 2003 AD Domain, where users was placed in more than one
1664 Organization Units. When I tried to give exim partial DN, the exit code
1665 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1666 But simultaneously result of request was absolutely normal ldap result,
1667 so I produce this patch..."
1669 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1670 REFERENCE, so I have modified the code to exclude the patch when that macro
1673 55. Some experimental protocols are using DNS PTR records for new purposes. The
1674 keys for these records are domain names, not reversed IP addresses. The
1675 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1676 leaves it alone. Component reversal etc. now happens only for IP addresses.
1679 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1681 57. Double the size of the debug message buffer (to 2048) so that more of very
1682 long debug lines gets shown.
1684 58. The exicyclog utility now does better if the number of log files to keep
1685 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1687 59. Two changes related to the smtp_active_hostname option:
1689 (1) $smtp_active_hostname is now available as a variable.
1690 (2) The default for smtp_banner uses $smtp_active_hostname instead
1691 of $primary_hostname.
1693 60. The host_aton() function is supposed to be passed a string that is known
1694 to be a valid IP address. However, in the case of IPv6 addresses, it was
1695 not checking this. This is a hostage to fortune. Exim now panics and dies
1696 if the condition is not met. A case was found where this could be provoked
1697 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1698 components; fortuitously, this particular loophole had already been fixed
1699 by change 4.50/55 above.
1701 If there are any other similar loopholes, the new check in host_aton()
1702 itself should stop them being exploited. The report I received stated that
1703 data on the command line could provoke the exploit when Exim was running as
1704 exim, but did not say which command line option was involved. All I could
1705 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1706 running as the user.
1709 61. There was a buffer overflow vulnerability in the SPA authentication code
1710 (which came originally from the Samba project). I have added a test to the
1711 spa_base64_to_bits() function which I hope fixes it.
1714 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1715 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1717 63. The daemon start-up calls getloadavg() while still root for those OS that
1718 need the first call to be done as root, but it missed one case: when
1719 deliver_queue_load_max is set with deliver_drop_privilege. This is
1720 necessary for the benefit of the queue runner, because there is no re-exec
1721 when deliver_drop_privilege is set.
1723 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1724 This has been fixed.
1726 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1727 was in use, was not putting the data itself into the right store pool;
1728 consequently, it could be overwritten for a subsequent message in the same
1729 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1732 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1734 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1735 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1736 the function were treating the return as a boolean value, which happened to
1737 work because 0=false and not-0=true, but is not correct code.
1739 68. The host_aton() function was not handling scoped IPv6 addresses (those
1740 with, for example, "%eth0" on the end) correctly.
1742 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1743 negated items (that is, ~something) in unsigned ints. Some compilers
1744 apparently mutter when there is no cast.
1746 70. If an address verification called from an ACL failed, and did not produce a
1747 user-specific message (i.e. there was only a "system" message), nothing was
1748 put in $acl_verify_message. In this situation, it now puts the system
1751 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1752 session; change 4.31/43 added the unwanted input to the log line - except
1753 that it did not do this in the start of session case. It now does.
1755 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1756 This is wrong and can cause the other end to generate a synchronization
1757 error if it is another Exim or anything else that does the synchronization
1758 check. A QUIT command is no longer sent after a timeout.
1760 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1761 during host lookups.
1763 74. The defer_ok option of callout verification was not working if it was used
1764 when verifying addresses in header lines, that is, for this case:
1766 verify = header_sender/callout=defer_ok
1768 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1769 those file descriptors could be used for SMTP connections. If anything
1770 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1771 could be sent to the SMTP client, causing chaos. The daemon now opens
1772 stdin, stdout, and stderr to /dev/null when it puts itself into the
1775 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1776 log by default. The user can override this with suitable Perl magic.
1778 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1779 the log message when discard triggers, was not working for the DATA ACL or
1780 for the non-SMTP ACL.
1782 78. Error message wording change in sieve.c.
1784 79. If smtp_accept_max_per_host was set, the number of connections could be
1785 restricted to fewer than expected, because the daemon was trying to set up
1786 a new connection before checking whether the processes handling previous
1787 connections had finished. The check for completed processes is now done
1788 earlier. On busy systems, this bug wouldn't be noticed because something
1789 else would have woken the daemon, and it would have reaped the completed
1792 80. If a message was submitted locally by a user whose login name contained one
1793 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1794 It caused a spool format error. I have fixed the spool reading code. A
1795 related problem was that the "from" clause in the Received: line became
1796 illegal because of the space(s). It is now covered by ${quote_local_part.
1798 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1801 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1803 83. Patch from Sieve maintainer to fix unterminated string problem in
1804 "vacation" handling.
1806 84. Some minor changes to the Linux configuration files to help with other
1807 OS variants using glibc.
1809 85. One more patch for Sieve to update vacation handling to latest spec.
1812 ----------------------------------------------------
1813 See the note above about the 4.44 and 4.50 releases.
1814 ----------------------------------------------------
1820 1. Change 4.43/35 introduced a bug that caused file counts to be
1821 incorrectly computed when quota_filecount was set in an appendfile
1824 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1825 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1828 3. Give more explanation in the error message when the command for a transport
1829 filter fails to execute.
1831 4. There are several places where Exim runs a non-Exim command in a
1832 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1833 was being done only for the command run by the queryprogram router. It is
1834 now done for all such subprocesses. The other cases are: ${run, transport
1835 filters, and the commands run by the lmtp and pipe transports.
1837 5. Some older OS have a limit of 256 on the maximum number of file
1838 descriptors. Exim was using setrlimit() to set 1000 as a large value
1839 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1840 systems. I've change it so that if it can't get 1000, it tries for 256.
1842 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1843 was an oversight, and furthermore, ever since the addition of extra
1844 controls (e.g. 4.43/32), the checks on when to allow different forms of
1845 "control" were broken. There should now be diagnostics for all cases when a
1846 control that does not make sense is encountered.
1848 7. $recipients is now available in the predata ACL (oversight).
1850 8. Tidy the search cache before the fork to do a delivery from a message
1851 received from the command line. Otherwise the child will trigger a lookup
1852 failure and thereby defer the delivery if it tries to use (for example) a
1853 cached ldap connection that the parent has called unbind on.
1855 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1856 of $address_data from the recipient verification was clobbered by the
1857 sender verification.
1859 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1860 was its contents. (It was OK if the option was not defined at all.)
1862 11. A "Completed" log line is now written for messages that are removed from
1863 the spool by the -Mrm option.
1865 12. $host_address is now set to the target address during the checking of
1866 ignore_target_hosts.
1868 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1869 being passed; this would have caused $sender_host_name to have been used if
1870 matching the list had actually called for a host name (not very likely,
1871 since this list is usually IP addresses). A host name is now passed as
1874 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1875 code that specifies a non-restarting handler (typically sigaction() in
1876 modern systems) in an attempt to fix a rare and obscure crash bug.
1878 15. Narrowed the window for a race in the daemon that could cause it to ignore
1879 SIGCHLD signals. This is not a major problem, because they are used only to
1880 wake it up if nothing else does.
1882 16. A malformed maildirsize file could cause Exim to calculate negative values
1883 for the mailbox size or file count. Odd effects could occur as a result.
1884 The maildirsize information is now recalculated if the size or filecount
1887 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1888 support for a long time. Removed HAVE_SYS_VFS_H.
1890 18. Updated exipick to current release from John Jetmore.
1892 19. Allow an empty sender to be matched against a lookup in an address list.
1893 Previously the only cases considered were a regular expression, or an
1896 20. Exim went into a mad DNS lookup loop when doing a callout where the
1897 host was specified on the transport, if the DNS lookup yielded more than
1900 21. The RFC2047 encoding function was originally intended for short strings
1901 such as real names; it was not keeping to the 75-character limit for
1902 encoded words that the RFC imposes. It now respects the limit, and
1903 generates multiple encoded words if necessary. To be on the safe side, I
1904 have increased the buffer size for the ${rfc2047: expansion operator from
1907 22. Failure to deliver a bounce message always caused it to be frozen, even if
1908 there was an errors_to setting on the router. The errors_to setting is now
1911 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1912 canonical form (fully expanded) before being placed in
1913 $sender_host_address.
1915 24. Updated eximstats to version 1.33
1917 25. Include certificate and key file names in error message when GnuTLS fails
1918 to set them up, because the GnuTLS error message doesn't include the name
1919 of the failing file when there is a problem reading it.
1921 26. Expand error message when OpenSSL has problems setting up cert/key files.
1924 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1925 (this can affect the format of dates).
1927 28. exim_tidydb, when checking for the continued existence of a message for
1928 which it has found a message-specific retry record, was not finding
1929 messages that were in split spool directories. Consequently, it was
1930 deleting retry records that should have stayed in existence.
1932 29. eximstats updated to version 1.35
1933 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1934 1.35 - bugfix such that pie charts by volume are generated correctly
1936 30. The SPA authentication driver was not abandoning authentication and moving
1937 on to the next authenticator when an expansion was forced to fail,
1938 contradicting the general specification for all authenticators. Instead it
1939 was generating a temporary error. It now behaves as specified.
1941 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1942 (the order specifies the preference for clients). The order is now AES256,
1943 AES128, 3DES, ARCFOUR128.
1945 31. Small patch to Sieve code - explicitly set From: when generating an
1948 32. Exim crashed if a remote delivery caused a very long error message to be
1949 recorded - for instance if somebody sent an entire SpamAssassin report back
1950 as a large number of 550 error lines. This bug was coincidentally fixed by
1951 increasing the size of one of Exim's internal buffers (big_buffer) that
1952 happened as part of the Exiscan merge. However, to be on the safe side, I
1953 have made the code more robust (and fixed the comments that describe what
1956 33. Some experimental protocols are using DNS PTR records for new purposes. The
1957 keys for these records are domain names, not reversed IP addresses. The
1958 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1959 leaves it alone. Component reversal etc. now happens only for IP addresses.
1962 34. The host_aton() function is supposed to be passed a string that is known
1963 to be a valid IP address. However, in the case of IPv6 addresses, it was
1964 not checking this. This is a hostage to fortune. Exim now panics and dies
1965 if the condition is not met. A case was found where this could be provoked
1966 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1967 components; fortuitously, this particular loophole had already been fixed
1968 by change 4.50/55 or 4.44/33 above.
1970 If there are any other similar loopholes, the new check in host_aton()
1971 itself should stop them being exploited. The report I received stated that
1972 data on the command line could provoke the exploit when Exim was running as
1973 exim, but did not say which command line option was involved. All I could
1974 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1975 running as the user.
1978 35. There was a buffer overflow vulnerability in the SPA authentication code
1979 (which came originally from the Samba project). I have added a test to the
1980 spa_base64_to_bits() function which I hope fixes it.
1983 36. The daemon start-up calls getloadavg() while still root for those OS that
1984 need the first call to be done as root, but it missed one case: when
1985 deliver_queue_load_max is set with deliver_drop_privilege. This is
1986 necessary for the benefit of the queue runner, because there is no re-exec
1987 when deliver_drop_privilege is set.
1989 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1990 was in use, was not putting the data itself into the right store pool;
1991 consequently, it could be overwritten for a subsequent message in the same
1992 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1995 38. Sometimes the final signoff response after QUIT could fail to get
1996 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1997 < 0 before doing a fflush(). This bug looks as though it goes back to the
1998 introduction of TLS in release 3.20, but "sometimes" must have been rare
1999 because the tests only now provoked it.
2005 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2006 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2007 Now it returns an integer. A number of calls were still expecting a T/F
2008 return. Fortuitously, in all cases, the tests worked in OK situations,
2009 which is the norm. However, things would have gone wrong on any write
2010 failures on the smtp file descriptor. This function is used when sending
2011 messages over SMTP and also when doing verify callouts.
2013 2. When Exim is called to do synchronous delivery of a locally submitted
2014 message (the -odf or -odi options), it no longer closes stderr before doing
2017 3. Implemented the mua_wrapper option.
2019 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2021 5. Implemented the functions header_remove(), header_testname(),
2022 header_add_at_position(), and receive_remove_recipient(), and exported them
2025 6. If an ACL "warn" statement specified the addition of headers, Exim already
2026 inserted X-ACL-Warn: at the start if there was no header name. However, it
2027 was not making this test for the second and subsequent header lines if
2028 there were newlines in the string. This meant that an invalid header could
2029 be inserted if Exim was badly configured.
2031 7. Allow an ACL "warn" statement to add header lines at the start or after all
2032 the Received: headers, as well as at the end.
2034 8. Added the rcpt_4xx retry error code.
2036 9. Added postmaster_mailfrom=xxx to callout verification option.
2038 10. Added mailfrom=xxxx to the callout verification option, for verify=
2041 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2042 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2044 12. Inserted some casts to stop certain compilers warning when using pointer
2045 differences as field lengths or precisions in printf-type calls (mostly
2046 affecting debugging statements).
2048 13. Added optional readline() support for -be (dynamically loaded).
2050 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2051 same clock tick as a message's arrival, so that its received time was the
2052 same as the "first fail" time on the retry record, and that message
2053 remained on the queue past the ultimate address timeout, every queue runner
2054 would try a delivery (because it was past the ultimate address timeout) but
2055 after another failure, the ultimate address timeout, which should have then
2056 bounced the address, did not kick in. This was a "< instead of <=" error;
2057 in most cases the first failure would have been in the next clock tick
2058 after the received time, and all would be well.
2060 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2061 being recognized when the domain list was tested by the match_domain
2062 condition in an expansion string.
2064 16. Added the ${str2b64: operator.
2066 17. Exim was always calling setrlimit() to set a large limit for the number of
2067 processes, without checking whether the existing limit was already
2068 adequate. (It did check for the limit on file descriptors.) Furthermore,
2069 errors from getrlimit() and setrlimit() were being ignored. Now they are
2070 logged to the main and panic logs, but Exim does carry on, to try to do its
2071 job under whatever limits there are.
2073 18. Imported PCRE 5.0.
2075 19. Trivial typo in log message " temporarily refused connection" (the leading
2078 20. If the log selector return_path_on_delivery was set and an address was
2079 redirected to /dev/null, the delivery process crashed because it assumed
2080 that a return path would always be set for a "successful" delivery. In this
2081 case, the whole delivery is bypassed as an optimization, and therefore no
2084 21. Internal re-arrangement: the function for sending a challenge and reading
2085 a response while authentication was assuming a zero-terminated challenge
2086 string. It's now changed to take a pointer and a length, to allow for
2087 binary data in such strings.
2089 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2091 23. Exim was not respecting finduser_retries when seeking the login of the
2092 uid under which it was called; it was always trying 10 times. (The default
2093 setting of finduser_retries is zero.) Also, it was sleeping after the final
2094 failure, which is pointless.
2096 24. Implemented tls_on_connect_ports.
2098 25. Implemented acl_smtp_predata.
2100 26. If the domain in control=submission is set empty, Exim assumes that the
2101 authenticated id is a complete email address when it generates From: or
2102 Sender: header lines.
2104 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2105 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2106 chown and chgrp in /bin and hostname in /usr/bin.
2108 28. Exim was keeping the "process log" file open after each use, just as it
2109 does for the main log. This opens the possibility of it remaining open for
2110 long periods when the USR1 signal hits a daemon. Occasional processlog
2111 errors were reported, that could have been caused by this. Anyway, it seems
2112 much more sensible not to leave this file open at all, so that is what now
2115 29. The long-running daemon process does not normally write to the log once it
2116 has entered its main loop, and it closes the log before doing so. This is
2117 so that log files can straightforwardly be renamed and moved. However,
2118 there are a couple of unusual error situations where the daemon does write
2119 log entries, and I had neglected to close the log afterwards.
2121 30. The text of an SMTP error response that was received during a remote
2122 delivery was being truncated at 512 bytes. This is too short for some of
2123 the long messages that one sometimes sees. I've increased the limit to
2126 31. It is now possible to make retry rules that apply only when a message has a
2127 specific sender, in particular, an empty sender.
2129 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2130 it possible to be selective about when SMTP synchronization is enforced.
2132 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2134 32. Implemented hosts_connection_nolog.
2136 33. Added an ACL for QUIT.
2138 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2141 35. Added mailbox_size and mailbox_filecount to appendfile.
2143 36. Added control = no_multiline_responses to ACLs.
2145 37. There was a bug in the logic of the code that waits for the clock to tick
2146 in the case where the clock went backwards by a substantial amount such
2147 that the microsecond fraction of "now" was more than the microsecond
2148 fraction of "then" (but the whole seconds number was less).
2150 38. Added support for the libradius Radius client library this is found on
2151 FreeBSD (previously only the radiusclient library was supported).
2157 1. When certain lookups returned multiple values in the form name=value, the
2158 quoting of the values was not always being done properly. Specifically:
2159 (a) If the value started with a double quote, but contained no whitespace,
2161 (b) If the value contained whitespace other than a space character (i.e.
2162 tabs or newlines or carriage returns) it was not quoted.
2163 This fix has been applied to the mysql and pgsql lookups by writing a
2164 separate quoting function and calling it from the lookup code. The fix
2165 should probably also be applied to nisplus, ibase and oracle lookups, but
2166 since I cannot test any of those, I have not disturbed their existing code.
2168 2. A hit in the callout cache for a specific address caused a log line with no
2169 reason for rejecting RCPT. Now it says "Previous (cached) callout
2170 verification failure".
2172 3. There was an off-by-one bug in the queryprogram router. An over-long
2173 return line was truncated at 256 instead of 255 characters, thereby
2174 overflowing its buffer with the terminating zero. As well as fixing this, I
2175 have increased the buffer size to 1024 (and made a note to document this).
2177 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2178 when Exim is waiting for an SMTP response from a remote server, Exim
2179 restarts its select() call on the socket, thereby resetting its timeout.
2180 This is not a problem when such interrupts are rare. Somebody set up a cron
2181 job to run exiwhat every 2 minutes, which is less than the normal select()
2182 timeout (5 or 10 minutes). This meant that the select() timeout never
2183 kicked in because it was always reset. I have fixed this by comparing the
2184 time when an interrupt arrives with the time at the start of the first call
2185 to select(). If more time than the timeout has elapsed, the interrupt is
2186 treated as a timeout.
2188 5. Some internal re-factoring in preparation for the addition of Sieve
2189 extensions (by MH). In particular, the "personal" test is moved to a
2190 separate function, and given an option for scanning Cc: and Bcc: (which is
2191 not set for Exim filters).
2193 6. When Exim created an email address using the login of the caller as the
2194 local part (e.g. when creating a From: or Sender: header line), it was not
2195 quoting the local part when it contained special characters such as @.
2197 7. Installed new OpenBSD configuration files.
2199 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2200 try to make them clearer.
2202 9. Callout options, other than the timeout value, were being ignored when
2203 verifying sender addresses in header lines. For example, when using
2205 verify = header_sender/callout=no_cache
2207 the cache was (incorrectly) being used.
2209 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2210 only the Cygwin environment.
2212 11. When return_path_on_delivery was set as a log selector, if different remote
2213 addresses in the same message used different return paths and parallel
2214 remote delivery occurred, the wrong values would sometimes be logged.
2215 (Whenever a remote delivery process finished, the return path value from
2216 the most recently started remote delivery process was logged.)
2218 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2219 lines when AUTH and/or TLS are in use. This is the "received protocol"
2220 field. Exim used to use "asmtp" for authenticated SMTP, without any
2221 indication (in the protocol name) for TLS use. Now it follows the RFC and
2222 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2223 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2224 names appear in log lines as well as in Received: header lines.
2226 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2227 extensions, and comparison tests, and to fix some bugs.
2229 14. Changes to the "personal" filter test:
2231 (1) The test was buggy in that it was just doing the equivalent of
2232 "contains" tests on header lines. For example, if a user's address was
2233 anne@some.where, the "personal" test would incorrectly be true for
2235 To: susanne@some.where
2237 This test is now done by extracting each address from the header in turn,
2238 and checking the entire address. Other tests that are part of "personal"
2239 are now done using regular expressions (for example, to check local parts
2240 of addresses in From: header lines).
2242 (2) The list of non-personal local parts in From: addresses has been
2243 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2244 taken from the Sieve specification recommendations.
2246 (3) If the message contains any header line starting with "List-" it is
2247 treated as non-personal.
2249 (4) The test for "circular" in the Subject: header line has been removed
2250 because it now seems ill-conceived.
2252 15. Minor typos in src/EDITME comments corrected.
2254 16. Installed latest exipick from John Jetmore.
2256 17. If headers_add on a router specified a text string that was too long for
2257 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2258 of string_sprintf() is now avoided.
2260 18. $message_body_size was not set (it was always zero) when running the DATA
2261 ACL and the local_scan() function.
2263 19. For the "mail" command in an Exim filter, no default was being set for
2264 the once_repeat time, causing a random time value to be used if "once" was
2265 specified. (If the value happened to be <= 0, no repeat happened.) The
2266 default is now 0s, meaning "never repeat". The "vacation" command was OK
2267 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2268 (I found it when inspecting the code).
2270 20. There is now an overall timeout for performing a callout verification. It
2271 defaults to 4 times the callout timeout, which applies to individual SMTP
2272 commands during the callout. The overall timeout applies when there is more
2273 than one host that can be tried. The timeout is checked before trying the
2274 next host. This prevents very long delays if there are a large number of
2275 hosts and all are timing out (e.g. when the network connections are timing
2276 out). The value of the overall timeout can be changed by specifying an
2277 additional sub-option for "callout", called "maxwait". For example:
2279 verify = sender/callout=5s,maxwait=20s
2281 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2282 to the end before writing, but this should make it even safer).
2284 22. Exim was forgetting that it had advertised PIPELINING for the second and
2285 subsequent messages on an SMTP connection. It was also not resetting its
2286 memory on STARTTLS and an internal HELO.
2288 23. When Exim logs an SMTP synchronization error within a session, it now
2289 records whether PIPELINING has been advertised or not.
2291 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2292 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2293 rather than long int.
2295 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2297 26. Added the never_mail option to autoreply.
2303 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2304 crash if the getsockname() call failed; this can happen if a connection is
2305 closed very soon after it is established. The problem was simply in the
2306 order in which certain operations were done, causing Exim to try to write
2307 to the SMTP stream before it had set up the file descriptor. The bug has
2308 been fixed by making things happen in the correct order.
2314 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2315 before the connection was closed, thus losing the rejection response.
2317 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2318 some early Solaris releases, but causes trouble in current releases where
2319 socklen_t is defined.
2321 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2324 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2327 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2328 that releasing the top of it at the end releases what was used for sub-
2329 expansions (unless the block got too big). However, discard this block if
2330 the first thing is a variable or header, so that we can use its block when
2331 it is dynamic (useful for very large $message_headers, for example).
2333 6. Lookups now cache *every* query, not just the most recent. A new, separate
2334 store pool is used for this. It can be recovered when all lookup caches are
2335 flushed. Lookups now release memory at the end of their result strings.
2336 This has involved some general refactoring of the lookup sources.
2338 7. Some code has been added to the store_xxx() functions to reduce the amount
2339 of flapping under certain conditions.
2341 8. log_incoming_interface used to affect only the <= reception log lines. Now
2342 it causes the local interface and port to be added to several more SMTP log
2343 lines, for example "SMTP connection from", and rejection lines.
2345 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2347 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2349 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2350 could be overwritten at the end of the current message (or the start of a
2351 new message if it was set in a HELO ACL). The value is now preserved for
2352 the duration of the SMTP connection.
2354 12. If a transport had a headers_rewrite setting, and a matching header line
2355 contained an unqualified address, that address was qualified, even if it
2356 did not match any rewriting rules. The underlying bug was that the values
2357 of the flags that permit the existence of unqualified sender and recipient
2358 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2359 non-local messages, and by -bnq for local messages) were not being
2360 preserved with the message after it was received.
2362 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2363 "next input=" as part of the text comprising the host identity instead of
2364 the correct text. The code was using the same buffer for two different
2365 strings. However, depending on which order the printing function evaluated
2366 its arguments, the bug did not always show up. Under Linux, for example, my
2367 test suite worked just fine.
2369 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2370 doesn't work with some older versions of Perl. It has been changed to "my",
2371 which in any case is probably the better facility to use.
2373 15. A really picky compiler found some instances of statements for creating
2374 error messages that either had too many or two few arguments for the format
2377 16. The size of the buffer for calls to the DNS resolver has been increased
2378 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2379 for addresses that have a lot of PTR records. This alleviates a problem; it
2380 does not fully solve it.
2382 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2383 buffer now truncates the list and logs the incident, which is the same
2384 action as happens when Exim is looking up a host name and its aliases.
2385 Previously in this situation something unpredictable would happen;
2386 sometimes it was "internal error: store_reset failed".
2388 18. If a server dropped the connection unexpectedly when an Exim client was
2389 using GnuTLS and trying to read a response, the client delivery process
2390 crashed while trying to generate an error log message.
2392 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2393 string, for example:
2395 warn message = H1: something\nH2: something
2397 the text was added as a single header line from Exim's point of view
2398 though it ended up OK in the delivered message. However, searching for the
2399 second and subsequent header lines using $h_h2: did not work. This has been
2400 fixed. Similarly, if a system filter added multiple headers in this way,
2401 the routers could not see them.
2403 20. Expanded the error message when iplsearch is called with an invalid key to
2404 suggest using net-iplsearch in a host list.
2406 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2407 ACLs are no longer actually imposed (and a message to that effect is
2410 22. If a "gecos" field in a passwd entry contained escaped characters, in
2411 particular, if it contained a \" sequence, Exim got it wrong when building
2412 a From: or a Sender: header from that name. A second bug also caused
2413 incorrect handling when an unquoted " was present following a character
2414 that needed quoting.
2416 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2417 was not being matched caselessly.
2419 24. Arranged for all hyphens in the exim.8 source to be escaped with
2422 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2423 properly. Recipient callout cache records were still being keyed to include
2424 the sender, even when use_sender was set false. This led to far more
2425 callouts that were necessary. The sender is no longer included in the key
2426 when use_sender is false.
2428 26. Added "control = submission" modifier to ACLs.
2430 27. Added the ${base62d: operator to decode base 62 numbers.
2432 28. dnsdb lookups can now access SRV records.
2434 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2435 the configuration file.
2437 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2438 (-v) mode. This makes the output for a verbose queue run more intelligible.
2440 31. Added a use_postmaster feature to recipient callouts.
2442 32. Added the $body_zerocount variable, containing the number of binary zero
2443 bytes in the message body.
2445 33. The time of last modification of the "new" subdirectory is now used as the
2446 "mailbox time last read" when there is a quota error for a maildir
2449 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2451 35. Added +ignore_unknown as a special item in host lists.
2453 36. Code for decoding IPv6 addresses in host lists is now included, even if
2454 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2455 address was recognized as an IP address, but was then not correctly decoded
2456 into binary, causing unexpected and incorrect effects when compared with
2463 1. Very minor rewording of debugging text in manualroute to say "list of
2464 hosts" instead of "hostlist".
2466 2. If verify=header_syntax was set, and a header line with an unqualified
2467 address (no domain) and a large number of spaces between the end of the
2468 name and the colon was received, the reception process suffered a buffer
2469 overflow, and (when I tested it) crashed. This was caused by some obsolete
2470 code that should have been removed. The fix is to remove it!
2472 3. When running in the test harness, delay a bit after writing a bounce
2473 message to get a bit more predictability in the log output.
2475 4. Added a call to search_tidyup() just before forking a reception process. In
2476 theory, someone could use a lookup in the expansion of smtp_accept_max_
2477 per_host which, without the tidyup, could leave open a database connection.
2479 5. Added the variables $recipient_data and $sender_data which get set from a
2480 lookup success in an ACL "recipients" or "senders" condition, or a router
2481 "senders" option, similar to $domain_data and $local_part_data.
2483 6. Moved the writing of debug_print from before to after the "senders" test
2486 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2487 problems for message scanning, either using a data ACL, or using
2488 local_scan() because the Received: header was not generated till after they
2489 were called (in order to set the time as the time of reception completion).
2490 I have revised the way this works. The header is now generated after the
2491 body is received, but before the ACL or local_scan() are called. After they
2492 are run, the timestamp in the header is updated.
2498 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2499 before starting a queue runner without re-exec. This happened only when
2500 deliver_drop_privilege was set or when the Exim user was set to root. The
2501 effect of the bug was that timeouts during subsequent deliveries caused
2502 crashes instead of being properly handled. The handler is now left at its
2503 default (and expected) setting.
2505 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2506 message, again when deliver_drop_privilege is set or Exim is run as root.
2507 The bug described in (1) was not present in this case, but the tidying up
2508 of the other signals was missing. I have made the two cases consistent.
2510 3. The ignore_target_hosts setting on a manualroute router was being ignored
2511 for hosts that were looked up using the /MX notation.
2513 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2516 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2517 operated on the sender address. After changing the $sender_address to <>
2518 for the sender address verify, Exim was re-instated it as the original
2519 (before rewriting) address, but remembering that it had rewritten it, so it
2520 wasn't rewriting it again. This bug also had the effect of breaking the
2521 sender address verification caching when the sender address was rewritten.
2523 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2524 This has been changed so that if the ip literal address matches
2525 ignore_target_hosts, the router declines.
2527 7. Added expansion conditions match_domain, match_address, and match_local_
2528 part (NOT match_host).
2530 8. The placeholder for the Received: header didn't have a length field set.
2532 9. Added code to Exim itself and to exim_lock to test for a specific race
2533 condition that could lead to file corruption when using MBX delivery. The
2534 issue is with the lockfile that is created in /tmp. If this file is removed
2535 after a process has opened it but before that process has acquired a lock,
2536 there is the potential for a second process to recreate the file and also
2537 acquire a lock. This could lead to two Exim processes writing to the file
2538 at the same time. The added code performs the same test as UW imapd; it
2539 checks after acquiring the lock that its file descriptor still refers to
2540 the same named file.
2542 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2543 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2546 11. Added the smtp_active_hostname option. If used, this will typically be made
2547 to depend on the incoming interface address. Because $interface_address is
2548 not set up until the daemon has forked a reception process, error responses
2549 that can happen earlier (such as "too many connections") no longer contain
2552 12. If an expansion in a condition on a "warn" statement fails because a lookup
2553 defers, the "warn" statement is abandoned, and the next ACL statement is
2554 processed. Previously this caused the whole ACL to be aborted.
2556 13. Added the iplsearch lookup type.
2558 14. Added ident_timeout as a log selector.
2560 15. Added tls_certificate_verified as a log selector.
2562 16. Added a global option tls_require_ciphers (compare the smtp transport
2563 option of the same name). This controls incoming TLS connections.
2565 17. I finally figured out how to make tls_require_ciphers do a similar thing
2566 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2567 before starting the TLS session.
2569 18. Tabs are now shown as \t in -bP output.
2571 19. If the log selector return_path_on_delivery was set, Exim crashed when
2572 bouncing a message because it had too many Received: header lines.
2574 20. If two routers both had headers_remove settings, and the first one included
2575 a superfluous trailing colon, the final name in the first list and the
2576 first name in the second list were incorrectly joined into one item (with a
2577 colon in the middle).
2583 1. Added -C and -D options to the exinext utility, mainly to make it easier
2584 to include in the automated testing, but these could be helpful when
2585 multiple configurations are in use.
2587 2. The exinext utility was not formatting the output nicely when there was
2588 an alternate port involved in the retry record key, nor when there was a
2589 message id as well (for retries that were specific to a specific message
2590 and a specific host). It was also confused by IPv6 addresses, because of
2591 the additional colons they contain. I have fixed the IPv4 problem, and
2592 patched it up to do a reasonable job for IPv6.
2594 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2595 delivery, the log line now contains "pipelined" if PIPELINING was used.
2597 4. An SMTP transport process used to panic and die if the bind() call to set
2598 an explicit outgoing interface failed. This has been changed; it is now
2599 treated in the same way as a connect() failure.
2601 5. A reference to $sender_host_name in the part of a conditional expansion
2602 that was being skipped was still causing a DNS lookup. This no longer
2605 6. The def: expansion condition was not recognizing references to header lines
2606 that used bh_ and bheader_.
2608 7. Added the _cache feature to named lists.
2610 8. The code for checking quota_filecount in the appendfile transport was
2611 allowing one more file than it should have been.
2613 9. For compatibility with Sendmail, the command line option
2621 and sets the incoming protocol and host name (for trusted callers). The
2622 host name and its colon can be omitted when only the protocol is to be set.
2623 Note the Exim already has two private options, -pd and -ps, that refer to
2624 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2625 "s", but I don't think that's a major issue.
2627 10. A number of refactoring changes to the code, none of which should affect
2630 (a) The number of logging options was getting close to filling up the
2631 32-bit word that was used as a bit map. I have split them into two classes:
2632 those that are passed in the argument to log_write(), and those that are
2633 only ever tested independently outside of that function. These are now in
2634 separate 32-bit words, so there is plenty of room for expansion again.
2635 There is no change in the user interface or the logging behaviour.
2637 (b) When building, for example, log lines, the code previously used a
2638 macro that called string_cat() twice, in order to add two strings. This is
2639 not really sufficiently general. Furthermore, there was one instance where
2640 it was actually wrong because one of the argument was used twice, and in
2641 one call a function was used. (As it happened, calling the function twice
2642 did not affect the overall behaviour.) The macro has been replaced by a
2643 function that can join an arbitrary number of extra strings onto a growing
2646 (c) The code for expansion conditions now uses a table and a binary chop
2647 instead of a serial search (which was left over from when there were very
2648 few conditions). Also, it now recognizes conditions like "pam" even when
2649 the relevant support is not compiled in: a suitably worded error message is
2650 given if an attempt is made to use such a condition.
2652 11. Added ${time_interval:xxxxx}.
2654 12. A bug was causing one of the ddress fields not to be passed back correctly
2655 from remote delivery subprocesses. The field in question was not being
2656 subsequently used, so this caused to problems in practice.
2658 13. Added new log selectors queue_time and deliver_time.
2660 14. Might have fixed a bug in maildirsizefile handling that threw up
2661 "unexpected character" debug warnings, and recalculated the data
2662 unnecessarily. In any case, I expanded the warning message to give more
2665 15. Added the message "Restricted characters in address" to the statements in
2666 the default ACL that block characters like @ and % in local parts.
2668 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2669 Three changes have been made:
2671 (a) There was a serious bug; a negative response to MAIL caused the whole
2672 recipient domain to be cached as invalid, thereby blocking all messages
2673 to all local parts at the same domain, from all senders. This bug has
2674 been fixed. The domain is no longer cached after a negative response to
2675 MAIL if the sender used is not empty.
2677 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2680 (c) A new callout option, "use_sender" has been added for people who want
2681 the modified behaviour.
2687 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2690 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2691 indeed breaks things for older releases.
2693 3. Added additional logging to the case where there is a problem reading data
2694 from a filter that is running in a subprocess using a pipe, in order to
2695 try to track down a specific problem.
2697 4. Testing facility fudge: when running in the test harness and attempting
2698 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2699 getting "No route to host". Convert this to a timeout.
2701 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2704 6. Some OS don't have socklen_t but use size_t instead. This affects the
2705 fifth argument of getsockopt() amongst other things. This is now
2706 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2707 can be set for individual OS. I have set it for SunOS5, OSF1, and
2708 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2709 some earlier ones do not.
2711 7. Change 4.30/15 was not doing the test caselessly.
2713 8. The standard form for an IPv6 address literal was being rejected by address
2714 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2715 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2716 this, as well as the form without the "IPv6" on the front (but only when
2717 address literals are enabled, of course).
2719 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2721 10. Exim crashed if a message with an empty sender address specified by -f
2722 encountered a router with an errors_to setting. This could be provoked only
2723 by a command such as
2727 where an empty string was supplied; "<>" did not hit this bug.
2729 11. Installed PCRE release 4.5.
2731 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2732 remained set. It is now erased.
2734 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2735 times from message ids (which are base 36 rather than the normal 62).
2737 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2738 were being counted as actual protocol errors, and logged if the log
2739 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2740 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2741 and DATA following a set of rejected RCPTs do not count as protocol errors.
2742 In other words, Exim assumes they were pipelined, though this may not
2743 actually be the case. Of course, in all cases the client gets an
2744 appropriate error code.
2746 15. If a lookup fails in an ACL condition, a message about the failure may
2747 be available; it is used if testing the ACL cannot continue, because most
2748 such messages specify what the cause of the deferral is. However, some
2749 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2750 that caused an old message to be retained and used if a later statement
2751 caused a defer, replacing the real cause of the deferral.
2753 16. If an IP address had so many PTR records that the DNS lookup buffer
2754 was not large enough to hold them, Exim could crash while trying to process
2755 the truncated data. It now detects and logs this case.
2757 17. Further to 4.21/58, another change has been made: if (and only if) the
2758 first line of a message (the first header line) ends with CRLF, a bare LF
2759 in a subsequent header line has a space inserted after it, so as not to
2760 terminate the header.
2762 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2763 unnecessarily, used atoi() instead of strtol(), and didn't check the
2764 termination when getting file sizes from file names by regex.
2766 19. Completely re-implemented the support for maildirsize files, in the light
2767 of a number of problems with the previous contributed implementation
2768 (4.30/29). In particular:
2770 . If the quota is zero, the maildirsize file is maintained, but no quota is
2773 . If the maildir directory does not exist, it is created before any attempt
2774 to write a maildirsize file.
2776 . The quota value in the file is just a cache; if the quota is changed in
2777 the transport, the new value overrides.
2779 . A regular expression is available for excluding directories from the
2782 20. The autoreply transport checks the characters in options that define the
2783 message's headers; it allows continued headers, but it was checking with
2784 isspace() after an embedded newline instead of explicitly looking for a
2787 21. If all the "regular" hosts to which an address was routed had passed their
2788 expiry times, and had not reached their retry times, the address was
2789 bounced, even if fallback hosts were defined. Now Exim should go on to try
2792 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2793 equivalent code in the SMTP transport. Some hosts send humungous responses
2794 to HELO/EHLO, more than 1024 it seems.
2796 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2797 gives compiler warnings in some environments. I've now done it "properly",
2800 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2801 (because of problems with the built-in one) was declared to return uschar *
2802 instead of char *, causing compiler failure.
2804 25. Fixed a file descriptor leak when processing alias/forward files.
2806 26. Fixed a minor format string issue in dbfn.c.
2808 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2810 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2811 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2814 29. When a sender address is verified, it is cached, to save repeating the test
2815 when there is more than one recipient in a message. However, when the
2816 verification involves a callout, it is possible for different callout
2817 options to be set for different recipients. It is too complicated to keep
2818 track of this in the cache, so now Exim always runs a verification when a
2819 callout is required, relying on the callout cache for the optimization.
2820 The overhead is duplication of the address routing, but this should not be
2823 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2824 to be verified with callout=postmaster, and the main callout worked but the
2825 postmaster check failed, the verification correctly failed. However, if a
2826 subsequent RCPT command asked for sender verification *without* the
2827 postmaster check, incorrect caching caused this verification also to fail,
2830 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2831 it was not caching the DNS options (qualify_single, search_parents) that
2832 were used when the lookup failed. A subsequent lookup with different
2833 options therefore always gave the same answer, though there were cases
2834 where it should not have. (Example: a "domains = !$mx_any" option on a
2835 dnslookup router: the "domains" option is always processed without any
2836 widening, but the router might have qualify_single set.) Now Exim uses the
2837 cached value only when the same options are set.
2839 32. Added John Jetmore's "exipick" utility to the distribution.
2841 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2842 than a timeout (e.g. a certificate is required, and is not provided), an
2843 Exim server now closes the connection immediately. Previously it waited for
2844 the client to close - but if the client is SSL, it seems that they each
2845 wait for each other, leading to a delay before one of them times out.
2847 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2848 maintained 0.8.x compatibility because I don't think many are using it, and
2849 it is clearly obsolete.
2851 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2854 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2855 client certificate was expired. A simple patch fixes this, though I don't
2856 understand the full logic of why the verify callback is called multiple
2859 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2860 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2861 which causes problems with some clients (such as the Certicom SSL Plus
2862 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2863 disables the coutermeasure allowing Eudora to connect."
2865 38. Exim was not checking that a write() to a log file succeeded. This could
2866 lead to Bad Things if a log got too big, in particular if it hit a file
2867 size limit. Exim now panics and dies if it cannot write to a log file, just
2868 as it does if it cannot open a log file.
2870 39. Modified OS/Makefile-Linux so that it now contains
2872 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2874 The two -D definitions ensure that Exim is compiled with large file
2875 support, which makes it possible to handle log files that are bigger than
2878 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2879 instance) a domain was checked against a named list that involved a lookup,
2880 causing $domain_data to be set, then another domain was checked against the
2881 same list, then the first domain was re-checked, the value of $domain_data
2882 after the final check could be wrong. In particular, if the second check
2883 failed, it could be set empty. This bug probably also applied to
2886 41. The strip_trailing_dot option was not being applied to the address given
2887 with the -f command-line option.
2889 42. The code for reading a message's header from the spool was incrementing
2890 $received_count, but never initializing it. This meant that the value was
2891 incorrect (doubled) while delivering a message in the same process in which
2892 it was received. In the most common configuration of Exim, this never
2893 happens - a fresh exec is done - but it can happen when
2894 deliver_drop_privilege is set.
2896 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2897 it now includes up to 150 characters of the unexpected data in the log
2900 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2901 and building data strings. The size of both of these buffers was 10 000
2902 bytes - far larger than anybody would *ever* want, thought I. Needless to
2903 say, somebody hit the limit. I have increased the maximum line length to
2904 20 000 and the maximum data length of concatenated lines to 100 000. I have
2905 also fixed two bugs, because there was no checking on these buffers. Tsk,
2906 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2907 buffer is too small.
2909 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2910 lsearch lookups. Now it does.
2912 46. When parsing a route_list item in a manualroute router, a fixed-length
2913 buffer was used for the list of hosts. I made this 1024 bytes long,
2914 thinking that nobody would ever have a list of hosts that long. Wrong.
2915 Somebody had a whole pile of complicated expansion conditions, and the
2916 string was silently truncated, leading to an expansion error. It turns out
2917 that it is easier to change to an unlimited length (owing to other changes
2918 that have happened since this code was originally written) than to build
2919 structure for giving a limitation error. The length of the item that
2920 expands into the list of hosts is now unlimited.
2922 47. The lsearch lookup could not handle data where the length of text line was
2923 more than 4095 characters. Such lines were truncated, leading to shortened
2924 data being returned. It should now handle lines of any length.
2926 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2927 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2930 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2931 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2932 no longer happens in comment lines. A list of replacements is now placed
2933 at the head of all of the source files, except those whose only change is
2934 to replace PERL_COMMAND in the very first #! line.
2936 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2937 messages on the queue, with a bottom-up merge sort, using code contributed
2938 by Michael Haardt. This should make operations like -bp somewhat faster on
2939 large queues. It won't affect queue runners, except when queue_run_in_order
2942 51. Installed eximstats 1.31 in the distribution.
2944 52. Added support for SRV lookups to the dnslookup router.
2946 53. If an ACL referred to $message_body or $message_body_end, the value was not
2947 reset for any messages that followed in the same SMTP session.
2949 54. The store-handling optimization for building very long strings was not
2950 differentiating between the different store pools. I don't think this
2951 actually made any difference in practice, but I've tidied it.
2953 55. While running the routers to verify a sender address, $sender_address
2954 was still set to the sender address. This is wrong, because when routing to
2955 send a bounce to the sender, it would be empty. Therefore, I have changed
2956 it so that, while verifying a sender address, $sender_address is set to <>.
2957 (There is no change to what happens when verifying a recipient address.)
2959 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2960 target A or AAAA records (if not already returned) without resetting the
2961 qualify_single or search_parents options of the DNS resolver. These are
2962 inappropriate in this case because the targets of MX and SRV records must
2963 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2964 target that, when qualified, matched something in the local domain. These
2965 two options are now turned off when doing these lookups.
2967 57. It seems that at least some releases of Reiserfs (which does not have the
2968 concept of a fixed number of inodes) returns zero and not -1 for the
2969 number of available inodes. This interacted badly with check_spool_inodes,
2970 which assumed that -1 was the "no such thing" setting. What I have done is
2971 to check that the total number of inodes is greater than zero before doing
2972 the test of how many are available.
2974 58. When a "warn" ACL statement has a log_message modifier, the message is
2975 remembered, and not repeated. This is to avoid a lot of repetition when a
2976 message has many recipients that cause the same warning to be written.
2977 Howewer, Exim was preserving the list of already written lines for an
2978 entire SMTP session, which doesn't seem right. The memory is now reset if a
2979 new message is started.
2981 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2982 debugging output unless log_rewrite was also set.
2984 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2985 of (int)(handle) when we know that handle contains (void *)(-1).
2987 61. The Exim daemon panic-logs an error return when it closes the incoming
2988 connection. However "connection reset by peer" seems to be common, and
2989 isn't really an error worthy of noting specially, so that particular error
2992 62. When Exim is trying to find all the local interfaces, it used to panic and
2993 die if the ioctl to get the interface flags failed. However, it seems that
2994 on at least one OS (Solaris 9) it is possible to have an interface that is
2995 included in the list of interfaces, but for which you get a failure error
2996 for this call. This happens when the interface is not "plumbed" into a
2997 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2998 failure of the "get flags" call assumes that the interface is down.
3000 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3001 makes life easier for people who are doing arithmetic on fields extracted
3002 from dates, where you often get leading zeros that should not be
3003 interpreted as octal.
3005 64. Added qualify_domain to the redirect router, to override the global
3008 65. If a pathologically long header line contained very many addresses (the
3009 report of this problem mentioned 10 000) and each of them was rewritten,
3010 Exim could use up a very large amount of memory. (It kept on making new
3011 copies of the header line as it rewrote, and never released the old ones.)
3012 At the expense of a bit more processing, the header rewriting function has
3013 been changed so that it no longer eats memory in this way.
3015 66. The generation of the Received: header has been moved from the time that a
3016 message starts to be received, to the time that it finishes. The timestamp
3017 in the Received: header should now be very close to that of the <= log
3018 line. There are two side-effects of this change:
3020 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3021 logged header lines no longer include the local Received: line, because
3022 it has not yet been created. The same applies to a copy of the message
3023 that is returned to a non-SMTP sender when a message is rejected.
3025 (b) When a filter file is tested using -bf, no additional Received: header
3026 is added to the test message. After some thought, I decided that this
3029 This change does not affect the value of $received_for. It is still set
3030 after address rewriting, but before local_scan() is called.
3032 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3034 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3035 gave an unhelpful panic error message, and a defer error. I have managed to
3036 change this behaviour so that it now rejects any supplied certificate,
3037 which seems right, as the list of acceptable certificates is empty.
3039 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3040 gave an unhelpful defer error. I have not managed to make this reject any
3041 supplied certificates, but the error message it gives is "no certificate
3042 supplied", which is not helpful.
3044 70. exigrep's output now also includes lines that are not associated with any
3045 message, but which match the given pattern. Implemented by a patch from
3046 Martin Sluka, which also tidied up the Perl a bit.
3048 71. Recipient callout verification, like sender verification, was using <> in
3049 the MAIL FROM command. This isn't really the right thing, since the actual
3050 sender may affect whether the remote host accepts the recipient or not. I
3051 have changed it to use the actual sender in the callout; this means that
3052 the cache record is now keyed on a recipient/sender pair, not just the
3053 recipient address. There doesn't seem to be a real danger of callout loops,
3054 since a callout by the remote host to check the sender would use <>.
3055 [SEE ABOVE: changed after hitting problems.]
3057 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3058 temporary errors. However, in the case of such a code being given after
3059 the end of a data transmission (i.e. after ".") Exim was failing to write
3060 a retry record for the message. (Yes, there was some broken host that was
3061 actually sending 8xx at this point.)
3063 73. An unknown lookup type in a host list could cause Exim to panic-die when
3064 the list was checked. (An example that provoked this was putting <; in the
3065 middle of a list instead of at the start.) If this happened during a DATA
3066 ACL check, a -D file could be left lying around. This kind of configuration
3067 error no longer causes Exim to die; instead it causes a defer errror. The
3068 incident is still logged to the main and panic logs.
3070 74. Buglet left over from Exim 3 conversion. The message "too many messages
3071 in one connection" was written to the rejectlog but not the mainlog, except
3072 when address rewriting (yes!) was being logged.
3074 75. Added write_rejectlog option.
3076 76. When a system filter was run not as root (that is, when system_filter_user
3077 was set), the values of the $n variables were not being returned to the
3078 main process; thus, they were not subsequently available in the $sn
3081 77. Added +return_path_on_delivery log selector.
3083 78. A connection timeout was being treated differently from recipients deferred
3084 when testing hosts_max_try with a message that was older than the host's
3085 retry timeout. (The host should not be counted, thus allowing all hosts to
3086 be tried at least once before bouncing.) This may have been the cause of an
3087 occasionally reported bug whereby a message would remain on the queue
3088 longer than the retry timeout, but would be bounced if a delivery was
3089 forced. I say "may" because I never totally pinned down the problem;
3090 setting up timeout/retry tests is difficult. See also the next item.
3092 79. The ultimate address timeout was not being applied to errors that involved
3093 a combination of host plus message (for example, a timeout on a MAIL
3094 command). When an address resolved to a number of possible hosts, and they
3095 were not all tried for each delivery (e.g. because of hosts_max_try), a
3096 message could remain on the queue longer than the retry timeout.
3098 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3101 81. Fixed an obscure SMTP outgoing bug which required at least the following
3102 conditions: (a) there was another message waiting for the same server;
3103 (b) the server returned 5xx to all RCPT commands in the first message so
3104 that the message was not completed; (c) the server dropped the connection
3105 or gave a negative response to the RSET that Exim sends to abort the
3106 transaction. The observed case was a dropped connection after DATA that had
3107 been sent in pipelining mode. That is, the server had advertised PIPELINING
3108 but was not implementing it correctly. The effect of the bug was incorrect
3109 behaviour, such as trying another host, and this could lead to a crash.
3115 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3116 and daemon.c were passed as pointers to ints; they should have been
3117 pointers to socklen_t variables (which are typically unsigned ints).
3119 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3122 3. Fixed a really odd bug that affected only the testing scheme; patching a
3123 certain fixed string in the binary changed the value of another string that
3124 happened to be identical to the end of the original first string.
3126 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3127 name", it returns that address as the IP address. On some operating
3128 systems (e.g. Solaris), it also passes back the IP address string as the
3129 "host name". However, on others (e.g. Linux), it passes back an empty
3130 string. Exim wasn't checking for this, and was changing the host name to an
3131 empty string, assuming it had been canonicized.
3133 5. Although rare, it is permitted to have more than one PTR record for a given
3134 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3135 all the names associated with an address, because they do in Solaris.
3136 However, it seems that they do not in Linux for data that comes from the
3137 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3138 I found this out when I moved to a new Linux workstation and tried to run
3139 the Exim test suite.
3141 To get round this problem I have changed the code so that it now does its
3142 own call to the DNS to look up PTR records when searching for a host name.
3143 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3144 addresses that are only in /etc/hosts are still found.
3146 This behaviour is, however, controlled by an option called host_lookup_
3147 order, which defaults to "bydns:byaddr". If people want to use the other
3148 order, or indeed, just use one or the other means of lookup, they can
3149 specify it in this variable.
3151 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3152 some operating systems, this comes back from gethostbyaddr() as an empty
3153 string, and this is what Exim used to test for. However, it seems that in
3154 other systems, "." is yielded. Exim now tests for this case too.
3156 7. The values of check_spool_space and check_log_space are now held internally
3157 as a number of kilobytes instead of an absolute number of bytes. If a
3158 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3159 kilobyte. This means that much larger values can be stored.
3161 8. Exim monitor: an attempt to get the action menu when not actually pointing
3162 at a message produces an empty menu entitled "No message selected". This
3163 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3164 no entries in it ("Shell widget menu has zero width and/or height"). So I
3165 have added a single, blank menu entry in this case.
3167 9. Added ${quote_local_part.
3169 10. MIME decoding is now applied to the contents of Subject: header lines when
3172 11. Now that a reference to $sender_host_address automatically causes a reverse
3173 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3174 host lookup before query-style lookups in lists that might use this
3175 variable. This has therefore been abolished, and the "net-" prefix is no
3176 longer necessary for query-style lookups.
3178 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3179 have been a typo for LFLAGS, so it has been changed.
3181 13. The install script calls Exim with "-C /dev/null" in order to find the
3182 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3183 to be output. Howeve, since Exim outputs its version number before the
3184 error, it didn't break the script. It just looked ugly. I fixed this by
3185 always allowing "-C /dev/null" if the caller is root.
3187 14. Ignore overlarge ACL variable number when reading spool file - insurance
3188 against a later release with more variables having written the file.
3190 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3191 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3192 this, as well as the form without the "IPv6" on the front.
3194 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3195 OS/Makefile-Darwin file.
3197 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3198 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3200 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3201 variable. However, some OS use ints for inodes. I've added cast to long int
3202 to get rid of the compiler warning.
3204 19. I had forgotten to lock out "/../" in configuration file names when
3205 ALT_CONFIG_PREFIX was set.
3207 20. Routers used for verification do not need to specify transports. However,
3208 if such a router generated a host list, and callout was configured, Exim
3209 crashed, because it could not find a port number from the (non-existent)
3210 transport. It now assumes port 25 in this circumstance.
3212 21. Added the -t option to exigrep.
3214 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3215 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3216 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3217 them is set, LOOKUP_LSEARCH is forced.
3219 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3221 24. Added sender and host information to the "rejected by local_scan()" log
3222 line; previously there was no indication of these.
3224 25. Added .include_if_exists.
3226 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3227 Linux. It turns out that not all file systems support this. Apparently some
3228 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3229 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3230 not supported on the file descriptor, is now ignored when Exim is trying to
3231 sync a directory. This applies only to Linux.
3233 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3235 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3236 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3239 29. Added support for maildirsize files from supplied patch (modified a bit).
3241 30. The use of :fail: followed by an empty string could lead Exim to respond to
3242 sender verification failures with (e.g.):
3244 550 Verification failed for <xxx>
3245 550 Sender verify failed
3247 where the first response line was missing the '-' that indicates it is not
3248 the final line of the response.
3250 31. The loop for finding the name of the user that called Exim had a hardwired
3251 limit of 10; it now uses the value of finduser_retries, which is used for
3252 all other user lookups.
3254 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3257 33. Exim was neglecting to zero errno before one call of strtol() when
3258 expanding a string and expecting an integer value. On some systems this
3259 resulted in spurious "integer overflow" errors. Also, it was casting the
3260 result into an int without checking.
3262 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3263 did not work. The code looks as if it has *never* worked, though it appears
3264 to have been documented since at least releast 1.62. I have made it work.
3266 35. The "timeout_DNS" error in retry rules, also documented since at least
3267 1.62, also never worked. As it isn't clear exactly what this means, and
3268 clearly it isn't a major issue, I have abolished the feature by treating it
3269 as "timeout", and writing a warning to the main and panic logs.
3271 36. The display of retry rules for -brt wasn't always showing the error code
3274 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3275 timeout_connect_A, timeout_connect_MX.
3277 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3278 to the empty sender.
3280 39. The daemon was not analysing the content of -oX till after it had closed
3281 stderr and disconnected from the controlling terminal. This meant that any
3282 syntax errors were only noted on the panic log, and the return code from
3283 the command was 0. By re-arranging the code a little, I've made the
3284 decoding happen first, so such errors now appear on stderr, and the return
3285 code is 1. However, the actual setting up of the sockets still happens in
3286 the disconnected process, so errors there are still only recorded on the
3289 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3290 connections (as happens on some IP stacks) was logged at start up time as
3291 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3292 it from "IPv6 and IPv4", which means that two separate sockets are being
3295 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3296 says whether AF_INET or AF_INET6 was passed as an argument.
3298 42. Exiwhat output was messed up when time zones were included in log
3301 43. Exiwhat now gives more information about the daemon's listening ports,
3302 and whether -tls-on-connect was used.
3304 44. The "port" option of the smtp transport is now expanded.
3306 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3307 silently ignored. Now an error message is written to the main and panic
3310 46. There's a new ACL modifier called "logwrite" which writes to a log file
3311 as soon as it is encountered.
3313 47. Added $local_user_uid and $local_user_gid at routing time.
3315 48. Exim crashed when trying to verify a sender address that was being
3318 49. Exim was recognizing only a space character after ".include". It now also
3319 recognizes a tab character.
3321 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3322 extracting the relevant information from the specification. The man page no
3323 longer contains scrambled data for the -d option, and I've added a section
3324 at the front about calling Exim under different names.
3326 51. Added "extra_headers" argument to the "mail" command in filter files.
3328 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3331 53. Installed eximstats 1.29.
3333 54. Added transport_filter_timeout as a generic transport option.
3335 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3336 Cc: header lines. This was required by RFC 822, but it not required by RFC
3339 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3340 incoming messages that did not have them. Now it does so only if the
3341 message originates locally, that is, if there is no associated remote host
3342 address. When Resent- header lines are present, this applies to the Resent-
3343 lines rather than the non-Resent- lines.
3345 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3346 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3348 58. Messages for configuration errors now include the name of the main
3349 configuration file - useful now that there may be more than one file in a
3350 list (.included file names were always shown).
3352 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3353 for those rare installations that do not start the daemon as root or run it
3354 setuid root. I've cut out the call to initgroups() if the daemon is not
3357 60. The Exim user and group can now be bound into the binary as text strings
3358 that are looked up at the start of Exim's processing.
3360 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3362 62. Added $mailstore_basename variable.
3364 63. Installed patch to sieve.c from Michael Haardt.
3366 64. When Exim failed to open the panic log after failing to open the main log,
3367 the original message it was trying to log was written to stderr and debug
3368 output, but if they were not available (the usual case in production), it
3369 was lost. Now it is written to syslog before the two lines that record the
3370 failures to open the logs.
3372 65. Users' Exim filters run in subprocesses under the user's uid. It is
3373 possible for a "deliver" command or an alias in a "personal" command to
3374 provoke an address rewrite. If logging of address rewriting is configured,
3375 this fails because the process is not running as root or exim. There may be
3376 a better way of dealing with this, but for the moment (because 4.30 needs
3377 to be released), I have disabled address rewrite logging when running a
3378 filter in a non-root, non-exim process.
3384 1. The buildconfig auxiliary program wasn't quoting the value set for
3385 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3386 not defined. This bug was masked in 4.22 by the effect that was fixed in
3389 2. Some messages that were rejected after a message id was allocated were
3390 shown as "incomplete" by exigrep. It no longer does this for messages that
3391 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3393 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3394 have allow_domain_literals set, the ID did not get logged in the <= line.
3395 Domain literals are now always recognized in Message-ID: header lines.
3397 4. The first argument for a ${extract expansion item is the key name or field
3398 number. Leading and trailing spaces in this item were not being ignored,
3399 causing some misleading effects.
3401 5. When deliver_drop_privilege was set, single queue runner processes started
3402 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3403 same command in the process it spins off) were not dropping privilege.
3405 6. When the daemon running as "exim" started a queue runner, it always
3406 re-executed Exim in the spun-off process. This is a waste of effort when
3407 deliver_drop_privilege is set. The new process now just calls the
3408 queue-runner function directly.
3414 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3417 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3418 ignored. Though the use of -oP was forcing the writing of a pid file, it
3419 was always written to the default place.
3421 3. If the message "no IP address found for host xxxx" is generated during
3422 incoming verification, it is now followed by identification of the incoming
3423 connection (so you can more easily find what provoked it).
3425 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3427 5. Added some features to "harden" Exim a bit more against certain attacks:
3429 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3430 be put in Local/Makefile. This is like the never_users runtime option,
3431 but it cannot be overridden. The default setting is "root".
3433 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3434 prefix string with which any file named in a -C command line option
3437 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3438 is retained for -C and -D only if the caller of Exim is root. Without
3439 it, the exim user may also use -C and -D and retain privilege.
3441 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3442 command line option is disabled.
3444 6. Macro names set by the -D option must start with an upper case letter, just
3445 like macro names defined in the configuration file.
3447 7. Added "dereference=" facility to LDAP.
3449 8. Two instances of the typo "uknown" in the source files are fixed.
3451 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3452 the Configure-Makefile script screwed up while processing it.
3454 10. Incorporated PCRE 4.4.
3456 11. The SMTP synchronization check was not operating right at the start of an
3457 SMTP session. For example, it could not catch a HELO sent before the client
3458 waited for the greeting. There is now a check for outstanding input at the
3459 point when the greeting is written. Because of the duplex, asynchronous
3460 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3461 way, but not yet received, when the check is performed.
3463 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3464 on TCP/IP sockets, because this apparently causes some broken clients to
3467 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3468 unchanged) from the Cygwin maintainer.
3470 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3471 was supported instead of testing for mbx. Effectively a typo.
3473 15. The spa authenticator server code was not checking that the input it
3474 received was valid base64.
3476 16. The debug output line for the "set" modifier in ACLs was not showing the
3477 name of the variable that was being set.
3479 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3481 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3482 lookup on its own. Something else was needed to trigger the lookup. For
3483 example, a match in host_lookup or the need for a host name in a host list.
3484 Now, if $sender_host_name is referenced and the host name has not yet been
3485 looked up, a lookup is performed. If the lookup fails, the variable remains
3486 empty, and $host_lookup_failed is set to "1".
3488 19. Added "eqi" as a case-independent comparison operator.
3490 20. The saslauthd authentication condition could segfault if neither service
3491 nor realm was specified.
3493 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3494 error message that was logged was misleading, and incoming SMTP
3495 connections were dropped. The message is now more accurate, and temporary
3496 errors are given to SMTP connections.
3498 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3499 (except RSET) with a 421 error, until QUIT is received. However, it was
3500 failing to send a response to QUIT.
3502 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3503 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3504 ACL is now run after the helo_try_verify_hosts code.
3506 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3507 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3508 case-independent, so other case variants are also recognized). Apparently
3509 some systems use these upper case variants.
3511 25. If more than two messages were waiting for the same host, and a transport
3512 filter was specified for the transport, Exim sent two messages over the
3513 same TCP/IP connection, and then failed with "socket operation on non-
3514 socket" when it tried to send the third.
3516 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3518 27. The extern definition of crypt16() in expand.c was not being excluded when
3519 the OS had its own crypt16() function.
3521 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3522 as a preferred synonym for return_size_limit, both as an option and as an
3525 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3527 30. Changed the default configuration ACL to relax the local part checking rule
3528 for addresses that are not in any local domains. For these addresses,
3529 slashes and pipe symbols are allowed within local parts, but the sequence
3530 /../ is explicitly forbidden.
3532 31. SPA server authentication was not clearing the challenge buffer before
3535 32. log_message in a "warn" ACL statement was writing to the reject log as
3536 well as to the main log, which contradicts the documentation and doesn't
3537 seem right (because no rejection is happening). So I have stopped it.
3539 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3540 However, I am unable to do any testing of this.
3542 34. Fixed an infelicity in the appendfile transport. When checking directories
3543 for a mailbox, to see if any needed to be created, it was accidentally
3544 using path names with one or more superfluous leading slashes; tracing
3545 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3547 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3548 contents are added to the log line that is written for every discarded
3549 recipient. (Previously a log_message setting was ignored.)
3551 36. The ${quote: operator now quotes the string if it is empty.
3553 37. The install script runs exim in order to find its version number. If for
3554 some reason other than non-existence or emptiness, which it checks, it
3555 could not run './exim', it was installing it with an empty version number,
3556 i.e. as "exim-". This error state is now caught, and the installation is
3559 38. An argument was missing from the function that creates an error message
3560 when Exim fails to connect to the socket for saslauthd authentication.
3561 This could cause Exim to crash, or give a corrupted message.
3563 39. Added isip, isip4, and isip6 to ${if conditions.
3565 40. The ACL variables $acl_xx are now saved with the message, and can be
3566 accessed later in routers, transports, and filters.
3568 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3569 strings in the file are not string-expanded.
3571 42. If a MAIL command specified a SIZE value that was too large to fit into an
3572 int variable, the check against message_size_limit failed. Such values are
3573 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3574 day this will have to be increased, but I don't think I want to be around
3575 when emails are that large.
3582 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3583 iconv() is not standard in FreeBSD.
3585 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3586 IPv6 enabled. The observed symptom was a segmentation fault on return from
3587 the function os_common_find_running_interfaces() in src/os.c.
3589 3. In the check_special_case() function in daemon.c I had used "errno" as an
3590 argument name, which causes warnings on some systems. This was basically a
3591 typo, since it was named "eno" in the comments!
3593 4. The code that waits for the clock to tick (at a resolution of some fraction
3594 of a second) so as to ensure message-id uniqueness was always waiting for
3595 at least one whole tick, when it could have waited for less. [This is
3596 almost certainly not relevant at current processor speeds, where it is
3597 unlikely to ever wait at all. But we try to future-proof.]
3599 5. The function that sleeps for a time interval that includes fractions of a
3600 second contained a race. It did not block SIGALRM between setting the
3601 timer, and suspending (a couple of lines later). If the interval was short
3602 and the sigsuspend() was delayed until after it had expired, the suspension
3603 never ended. On busy systems this could lead to processes getting stuck for
3606 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3607 process, before it forks any delivery processes. The open lookup caching
3608 mechanism meant that the open file or database connection was passed into
3609 the delivery process. The problem was that delivery processes always tidy
3610 up cached lookup data. This could cause a problem for the next delivery
3611 process started by the queue runner, because the external queue runner
3612 process does not know about the closure. So the next delivery process
3613 still has data in the lookup cache. In the case of a file lookup, there was
3614 no problem because closing a file descriptor in a subprocess doesn't affect
3615 the parent. However, if the lookup was caching a connection to a database,
3616 the connection was closed, and the second delivery process was likely to
3617 see errors such as "PGSQL: query failed: server closed the connection
3618 unexpectedly". The problem has been fixed by closing all cached lookups
3619 in a queue runner before running a delivery process.
3621 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3622 seem to have the "const" qualifier which it has on other OS. I've
3625 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3626 *of the same type* (client or server) with the same public name that an
3627 error should be diagnosed.
3629 9. When Exim looked up a host name for an IP address, but failed to find the
3630 original IP address when looking up the host name (a safety check), it
3631 output the message "<ip address> does not match any IP for NULL", which was
3632 confusing, to say the least. The bug was that the host name should have
3633 appeared instead of "NULL".
3635 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3636 user that is built into the binary, and the -C or -D options is used, root
3637 privilege is dropped before the configuration file is read. In addition,
3638 logging is switched to stderr instead of the normal log files. If the
3639 configuration then re-defines the Exim user, the unprivileged environment
3640 is probably not what is expected, so Exim logs a panic warning message (but
3643 However, if deliver_drop_privilege is set, the unprivileged state may well
3644 be exactly what is intended, so the warning has been cut out in that case,
3645 and Exim is allowed to try to write to its normal log files.
3651 1. smtp_return_error_details was not giving details for temporary sender
3652 or receiver verification errors.
3654 2. Diagnose a configuration error if two authenticators have the same public
3657 3. Exim used not to create the message log file for a message until the first
3658 delivery attempt. This could be confusing when incoming messages were held
3659 for policy or load reasons. The message log file is now created at the time
3660 the message is received, and an initial "Received" line is written to it.
3662 4. The automatically generated man page for command line options had a minor
3663 bug that caused no ill effects; however, a more serious problem was that
3664 the procedure for building the man page automatically didn't always
3665 operate. Consequently, release 4.20 contains an out-of-date version. This
3666 shouldn't happen again.
3668 5. When building Exim with embedded Perl support, the script that builds the
3669 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3670 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3672 6. The freeze_tell option was not being used for messages that were frozen on
3673 arrival, either by an ACL or by local_scan().
3675 7. Added the smtp_incomplete_transaction log selector.
3677 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3678 was accepting AUTH without a new EHLO.
3680 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3681 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3684 10. Logging of TCP/IP connections (when configured) now happens in the main
3685 daemon process instead of the child process, so that the TCP/IP connection
3686 count is more accurate (but it can never be perfect).
3688 11. The use of "drop" in a nested ACL was not being handled correctly in the
3689 outer ACL. Now, if condition failure induced by the nested "drop" causes
3690 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3691 or "require"), the connection is dropped.
3693 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3694 that yield "discard" can now be used with an "accept" or a "discard" verb,
3695 but an error is generated for any others (because I can't see a useful way
3696 to define what should happen).
3698 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3699 are now processed in the same way as lines in the Exim configuration file.
3700 In particular, continuation lines are supported.
3702 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3704 15. Added -ti meaning -t -i.
3706 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3707 domains, and warn by logging if others are found.
3709 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3710 returned from GIFCONF when Exim is trying to find the list of interfaces on
3711 a host. The code in os.c has been modified to copy each ifreq to an aligned
3712 structure in all cases.
3714 Also, in some cases, the returned ifreq's were being copied to a 'struct
3715 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3716 means the last couple of bytes of an IPv6 address could be chopped if the
3717 ifreq contained only a normal sockaddr (14 bytes storage).
3719 18. Named domain lists were not supported in the hosts_treat_as_local option.
3720 An entry such as +xxxx was not recognized, and was treated as a literal
3723 19. Ensure that header lines added by a DATA ACL are included in the reject log
3724 if the ACL subsequently rejects the message.
3726 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3727 MD5 (which is deprecated).
3729 21. When testing a filter file using -bf, Exim was writing a message when it
3730 took the sender from a "From " line in the message, but it was not doing so
3731 when it took $return_path from a Return-Path: header line. It now does.
3733 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3734 with a valid header line field name (a series of printing characters
3735 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3737 23. Changed "disc" in the source to "disk" to conform to the documentation and
3738 the book and for uniformity.
3740 24. Ignore Sendmail's -Ooption=value command line item.
3742 25. When execve() failed while trying to run a command in a pipe transport,
3743 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3744 could be confused with a return value of 69 from the command itself. This
3745 has been changed to 127, the value the shell returns if it is asked to run
3746 a non-existent command. The wording for the related log line suggests a
3747 non-existent command as the problem.
3749 26. If received_header_text expands to an empty string, do not add a Received:
3750 header line to the message. (Well, it adds a token one on the spool, but
3751 marks it "old" so that it doesn't get used or transmitted.)
3753 27. Installed eximstats 1.28 (addition of -nt option).
3755 28. There was no check for failure on the call to getsockname() in the daemon
3756 code. This can fail if there is a shortage of resources on the system, with
3757 ENOMEM, for example. A temporary error is now given on failure.
3759 29. Contrary to the C standard, it seems that in some environments, the
3760 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3761 program. Exim now does this explicitly; it affects the formatting of
3762 timestamps using strftime().
3764 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3765 complaints. I've now initialized all the variables, to avoid this.
3767 32. Header lines added by a system filter were not being "seen" during
3768 transport-time rewrites.
3770 33. The info_callback() function passed to OpenSSL is set up with type void
3771 (*)(SSL *, int, int), as described somewhere. However, when calling the
3772 function (actually a macro) that sets it up, the type void(*)() is
3773 expected. I've put in a cast to prevent warnings from picky compilers.
3775 34. If a DNS black list lookup found a CNAME record, but there were no A
3776 records associated with the domain it pointed at, Exim crashed.
3778 35. If a DNS black list lookup returned more than one A record, Exim ignored
3779 all but the first. It now scans all returned addresses if a particular IP
3780 value is being sought. In this situation, the contents of the
3781 $dnslist_value variable are a list of all the addresses, separated by a
3784 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3785 to accept a host name and all its aliases if the forward lookup for any of
3786 them yielded the IP address of the incoming connection. Now it accepts only
3787 those names whose forward lookup yields the correct IP address. Any other
3788 names are discarded. This closes a loophole whereby a rogue DNS
3789 administrator could create reverse DNS records to break through a
3790 wildcarded host restriction in an ACL.
3792 37. If a user filter or a system filter that ran in a subprocess used any of
3793 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3794 the wrong values were passed to the pipe command ($thisaddress had the
3795 value of $0, $0 had the value of $1, etc). This bug was introduced by
3796 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3798 38. Improved the line breaking for long SMTP error messages from ACLs.
3799 Previously, if there was no break point between 40 and 75 characters, Exim
3800 left the rest of the message alone. Two changes have been made: (a) I've
3801 reduced the minimum length to 35 characters; (b) if it can't find a break
3802 point between 35 and 75 characters, it looks ahead and uses the first one
3803 that it finds. This may give the occasional overlong line, but at least the
3804 remaining text gets split now.
3806 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3807 file descriptors might be low, and that setting 1000 would always raise it.
3808 It turns out that in some environments, the limit is already over 1000 and
3809 that lowering it causes trouble. So now Exim takes care not to decrease it.
3811 40. When delivering a message, the value of $return_path is set to $sender_
3812 address at the start of routing (routers may change the value). By an
3813 oversight, this default was not being set up when an address was tested by
3814 -bt or -bv, which affected the outcome if any router or filter referred to
3817 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3818 log only when "message" or "log_message" are set. However, if one of the
3819 conditions was an address verification, or a call to a nested ACL, the
3820 messages generated by the underlying test were being passed through. This
3821 no longer happens. The underlying message is available in $acl_verify_
3822 message for both "message" and "log_message" expansions, so it can be
3823 passed through if needed.
3825 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3826 new expansion $bh_ to give the encoded byte string without charset
3827 translation. Translation happens only if iconv() is available; HAVE_ICONV
3828 indicates this at build time. HEADERS_CHARSET gives the charset to
3829 translate to; headers_charset can change it in the configuration, and
3830 "headers charset" can change it in an individual filter file.
3832 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3833 that creates RFC 2047 encoded "words" labels them as that charset instead
3834 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3835 expansion operator; (ii) when Exim creates a From: line for a local
3836 message; (iii) when a header line is rewritten to include a "phrase" part.
3838 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3839 buggy, causing it to skip the first lines of messages whose message ID
3840 ended in 'D'. This would not have bitten before Exim release 4.14, because
3841 message IDs were unlikely to end in 'D' before then. The effect was to have
3842 incorrect size information for certain domains.
3844 45. #include "config.h" was missing at the start of the crypt16.c module. This
3845 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3848 46. If there was a timeout during a "random" callout check, Exim treated it as
3849 a failure of the random address, and carried on sending RSET and the real
3850 address. If the delay was just some slowness somewhere, the response to the
3851 original RCPT would be taken as a response to RSET and so on, causing
3852 mayhem of various kinds.
3854 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3855 when I implemented it. It didn't allow for the fact that some option values
3856 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3857 the right test for positive values.
3859 48. Domain names in DNS records are case-independent. Exim always looks them up
3860 in lower case. Some resolvers return domain names in exactly the case they
3861 appear in the zone file, that is, they may contain uppercase letters. Not
3862 all resolvers do this - some return always lower case. Exim was treating a
3863 change of case by a resolver as a change of domain, similar to a widening
3864 of a domain abbreviation. This triggered its re-routing code and so it was
3865 trying to route what was effectively the same domain again. This normally
3866 caused routing to fail (because the router wouldn't handle the domain
3867 twice). Now Exim checks for this case specially, and just changes the
3868 casing of the domain that it ultimately uses when it transmits the message
3871 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3874 50. If a filter generated a file delivery with a non-absolute name (possible if
3875 no home directory exists for the router), the forbid_file option was not
3878 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3879 the existing equality matching.
3881 52. Exim was using ints instead of ino_t variables in some places where it was
3882 dealing with inode numbers.
3884 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3885 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3886 called TMPDIR, and if it finds it is different, it changes its value.
3888 54. The smtp_printf() function is now made available to local_scan() so
3889 additional output lines can be written before returning. There is also an
3890 smtp_fflush() function to enable the detection of a dropped connection.
3891 The variables smtp_input and smtp_batched_input are exported to
3894 55. Changed the default runtime configuration: the message "Unknown user"
3895 has been removed from the ACL, and instead placed on the localuser router,
3896 using the cannot_route_message feature. This means that any verification
3897 failures that generate their own messages won't get overridden. Similarly,
3898 the "Unrouteable address" message that was in the ACL for unverifiable
3899 relay addresses has also been removed.
3901 56. Added hosts_avoid_esmtp to the smtp transport.
3903 57. The exicyclog script was not checking for the esoteric option
3904 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3905 will work only if exicyclog is run under the appropriate euid.
3907 58. Following a discussion on the list, the rules by which Exim recognises line
3908 endings on incoming messages have been changed. The -dropcr and drop_cr
3909 options are now no-ops, retained only for backwards compatibility. The
3910 following line terminators are recognized: LF CRLF CR. However, special
3911 processing applies to CR:
3913 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3914 nor a local message in the state where . is a terminator.
3916 (ii) If a bare CR is encountered in a header line, an extra space is added
3917 after the line terminator so as not to end the header. The reasoning
3918 behind this is that bare CRs in header lines are most likely either
3919 to be mistakes, or people trying to play silly games.
3921 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3922 was being incorrectly given as 18 bytes larger than it should have been.
3923 This is a VOB (very old bug).
3925 60. This may never have affected anything current, but just in case it has:
3926 When the local host is found other than at the start of a list of hosts,
3927 the local host, those with the same MX, and any that follow, are discarded.
3928 When the list in question was part of a longer list of hosts, the following
3929 hosts (not currently being processed) were also being discarded. This no
3930 longer happens. I'm not sure if this situation could ever has previously
3933 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3936 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3937 header. This is something that is recommended in a new Internet Draft, and
3938 is something that is documented as being done by Sendmail. There are two
3939 possible values. For messages generated by the autoreply transport, Exim
3942 Auto-Submitted: auto-replied
3944 whereas for all other generated messages (e.g. bounces) it adds
3946 Auto-Submitted: auto-generated
3948 63. The "personal" condition in filters now includes a test for the
3949 Auto-Submitted: header. If it contains the string "auto-" the message it
3950 not considered personal.
3952 64. Added rcpt_include_affixes as a generic transport option.
3954 65. Added queue_only_override (default true).
3956 66. Added the syslog_duplication option.
3958 67. If what should have been the first header line of a message consisted of
3959 a space followed by a colon, Exim was mis-interpreting it as a header line.
3960 It isn't of course - it is syntactically invalid and should therefore be
3961 treated as the start of the message body. The misbehaviour could have
3962 caused a number of strange effects, including loss of data in subsequent
3963 header lines, and spool format errors.
3965 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3966 client host had authenticated. This control can now be exercised by an ACL
3967 for more flexibility.
3969 69. By default, callouts do not happen when testing with -bh. There is now a
3970 variant, -bhc, which does actually run the callout code, including
3971 consulting and updating the callout cache.
3973 70. Added support for saslauthd authentication, courtesy of Alexander
3976 71. If statvfs() failed on the spool or log directories while checking their
3977 size for availability, Exim confusingly gave the error "space shortage".
3978 Furthermore, in debugging mode it crashed with a floating point exception.
3979 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3980 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3981 serious problem, Exim now writes to the main and panic logs when this
3982 happens, with details of the failure. It then refuses to accept the
3983 incoming message, giving the message "spool directory problem" or "log
3984 directory problem" with a 421 code for SMTP messages.
3986 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3987 0, 1, and 2 exist, because some OS complain for execs without them (see
3988 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3989 descriptors. However, the code omitted to check that the open succeeded,
3990 causing mysterious errors if for some reason the permissions on /dev/null
3991 got screwed. Now Exim writes a message to the main and panic logs, and
3992 bombs out if it can't open /dev/null.
3994 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3995 interact so that it is all more flexible. It is supposed to remain
3996 backwards compatible. Also added extra_local_interfaces.
3998 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3999 to bomb out with an assertion failure - to the client this appears as a
4000 connection drop. This problem occurs in the part of the code that was taken
4001 from the Samba project. Fortunately, the assertion is in a very simple
4002 function, so I have fixed this by reproducing the function inline in the
4003 one place where it is called, and arranging for authentication to fail
4004 instead of killing the process with assert().
4006 75. The SPA client code was not working when the server requested OEM rather
4007 than Unicode encoding.
4009 76. Added code to make require_files with a specific uid setting more usable in
4010 the case where statting the file as root fails - usually a non-root-mounted
4011 NFS file system. When this happens and the failure is EACCES, Exim now
4012 forks a subprocess and does the per-uid checking as the relevant uid.
4014 77. Added process_log_path.
4016 78. If log_file_path was not explicitly set, a setting of check_log_space or
4017 check_log_inodes was ignored.
4019 79. If a space check for the spool or log partitions fails, the incident is now
4020 logged. Of course, in the latter case the data may get lost...
4022 80. Added the %p formatting code to string_format() so that it can be used to
4023 print addresses in debug_print(). Adjusted all the address printing in the
4024 debugging in store.c to use %p rather than %d.
4026 81. There was a concern that a line of code in smtp_in.c could overflow a
4027 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4028 initially expressed, the concern was not well-founded, because trailing
4029 spaces are removed early. However, if the trailing spaces were followed by
4030 a NULL, they did not get removed, so the overflow was possible. Two fixes
4033 (a) I re-wrote the offending code in a cleaner fashion.
4034 (b) If an incoming SMTP command contains a NULL character, it is rejected
4037 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4038 runs initgroups(), so that if the Exim user is in any additional groups,
4039 they will be used during message reception.
4045 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob