1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.372 2006/07/07 14:36:04 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 SC/01 Use a glob alias rather than an array ref in eximstats generated
10 parser. This improves both readability and performance.
12 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
13 Don't display local sender or destination tables in eximstats unless
14 there is data to show.
15 Added average volumes into the eximstats top table text output.
17 SC/03 Collect data on the number of addresses (recipients) as well
18 as the number of messages in eximstats.
20 TF/01 Correct an error in the documentation for the redirect router. Exim
21 does (usually) call initgroups() when daemonizing.
23 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
24 with consistent privilege compared to when running as a daemon.
26 TF/03 Note in the spec that $authenticated_id is not set for local
27 submissions from trusted users.
29 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
30 Thanks to Dean Brooks <dean@iglou.com> for the patch.
32 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
33 by adding some example configuration directives to the default
34 configuration file. A little bit of work is required to uncomment the
35 directives and define how usernames and passwords are checked, but
36 there is now a framework to start from.
38 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
39 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
40 without this. I don't know how relevant this is to other LDAP libraries.
42 PH/02 Add the verb name to the "unknown ACL verb" error.
44 PH/03 Magnus Holmgren's patch for filter_prepend_home.
46 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
48 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
49 directory not expanded when it should be if an expanded home directory
50 was set for the address (which is overridden by the transport).
52 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
55 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
56 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
57 because it is too late at that time, and has no effect.
59 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
60 security issue with \' (bugzilla #107). I could not use the
61 PQescapeStringConn() function, because it needs a PGconn value as one of
64 PH/08 When testing addresses using -bt, indicate those final addresses that
65 are duplicates that would not cause an additional delivery. At least one
66 person was confused, thinking that -bt output corresponded to deliveries.
67 (Suppressing duplicates isn't a good idea as you lose the information
68 about possibly different redirections that led to the duplicates.)
70 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
71 systems where poll() doesn't work, in particular OS X.
73 PH/10 Added more information to debugging output for retry time not reached.
75 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
76 operations in malware.c.
78 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
81 PH/13 If write_rejectlog was set false when logging was sent to syslog with
82 syslog_duplication set false, log lines that would normally be written
83 both the the main log and to the reject log were not written to syslog at
86 PH/14 In the default configuration, change the use of "message" in ACL warn
87 statements to "add_header".
89 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
90 not followed by a command (e.g. "seen endif").
96 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
97 other effects) broke the use of negated acl sub-conditions.
99 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
102 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
103 "Deny" causes Exim to reject the incoming connection with a 554 error.
104 Unfortunately, if there is a major crisis, such as a disk failure,
105 tcp-wrappers gives "deny", whereas what one would like would be some
106 kind of temporary error. A kludge has been added to help with this.
107 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
108 554 error is used if errno is still zero or contains ENOENT (which occurs
109 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
112 PH/03 Add -lutil to the default FreeBSD LIBS setting.
114 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
115 errors. Otherwise a message that provokes a temporary error (when other
116 messages do not) can cause a whole host to time out.
118 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
119 addresses were routed directly to files or pipes from a redirect router.
120 File deliveries just didn't batch; pipe deliveries might have suffered
123 PH/06 A failure to get a lock for a hints database would erroneously always say
124 "Failed to get write lock", even when it was really a read lock.
126 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
127 of 0600. This has been changed to use the value of the lockfile_mode
128 option (which defaults to 0600).
130 PH/08 Applied small patch from the Sieve maintainer.
132 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
133 folder from quota calculations, a direct delivery into this folder messed
134 up the contents of the maildirsize file. This was because the regex was
135 used only to exclude .Trash (or whatever) when the size of the mailbox
136 was calculated. There was no check that a delivery was happening into an
137 excluded directory. This bug has been fixed by ignoring all quota
138 processing for deliveries into excluded directories.
140 PH/10 Added the maildirfolder_create_regex option to appendfile.
146 PH/01 The code for finding all the local interface addresses on a FreeBSD
147 system running IPv6 was broken. This may well have applied to all BSD
148 systems, as well as to others that have similar system calls. The broken
149 code found IPv4 interfaces correctly, but gave incorrect values for the
150 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
151 that it would not match correctly against @[] and not recognize the IPv6
154 PH/02 The ipliteral router was not recognizing addresses of the form user@
155 [ipv6:....] because it didn't know about the "ipv6:" prefix.
157 PH/03 Added disable_ipv6.
159 PH/04 Changed $reply_address to use the raw form of the headers instead of the
160 decoded form, because it is most often used to construct To: headers
161 lines in autoreplies, and the decoded form may well be syntactically
162 invalid. However, $reply_address has leading white space removed, and all
163 newlines turned into spaces so that the autoreply transport does not
166 PH/05 If group was specified without a user on a router, and no group or user
167 was specified on a transport, the group from the router was ignored.
169 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
170 for visible compile-time settings that can be used to change these
171 numbers, for those that want even more. Backwards compatibility with old
172 spool files has been maintained. However, going back to a previous Exim
173 release will lost any variables that are in spool files.
175 PH/07 Two small changes when running in the test harness: increase delay when
176 passing a TCP/IP connection to a new process, in case the original
177 process has to generate a bounce, and remove special handling of
178 127.0.0.2 (sic), which is no longer necessary.
180 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
181 be the same on different OS.
183 PH/09 Moved a debug statement in filter processing to avoid a race problem when
186 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
187 whether --show-vars was specified or not
189 JJ/02 exipick: Added support for new ACL variable spool format introduced
192 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
193 syntactically invalid From: or Reply-to: line, and a filter used this to
194 generate an autoreply, and therefore failed to obtain an address for the
195 autoreply, Exim could try to deliver to a non-existent relative file
196 name, causing unrelated and misleading errors. What now happens is that
197 it logs this as a hard delivery error, but does not attempt to create a
200 PH/11 The exinext utility has a -C option for testing purposes, but although
201 the given file was scanned by exinext itself; it wasn't being passed on
202 when Exim was called.
204 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
205 an end-of-file indication when reading a command response.
207 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
208 compiled. In many other places in Exim, IPv6 addresses are always
209 recognized, so I have changed this. It also means that IPv4 domain
210 literals of the form [IPV4:n.n.n.n] are now always recognized.
212 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
213 used if the router is not running as root, for example, when verifying at
214 ACL time, or when using -bh. The debugging output from this situation was
215 non-existent - all you got was a failure to exec. I have made two
218 (a) Failures to set uid/gid, the current directory, or a process leader
219 in a subprocess such as that created by queryprogram now generate
220 suitable debugging ouput when -d is set.
222 (b) The queryprogram router detects when it is not running as root,
223 outputs suitable debugging information if -d is set, and then runs
224 the subprocess without attempting to change uid/gid.
226 PH/15 Minor change to Makefile for building test_host (undocumented testing
229 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
230 additional section of a DNS packet that returns MX or SRV records.
231 Instead, it always explicitly searches for A/AAAA records. This avoids
232 major problems that occur when a DNS server includes only records of one
233 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
234 fixed another bug: if SRV records were looked up and the corresponding
235 address records were *not* found in the additional section, the port
236 values from the SRV records were lost.
238 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
239 using the correct key (the original address) when searching the retry
240 rules in order to find which one to use for generating the retry hint.
242 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
243 adding the default one. Similarly, if it contains a Reply-To: header, the
244 errors_reply_to option, if set, is not used.
246 PH/19 When calculating a retry time, Exim used to measure the "time since
247 failure" by looking at the "first failed" field in the retry record. Now
248 it does not use this if it is later than than the arrival time of the
249 message. Instead it uses the arrival time. This makes for better
250 behaviour in cases where some deliveries succeed, thus re-setting the
251 "first failed" field. An example is a quota failure for a huge message
252 when small messages continue to be delivered. Without this change, the
253 "time since failure" will always be short, possible causing more frequent
254 delivery attempts for the huge message than are intended.
255 [Note: This change was subsequently modified - see PH/04 for 4.62.]
257 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
258 $1, $2, $3) because the numerical variables can be reset during some
259 expansion items (e.g. "match"), thereby losing the authentication data.
261 PH/21 Make -bV show the size of off_t variables so that the test suite can
262 decide whether to run tests for quotas > 2G.
264 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
265 mailbox_size, and mailbox_filecount in the appendfile transport. If a
266 filecount value is greater than 2G or if a quota value is greater than 2G
267 on a system where the size of off_t is not greater than 4, a panic error
270 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
271 never match. The debug and -bh output now contains an explicit error
272 message indicating a malformed IPv4 address or mask.
274 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
275 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
278 PH/25 Do not write to syslog when running in the test harness. The only
279 occasion when this arises is a failure to open the main or panic logs
280 (for which there is an explicit test).
282 PH/26 Added the /no_tell option to "control=freeze".
284 PH/27 If a host name lookup failed very early in a connection, for example, if
285 the IP address matched host_lookup and the reverse lookup yielded a name
286 that did not have a forward lookup, an error message of the form "no IP
287 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
288 could be logged. Now it outputs the IP address instead of "NULL".
290 PH/28 An enabling patch from MH: add new function child_open_exim2() which
291 allows the sender and the authenticated sender to be set when
292 submitting a message from within Exim. Since child_open_exim() is
293 documented for local_scan(), the new function should be too.
295 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
296 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
297 results in an empty string is now treated as unset.
299 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
301 PH/31 Added sender_verify_fail logging option.
303 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
304 needed by RFC 822 but not by RFC 2822 was commented out. I have now
305 tidied the source and removed it altogether.
307 PH/33 When a queue run was abandoned because the load average was too high, a
308 log line was always written; now it is written only if the queue_run log
309 selector is set. In addition, the log line for abandonment now contains
310 information about the queue run such as the pid. This is always present
311 in "start" and "stop" lines but was omitted from the "abandon" line.
313 PH/34 Omit spaces between a header name and the colon in the error message that
314 is given when verify = headers_syntax fails (if there are lots of them,
315 the message gets confusing).
317 PH/35 Change the default for dns_check_names_pattern to allow slashes within
318 names, as there are now some PTR records that contain slashes. This check
319 is only to protect against broken name servers that fall over on strange
320 characters, so the fact that it applies to all lookups doesn't matter.
322 PH/36 Now that the new test suite is complete, we can remove some of the
323 special code in Exim that was needed for the old test suite. For example,
324 sorting DNS records because real resolvers return them in an arbitrary
325 order. The new test suite's fake resolver always returns records in the
328 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
329 bounces) except when queue_only is set, to avoid logging races between
330 the different processes.
332 PH/38 Panic-die if .include specifies a non-absolute path.
334 PH/39 A tweak to the "H" retry rule from its user.
336 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
337 a label. They prevented compilation on older perls.
339 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
340 a warning to be raised on newish perls.
342 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
343 on queue. Changes to match documented behaviour of showing count of
344 messages matching specified criteria.
346 PH/40 Changed the default ident timeout from 30s to 5s.
348 PH/41 Added support for the use of login_cap features, on those BSD systems
349 that have them, for controlling the resources used by pipe deliveries.
351 PH/42 The content-scanning code uses fopen() to create files in which to put
352 message data. Previously it was not paying any attention to the mode of
353 the files. Exim runs with umask(0) because the rest of the code creates
354 files with open(), and sets the required mode explicitly. Thus, these
355 files were ending up world-writeable. This was not a big issue, because,
356 being within the spool directory, they were not world-accessible. I have
357 created a function called modefopen, which takes an additional mode
358 argument. It sets umask(777), creates the file, chmods it to the required
359 mode, then resets the umask. All the relevant calls to fopen() in the
360 content scanning code have been changed to use this function.
362 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
363 to 24 hours. This avoids potential overflow problems when processing G
364 and H retry rules. I suspect nobody ever tinkers with this value.
366 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
368 PH/45 When the plaintext authenticator is running as a client, the server's
369 challenges are checked to ensure they are valid base64 strings. By
370 default, the authentication attempt is cancelled if an invalid string is
371 received. Setting client_ignore_invalid_base64 true ignores these errors.
372 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
373 they are received. Thus, the responses can be made to depend on the
374 challenges. If an invalid string is ignored, an empty string is placed in
377 PH/46 Messages that are created by the autoreply transport now contains a
378 References: header, in accordance with RFCs 2822 and 3834.
380 PH/47 Added authenticated_sender_force to the smtp transport.
382 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
384 PH/49 Installed latest patch from the Sieve maintainer.
386 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
387 was also set, the appendfile transport was unnecessarily scanning a
388 directory of message files (e.g. for maildir delivery) to find the count
389 of files (along with the size), even though it did not need this
390 information. It now does the scan only if it needs to find either the
391 size of the count of files.
393 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
395 PH/52 Two bugs concerned with error handling when the smtp transport is
398 (i) Exim was not creating retry information for temporary errors given
399 for individual recipients after the DATA command when the smtp transport
400 was used in LMTP mode. This meant that they could be retried too
401 frequently, and not timed out correctly.
403 (ii) Exim was setting the flag that allows error details to be returned
404 for LMTP errors on RCPT commands, but not for LMTP errors for individual
405 recipients that were returned after the DATA command.
407 PH/53 This is related to PH/52, but is more general: for any failing address,
408 when detailed error information was permitted to be returned to the
409 sender, but the error was temporary, then after the final timeout, only
410 "retry timeout exceeded" was returned. Now it returns the full error as
411 well as "retry timeout exceeded".
413 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
414 do this, and (what is worse) MTAs that accept it.
416 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
417 will now be deprecated.
419 PH/56 New os.c-cygwin from the Cygwin maintainer.
421 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
422 formats (previously only available in exim formats via -bpr, -bpru,
423 and -bpra. Now also available in native and exiqgrep formats)
425 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
426 with very large, slow to parse queues
428 JJ/08 exipick: added ! as generic prefix to negate any criteria format
430 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
432 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
433 responses to authentication challenges, though it was showing the
434 challenges; (ii) I've removed the CR characters from the debug output for
437 PH/58 Allow for the insertion of a newline as well as a space when a string
438 is turned into more than one encoded-word during RFC 2047 encoding. The
439 Sieve code now uses this.
441 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
442 data_4xx, lost_connection, tls_required.
444 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
445 message was being sent as an SMTP response.
447 PH/61 Add -l and -k options to exicyclog.
449 PH/62 When verifying, if an address was redirected to one new address, so that
450 verification continued, and the new address failed or deferred after
451 having set something in $address_data, the value of $address_data was not
452 passed back to the ACL. This was different to the case when no
453 redirection occurred. The value is now passed back in both cases.
455 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
456 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
457 use login_cap.h, so on its own it isn't the distinguishing feature. The
458 new name refers directly to the setclassresources() function.
460 PH/65 Added configuration files for NetBSD3.
462 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
464 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
465 is preferred over IPv4.
467 PH/68 The bounce_return_message and bounce_return_body options were not being
468 honoured for bounces generated during the reception of non-SMTP messages.
469 In particular, this applied to messages rejected by the ACL. This bug has
470 been fixed. However, if bounce_return_message is true and bounce_return_
471 body is false, the headers that are returned for a non-SMTP message
472 include only those that have been read before the error was detected.
473 (In the case of an ACL rejection, they have all been read.)
475 PH/69 The HTML version of the specification is now built in a directory called
476 spec_html instead of spec.html, because the latter looks like a path with
477 a MIME-type, and this confuses some software.
479 PH/70 Catch two compiler warnings in sieve.c.
481 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
482 function verify_get_ident() calls ip_connect() to connect a socket, but
483 if the "connect()" function timed out, ip_connect() used to close the
484 socket. However, verify_get_ident() also closes the socket later, and in
485 between Exim writes to the log, which may get opened at this point. When
486 the socket was closed in ip_connect(), the log could get the same file
487 descriptor number as the socket. This naturally causes chaos. The fix is
488 not to close the socket in ip_connect(); the socket should be closed by
489 the function that creates it. There was only one place in the code where
490 this was missing, in the iplookup router, which I don't think anybody now
491 uses, but I've fixed it anyway.
493 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
494 well as to direct DNS lookups. Otherwise the handling of names in host
495 lists is inconsistent and therefore confusing.
501 PH/01 Two changes to the default runtime configuration:
503 (1) Move the checks for relay_from_hosts and authenticated clients from
504 after to before the (commented out) DNS black list checks.
506 (2) Add control=submission to the relay_from_hosts and authenticated
507 clients checks, on the grounds that messages accepted by these
508 statements are most likely to be submissions.
510 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
512 (1) Generate an error if the third argument for the ${prvs expansion is
515 (2) Treat a missing third argument of ${prvscheck as if it were an empty
518 (3) Reset the variables that are obtained from the first argument of
519 ${prvscheck and used in the second argument before leaving the code,
520 because their memory is reclaimed, so using them afterwards may do
523 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
524 one (it's much easier than Tom thought :-).
526 (5) Because of (4), we can now allow for the use of $prvscheck_result
527 inside the third argument.
529 PH/03 For some reason, the default setting of PATH when running a command from
530 a pipe transport was just "/usr/bin". I have changed it to
533 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
534 anything to be listed in the output from -bV.
536 PH/05 When a filter generated an autoreply, the entire To: header line was
537 quoted in the delivery log line, like this:
539 => >A.N.Other <ano@some.domain> <original@ddress> ...
541 This has been changed so that it extracts the operative address. There
542 may be more than one such address. If so, they are comma-separated, like
545 => >ano@some.domain,ona@other.domain <original@ddress> ...
547 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
548 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
549 not being looked up in the rDNS to get a host name, Exim was showing the
550 IP address twice in Received: lines, even though the IP addresses were
551 identical. For example:
553 Received: from [1.2.3.4] (helo=[1.2.3.4])
555 However, if the real host name was known, it was omitting the HELO data
556 if it matched the actual IP address. This has been tidied up so that it
557 doesn't show the same IP address twice.
559 PH/07 When both +timestamp and +memory debugging was on, the value given by
560 $tod_xxx expansions could be wrong, because the tod_stamp() function was
561 called by the debug printing, thereby overwriting the timestamp buffer.
562 Debugging no longer uses the tod_stamp() function when +timestamp is set.
564 PH/08 When the original message was included in an autoreply transport, it
565 always said "this is a copy of the message, including all the headers",
566 even if body_only or headers_only was set. It now gives an appropriate
569 PH/09 Applied a patch from the Sieve maintainer which:
571 o fixes some comments
572 o adds the (disabled) notify extension core
573 o adds some debug output for the result of if/elsif tests
574 o points to the current vacation draft in the documentation
575 and documents the missing references header update
579 o fixes a bug in processing the envelope test (when testing
580 multiple envelope elements, the last element determinted the
583 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
584 Electronic Mail") by including:
586 Auto-submitted: auto-generated
588 in the messages that it generates (bounce messages and others, such as
589 warnings). In the case of bounce messages for non-SMTP mesages, there was
590 also a typo: it was using "Auto_submitted" (underscore instead of
591 hyphen). Since every message generated by Exim is necessarily in response
592 to another message, thes have all been changed to:
594 Auto-Submitted: auto-replied
596 in accordance with these statements in the RFC:
598 The auto-replied keyword:
600 - SHOULD be used on messages sent in direct response to another
601 message by an automatic process,
603 - MUST NOT be used on manually-generated messages,
605 - MAY be used on Delivery Status Notifications (DSNs) and Message
606 Disposition Notifications (MDNs),
608 - MUST NOT be used on messages generated by automatic or periodic
609 processes, except for messages which are automatic responses to
612 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
613 to the default Received: header definition.
615 PH/12 Added log selector acl_warn_skipped (default on).
617 PH/13 After a successful wildlsearch lookup, discard the values of numeric
618 variables because (a) they are in the wrong storage pool and (b) even if
619 they were copied, it wouldn't work properly because of the caching.
621 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
622 checking when decoding. Apparently there are clients that generate
623 overlong encoded strings. Why am I not surprised?
625 PH/15 If the first argument of "${if match_address" was not empty, but did not
626 contain an "@" character, Exim crashed. Now it writes a panic log message
627 and treats the condition as false.
629 PH/16 In autoreply, treat an empty string for "once" the same as unset.
631 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
632 extension "envelope-auth". The code is finished and in agreement with
633 other implementations, but there is no documentation so far and in fact,
634 nobody wrote the draft yet. This extension is currently #undef'ed, thus
635 not changing the active code.
637 Print executed "if" and "elsif" statements when debugging is used. This
638 helps a great deal to understand what a filter does.
640 Document more things not specified clearly in RFC3028. I had all this
641 sorted out, when out of a sudden new issues came to my mind. Oops."
643 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
646 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
647 canonical form (as documented). However, after a host name lookup from
648 the IP address, check_host() was doing a simple string comparison with
649 addresses acquired from the DNS when checking that the found name did
650 have the original IP as one of its addresses. Since any found IPv6
651 addresses are likely to be in abbreviated form, the comparison could
652 fail. Luckily, there already exists a function for doing the comparison
653 by converting both addresses to binary, so now that is used instead of
656 PH/20 There was another similar case to PH/19, when a complete host name was
657 given in a host list; looking up its IP address could give an abbreviated
658 form, whereas the current host's name might or might not be abbreviated.
659 The same fix has been applied.
665 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
666 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
669 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
670 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
672 PH/03 Typo: missing ".o" in src/pcre/Makefile.
674 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
675 header line, restrict the check to what is listed in RFCs 2369 and 2929.
676 Also, for "Auto-Submitted", treat anything other than "no" as
677 non-personal, in accordance with RFC 3834. (Previously it treated
678 anything starting "auto-" as non-personal.)
680 TF/01 The control=submission/name=... option had a problem with syntax
681 errors if the name included a slash character. The /name= option
682 now slurps the rest of the string, so it can include any characters
683 but it must come last in the list of options (after /sender_retain
686 PH/05 Some modifications to the interface to the fake nameserver for the new
694 TK/01 Added the "success_on_redirect" address verification option. See
695 NewStuff for rationale and an example.
697 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
699 PH/02 Patch to exigrep to allow it to work on syslog lines.
701 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
702 fread() to skip over the body file's header line, because in Cygwin the
703 header line is locked and is inaccessible.
705 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
706 co-exist for some time) to make it clear that it is the Exim ID that is
707 referenced, not the Message-ID: header line.
709 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
710 string_format() function, because snprintf() does not exist on all
713 PH/06 The use of forbid_filter_existstest now also locks out the use of the
714 ${stat: expansion item.
716 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
717 protocol synchronization error", to keep the pedants happy.
719 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
720 well as for IRIX systems, when gcc is being used. See the host.c source
723 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
725 PH/10 Named domain lists were not working if used in a queue_smtp_domains
728 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
729 transport and to the smtp transport in LMTP mode.
731 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
733 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
734 run a filter in a subprocess. This could lead to confusion in subsequent
735 lookups in the parent process. There should also be a search_tidyup() at
736 the end of the subprocess.
738 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
739 only if the host matched helo_try_verify_hosts, which caused the
740 verification to occur when the EHLO/HELO command was issued. The ACL just
741 tested the remembered result. Now, if a previous verification attempt has
742 not happened, "verify = helo" does it there and then.
744 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
746 TK/03 Fix log output including CR from clamd.
748 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
749 exist provoked a memory error which could cause a segfault.
751 PH/15 Installed PCRE 6.2
753 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
755 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
756 of the problem. Specifically, suggested +O2 rather than +O1 for the
759 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
761 PH/20 If a delivery was routed to a non-standard port by means of an SRV
762 record, the port was not correctly logged when the outgoing_port log
763 selector was set (it logged the transort's default port).
765 PH/21 Added support for host-specific ports to manualroute, queryprogram,
766 fallback_hosts, and "hosts" in the smtp transport.
768 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
769 host errors such as "Connection refused".
771 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
772 authentication with radiusclient 0.4.9:
774 - Error returned from rc_read_config was caught wrongly
775 - Username/password not passed on to radius server due to wrong length.
777 The presumption is that some radiusclient API changes for 4.51/PH/17
778 were not taken care of correctly. The code is still untested by me (my
779 Linux distribution still has 0.3.2 of radiusclient), but it was
780 contributed by a Radius user.
782 PH/24 When doing a callout, the value of $domain wasn't set correctly when
783 expanding the "port" option of the smtp transport.
785 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
786 while reading a MIME header. Thanks to Tom Hughes for a patch.
788 PH/24 Include config.h inside local_scan.h so that configuration settings are
791 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
792 that in an ACL for RCPT (for example), you can examine exactly what was
795 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
796 commands, but it was not correctly comparing the address with the actual
797 client host address. Thus, it would show the EHLO address in Received:
798 header lines when this was not necessary.
800 PH/27 Added the % operator to ${eval:}.
802 PH/28 Exim tries to create and chdir to its spool directory when it starts;
803 it should be ignoring failures (because with -C, for example, it has lost
804 privilege). It wasn't ignoring creation failures other than "already
807 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
810 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
811 because an input line was too long, either on its own, or by virtue of
812 too many continuations, the temporary file was not being removed, and the
813 return code was incorrect.
815 PH/31 Missing "BOOL" in function definition in filtertest.c.
817 PH/32 Applied Sieve patches from the maintainer.
819 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
821 PH/33 Added "verify = not_blind".
823 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
824 Local/Makefile (with some defaults set). These are used in built scripts
825 such as exicyclog, but they have never been used in the exim_install
826 script (though there are many overriding facilities there). I have
827 arranged that the exim_install script now takes note of these two
830 PH/35 Installed configuration files for Dragonfly.
832 PH/36 When a locally submitted message by a trusted user did not contain a
833 From: header, and the sender address was obtained from -f or from an SMTP
834 MAIL command, and the trusted user did not use -F to supply a sender
835 name, $originator_name was incorrectly used when constructing a From:
836 header. Furthermore, $originator_name was used for submission mode
837 messages from external hosts without From: headers in a similar way,
838 which is clearly wrong.
840 PH/37 Added control=suppress_local_fixups.
842 PH/38 When log_selector = +received_sender was set, and the addition of the
843 sender made the log line's construction buffer exactly full, or one byte
844 less than full, an overflow happened when the terminating "\n" was
847 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
848 when the result of a list match is failure because a DNS lookup failed.
850 PH/40 RM_COMMAND is now used in the building process.
852 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
853 the "build-* directories that it finds.
855 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
856 address in a domain literal was a prefix of an interface address.
858 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
859 when verifying a sender address, unless rewrite_headers is false.
861 PH/44 Wrote a long comment about why errors_to addresses are verified as
862 recipients, not senders.
864 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
865 the ratelimit ACL was added.
867 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
869 PH/46 Added extra information about PostgreSQL errors to the error string.
871 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
872 avoiding the need to install special zones in a real server. This is
873 backwards compatible; if it can't find the fake resolver, it drops back.
874 Thus, both old and new test suites can be run.
876 TF/02 Added util/ratelimit.pl
878 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
879 clock is set back in time.
881 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
882 Candler <B.Candler@pobox.com>.
884 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
885 OK for addresses that are the result of redirections.
887 PH/48 A number of further additions for the benefit of the new test suite,
888 including a fake gethostbyname() that interfaces to the fake DNS resolver
891 TF/06 The fix for widen_domains has also been applied to qualify_single and
892 search_parents which are the other dnslookup options that can cause
895 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
898 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
900 TF/07 Exim produced the error message "an SRV record indicated no SMTP
901 service" if it encountered an MX record with an empty target hostname.
902 The message is now "an MX or SRV record indicated no SMTP service".
904 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
905 if the DNS of the sending site is misconfigured. This is quite a
906 common situation. This change restores the behaviour of treating a
907 helo verification defer as a failure.
909 PH/51 If self=fail was set on a router, the bounce message did not include the
910 actual error message.
916 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
918 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
919 command itself ended in error, the underlying message about the transport
920 filter timeout was being overwritten with the pipe command error. Now the
921 underlying error message should be appended to the second error message.
923 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
925 PH/02 Reduce the amount of output that "make" produces by default. Full output
926 can still be requested.
928 PH/03 The warning log line about a condition test deferring for a "warn" verb
929 was being output only once per connection, rather than after each
930 occurrence (because it was using the same function as for successful
931 "warn" verbs). This seems wrong, so I have changed it.
933 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
934 it should not have, which might have caused a crash in the right
935 circumstances, but probably never did.
937 PH/04 Installed a modified version of Tony Finch's patch to make submission
938 mode fix the return path as well as the Sender: header line, and to
939 add a /name= option so that you can make the user's friendly name appear
942 TF/03 Added the control = fakedefer ACL modifier.
944 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
945 Mark Lowes for thorough testing.
947 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
949 TK/03 Merged latest SRS patch from Miles Wilton.
951 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
952 with the definition in sysexits.h (which is #included earlier).
953 Fortunately, Exim does not actually use EX_OK. The code used to try to
954 preserve the sysexits.h value, by assumimg that macro definitions were
955 scanned for macro replacements. I have been disabused of this notion,
956 so now the code just undefines EX_OK before #including unistd.h.
958 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
959 in the smtp transport. When a block could not be written in a single
960 write() function, the timeout was being re-applied to each part-write.
961 This seems wrong - if the receiver was accepting one byte at a time it
962 would take for ever. The timeout is now adjusted when this happens. It
963 doesn't have to be particularly precise.
965 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
966 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
968 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
969 without a domain if the check to <postmaster@domain> fails.
971 SC/01 Eximstats: added -xls and the ability to specify output files
972 (patch written by Frank Heydlauf).
974 SC/02 Eximstats: use FileHandles for outputing results.
976 SC/03 Eximstats: allow any combination of xls, txt, and html output.
978 SC/04 Eximstats: fixed display of large numbers with -nvr option
980 SC/05 Eximstats: fixed merging of reports with empty tables.
982 SC/06 Eximstats: added the -include_original_destination flag
984 SC/07 Eximstats: removed tabs and trailing whitespace.
986 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
988 TK/06 MBOX spool code: Add real "From " MBOX separator line
989 so the .eml file is really in mbox format (even though
990 most programs do not really care). Patch from Alex Miller.
992 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
993 The latter is generated from $received_to and is only set if the
994 message has one envelope recipient. SA can use these headers,
995 obviously out-of-the-box. Patch from Alex Miller.
997 PH/08 The ${def test on a variable was returning false if the variable's
998 value was "0", contrary to what the specification has always said!
999 The result should be true unless the variable is empty.
1001 PH/09 The syntax error of a character other than { following "${if
1002 def:variable_name" (after optional whitespace) was not being diagnosed.
1003 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1004 accidental colon was present, for example, could give incorrect results.
1006 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1007 result is used (not including appendfile, where other changes are about
1010 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1011 This involved changing a lot of size variables from int to off_t. It
1012 should work with maildirs and everything.
1014 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1015 spamd dying while we are connected to it.
1017 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1020 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1021 function for BDB 4.3.
1023 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1025 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1026 the sources for PCRE have been re-arranged and more files are now
1029 PH/15 The code I had for printing potentially long long variables in PH/11
1030 above was not the best (it lost precision). The length of off_t variables
1031 is now inspected at build time, and an appropriate printing format (%ld
1032 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1033 to be "long long int" or "long int". This is needed for the internal
1034 formatting function string_vformat().
1036 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1037 the configuration file to be ":syslog", then the script "guesses" where
1038 the logs files are, rather than using the compiled in default. In our
1039 case the guess is not the same as the compiled default, so the script
1040 suddenly stopped working when I started to use syslog. The patch checks
1041 to see if log_file_path is "". If so, it attempts to read it from exim
1042 with no configuration file to get the compiled in version, before it
1043 falls back to the previous guessing code."
1045 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1046 implementing BATV in an Exim configuration. See NewStuff for the gory
1049 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1050 Makefile that are specific to HP-UX.
1052 PH/18 If the "use_postmaster" option was set for a recipient callout together
1053 with the "random" option, the postmaster address was used as the MAIL
1054 FROM address for the random test, but not for the subsequent recipient
1055 test. It is now used for both.
1057 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1058 patch removes a few documentation additions to RFC 3028, because the
1059 latest draft now contains them. It adds the new en;ascii-case comparator
1060 and a new error check for 8bit text in MIME parts. Comparator and
1061 require names are now matched exactly. I enabled the subaddress
1062 extension, but it is not well tested yet (read: it works for me)."
1064 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1065 rework some of the code of TK/09 above to avoid the hardwired use of
1066 "%lld" and "long long". Replaced the call to snprintf() with a call to
1069 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1070 records point to non-existent hosts", "retry timeout exceeded", and
1071 "retry time not reached for any host after a long failure period".
1073 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1074 experimental DomainKeys support:
1076 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1077 (2) On an error such as an illegally used "control", the wrong name for
1078 the control was given.
1080 These problems did NOT occur unless DomainKeys support was compiled.
1082 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1084 PH/24 Added ${if match_ip condition.
1086 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1087 databases so that it will be absolutely obvious if a crash occurs in the
1088 DB library. This is a regular occurrence (often caused by mis-matched
1091 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1092 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1093 picked up on a user's system that detects such things. There doesn't seem
1094 to be a gcc warning option for this - only an attribute that has to be
1095 put on the function's prototype. It seems that in Fedora Core 4 they have
1096 set this on a number of new functions. No doubt there will be more in due
1099 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1100 specify a transport. However, if an address that was verified by such a
1101 router was the subject of a callout, Exim crashed because it tried to
1102 read the rcpt_include_affixes from the non-existent transport. Now it
1103 just assumes that the setting of that option is false. This bug was
1104 introduced by 4.51/PH/31.
1106 PH/28 Changed -d+all to exclude +memory, because that information is very
1107 rarely of interest, but it makes the output a lot bigger. People tend to
1108 do -d+all out of habit.
1110 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1111 code in os-type was giving problems when libc.so lives in lib64, like on
1114 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1115 aren't the modern standard, and it seems that some systems' include files
1116 don't always have them. Exim was already checking for some of the newer
1117 ones like T_AAAA, and defining it itself. I've added checks for all the
1118 record types that Exim uses.
1120 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1121 not automatically generating a new one, as it is supposed to. This
1122 prevented TLS from working. If the file did exist, but contained invalid
1123 data, a new version was generated, as expected. It was only the case of a
1124 non-existent file that was broken.
1126 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1127 with a change in libdomainkeys > 0.64.
1129 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1130 from DNS. If the selector record carries the flag, it now has
1131 precedence over the domain-wide flag.
1133 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1135 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1136 the use of a port name that isn't defined in /etc/services) occurred, the
1137 message was deferred as in a normal delivery, and thus remained on the
1138 spool, instead of being failed because of the mua_wrapper setting. This
1139 is now fixed, and I tidied up some of the mua_wrapper messages at the
1142 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1143 the messages in a hash of arrays rather than using individual hashes.
1144 This is a bit cleaner and results in dramatic memory savings, albeit
1145 at a slight CPU cost.
1147 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1148 as requested by Marc Sherman.
1150 SC/10 Eximstats: added histograms for user specified patterns as requested
1153 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1155 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1156 fopen() in the content-scanning modules that did not already have it.
1162 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1163 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1165 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1167 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1169 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1170 to test Sieve filters that use "vacation".
1172 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1173 that changes the way the GnuTLS parameters are stored in the cache file.
1174 The new format can be generated externally. For backward compatibility,
1175 if the data in the cache doesn't make sense, Exim assumes it has read an
1176 old-format file, and it generates new data and writes a new file. This
1177 means that you can't go back to an older release without removing the
1180 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1181 work if there are any delivery delays because "one_time" forces the
1182 parent to be marked "delivered", so its unseen clone is never tried
1183 again. For this reason, Exim now forbids the simultaneous setting of
1186 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1187 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1188 entry if you want to know the details. The fix, however, neglected to
1189 consider the case when local delivery batching is involved. The test for
1190 "previously delivered" was not happening when checking to see if an
1191 address could be batched with a previous (undelivered) one; under
1192 certain circumstances this could lead to multiple deliveries to the same
1195 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1196 in its include files, and this causes problems building Exim.
1198 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1199 header_syntax) but Exim was just ignoring anything given after a slash.
1200 In particular, this caused confusion with an attempt to use "verify =
1201 reverse_host_lookup/defer_ok". An error is now given when options are
1202 supplied for verify items that do not have them. (Maybe reverse_host_
1203 lookup should have a defer_ok option, but that's a different point.)
1205 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1206 defined by RFC 821) to 2048, because there were problems with some AUTH
1207 commands, and RFC 1869 says the size should be increased for extended
1208 SMTP commands that take arguments.
1210 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1213 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1214 "unknown" error; now it says that the functionality isn't in the binary.
1216 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1217 an address' error message when a string expansion fails (syntax or
1218 whatever). Otherwise the password may appear in the log. Following change
1219 PH/42 below, there is no longer a chance of it appearing in a bounce
1222 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1224 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1225 of its addresses was ever tried. (Bugzilla bug #2.)
1227 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1228 the result incorrectly in the debug output. (It correctly added a newline
1229 to what was transported.)
1231 TF/01 Added $received_time.
1233 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1234 commented out examples of how to interface to a virus scanner and to
1235 SpamAssassin. Also added commented examples of av_scanner and
1236 spamd_address settings.
1238 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1239 and controls are allowed in which ACLs. There were a couple of minor
1240 errors. Some of the entries in the conditions table (which is a table of
1241 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1242 negation of where the condition IS allowed.
1244 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1246 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1247 header file does not have a version number, so I've had to invent a new
1248 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1249 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1250 radiusclient), but it was contributed by a Radius user.
1252 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1253 files or directories, for OpenSSL.
1255 PH/19 When an Exim process that is running as root has to create an Exim log
1256 file, it does so in a subprocess that runs as exim:exim so as to get the
1257 ownership right at creation (otherwise, other Exim processes might see
1258 the file with the wrong ownership). There was no test for failure of this
1259 fork() call, which would lead to the process getting stuck as it waited
1260 for a non-existent subprocess. Forks do occasionally fail when resources
1261 run out. I reviewed all the other calls to fork(); they all seem to check
1264 PH/20 When checking for unexpected SMTP input at connect time (before writing
1265 the banner), Exim was not dealing correctly with a non-positive return
1266 from the read() function. If the client had disconnected by this time,
1267 the result was a log entry for a synchronization error with an empty
1268 string after "input=" when read() returned zero. If read() returned -1
1269 (an event I could not check), uninitialized data bytes were printed.
1270 There were reports of junk text (parts of files, etc) appearing after
1273 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1275 PH/22 Added support for macro redefinition, and (re)definition in between
1276 driver and ACL definitions.
1278 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1279 forgetting to use the resulting value; it was using the unexpanded value.
1281 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1282 hadn't been configured. The fix is from Juergen Kreileder, who
1283 understands it better than I do:
1285 "Here's what I see happening with three configured cyrus_sasl
1286 authenticators configured (plain, login, cram-md5):
1288 On startup auth_cyrus_sasl_init() gets called for each of these.
1289 This means three calls to sasl_listmech() without a specified mech_list.
1290 => SASL tests which mechs of all available mechs actually work
1291 => three warnings about OTP not working
1292 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1294 With the patch, sasl_listmech() also gets called three times. But now
1295 SASL's mech_list option is set to the server_mech specified in the the
1296 authenticator. Or in other words, the answer from sasl_listmech()
1297 gets limited to just the mech you're testing for (which is different
1299 => the return list contains just 'plain' or 'login', 'cram-md5' or
1300 nothing depending on the value of ob->server_mech.
1302 I've just tested the patch: Authentication still works fine,
1303 unavailable mechs specified in the exim configuration are still
1304 caught, and the auth.log warnings about OTP are gone."
1306 PH/25 When debugging is enabled, the contents of the command line are added
1307 to the debugging output, even when log_selector=+arguments is not
1310 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1311 answer is "GNU", and only if the return is "GNU/something" is the answer
1314 PH/27 $acl_verify_message is now set immediately after the failure of a
1315 verification in an ACL, and so is available in subsequent modifiers. In
1316 particular, the message can be preserved by coding like this:
1318 warn !verify = sender
1319 set acl_m0 = $acl_verify_message
1321 Previously, $acl_verify_message was set only while expanding "message"
1322 and "log_message" when a very denied access.
1324 PH/28 Modified OS/os.c-Linux with
1326 -#ifndef OS_LOAD_AVERAGE
1327 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1329 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1330 nomenclature these days.)
1332 PH/29 Installed patch from the Sieve maintainer that adds the options
1333 sieve_useraddress and sieve_subaddress to the redirect router.
1335 PH/30 In these circumstances:
1336 . Two addresses routed to the same list of hosts;
1337 . First host does not offer TLS;
1338 . First host accepts first address;
1339 . First host gives temporary error to second address;
1340 . Second host offers TLS and a TLS session is established;
1341 . Second host accepts second address.
1342 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1343 and peerdn, if requested) that were in fact used only for the second
1346 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1347 attention to any local part prefix or suffix that was matched by the
1348 router that accepted the address. It now behaves in the same way as it
1349 does for delivery: the affixes are removed from the local part unless
1350 rcpt_include_affixes is set on the transport.
1352 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1353 timeout during the DATA phase of an incoming message.
1355 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1356 applied a patch sanctioned by the Sieve maintainer.
1358 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1359 the uid or gid is negative. A case of a negative gid caused this to be
1360 noticed. The fix allows for either to be negative.
1362 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1363 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1366 PH/36 The change PH/12 above was broken. Fixed it.
1368 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1369 the grounds that routing the same address twice would always produce the
1370 same answer. This might have been true once, but it is certainly no
1371 longer true now. Routing a child address may depend on the previous
1372 routing that produced that child. Some complicated redirection strategies
1373 went wrong when messages had multiple recipients, and made Exim's
1374 behaviour dependent on the order in which the addresses were given.
1376 I have moved the duplicate checking until after the routing is complete.
1377 Exim scans the addresses that are assigned to local and remote
1378 transports, and removes any duplicates. This means that more work will be
1379 done, as duplicates will always all be routed, but duplicates are
1380 presumably rare, so I don't expect this is of any significance.
1382 For deliveries to pipes, files, and autoreplies, the duplicate checking
1383 still happens during the routing process, since they are not going to be
1386 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1387 It corrects a timeout issue with spamd. This is Ian's comment: "The
1388 background is that sometimes spamd either never reads data from a
1389 connection it has accepted, or it never writes response data. The exiscan
1390 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1391 blindly assumes that writes won't block so it may never time out."
1393 PH/39 Allow G after quota size as well as K and M.
1395 PH/40 The value set for $authenticated_id in an authenticator may not contain
1396 binary zeroes or newlines because the value is written to log lines and
1397 to spool files. There was no check on this. Now the value is run through
1398 the string_printing() function so that such characters are converted to
1399 printable escape sequences.
1401 PH/41 $message_linecount is a new variable that contains the total number of
1402 lines in the message. Compare $body_linecount, which is the count for the
1405 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1406 bounce and delay warning messages, except in certain special cases, which
1409 (a) An SMTP error message from a remote host;
1410 (b) A message specified in a :fail: redirection;
1411 (c) A message specified in a "fail" command in a system filter;
1412 (d) A message specified in a FAIL return from the queryprogram router;
1413 (e) A message specified by the cannot_route_message router option.
1415 In these cases only, Exim does include the error details in bounce and
1416 warning messages. There are also a few cases where bland messages such
1417 as "unrouteable address" or "local delivery error" are given.
1419 PH/43 $value is now also set for the "else" part of a ${run expansion.
1421 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1422 being worked on, but at least Exim now implements the latest version to
1425 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1426 process to complete was treated as a delivery failure, a timeout while
1427 writing the message to the pipe was logged, but erroneously treated as a
1428 successful delivery. Such timeouts include transport filter timeouts. For
1429 consistency with the overall process timeout, these timeouts are now
1430 treated as errors, giving rise to delivery failures by default. However,
1431 there is now a new Boolean option for the pipe transport called
1432 timeout_defer, which, if set TRUE, converts the failures into defers for
1433 both kinds of timeout. A transport filter timeout is now identified in
1436 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1437 systems where "make" and "gmake" are different, calling "gmake" at top
1438 level broke things. I've arranged for the value of $(MAKE) to be passed
1439 from the Makefile to this script so that it can call the same version of
1443 A note about Exim versions 4.44 and 4.50
1444 ----------------------------------------
1446 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1447 changes of various kinds. As a consequence, a big documentation update was
1448 needed. This delayed the release for rather longer than seemed good, especially
1449 in the light of a couple of (minor) security issues. Therefore, the changes
1450 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1451 release. So 4.44 and 4.50 are in effect two different branches that both start
1454 I have left the 4.50 change log unchanged; it contains all the changes since
1455 4.43. The change log for 4.44 is below; many of its items are identical to
1456 those for 4.50. This seems to be the most sensible way to preserve the
1457 historical information.
1463 1. Minor wording change to the doc/README.SIEVE file.
1465 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1466 computation of the current number of files was incorrect.
1468 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1469 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1472 4. Give more explanation in the error message when the command for a transport
1473 filter fails to execute.
1475 5. There are several places where Exim runs a non-Exim command in a
1476 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1477 was being done only for the command run by the queryprogram router. It is
1478 now done for all such subprocesses. The other cases are: ${run, transport
1479 filters, and the commands run by the lmtp and pipe transports.
1481 6. Added CONFIGURE_GROUP build-time option.
1483 7. Some older OS have a limit of 256 on the maximum number of file
1484 descriptors. Exim was using setrlimit() to set 1000 as a large value
1485 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1486 systems. I've change it so that if it can't get 1000, it tries for 256.
1488 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1489 was an oversight, and furthermore, ever since the addition of extra
1490 controls (e.g. 4.43/32), the checks on when to allow different forms of
1491 "control" were broken. There should now be diagnostics for all cases when a
1492 control that does not make sense is encountered.
1494 9. Added the /retain_sender option to "control=submission".
1496 10. $recipients is now available in the predata ACL (oversight).
1498 11. Tidy the search cache before the fork to do a delivery from a message
1499 received from the command line. Otherwise the child will trigger a lookup
1500 failure and thereby defer the delivery if it tries to use (for example) a
1501 cached ldap connection that the parent has called unbind on.
1503 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1504 of $address_data from the recipient verification was clobbered by the
1505 sender verification.
1507 13. The value of address_data from a sender verification is now available in
1508 $sender_address_data in subsequent conditions in the ACL statement.
1510 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1512 15. Added a new option "connect=<time>" to callout options, to set a different
1515 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1516 was its contents. (It was OK if the option was not defined at all.)
1518 17. A "Completed" log line is now written for messages that are removed from
1519 the spool by the -Mrm option.
1521 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1522 information about exactly what failed.
1524 19. Added -dd to debug only the daemon process.
1526 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1527 handles timeouts, both on the server side and network timeouts. Renamed the
1528 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1530 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1531 It is now set to "smtps".
1533 22. $host_address is now set to the target address during the checking of
1534 ignore_target_hosts.
1536 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1537 being passed; this would have caused $sender_host_name to have been used if
1538 matching the list had actually called for a host name (not very likely,
1539 since this list is usually IP addresses). A host name is now passed as
1542 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1543 code that specifies a non-restarting handler (typically sigaction() in
1544 modern systems) in an attempt to fix a rare and obscure crash bug.
1546 25. Narrowed the window for a race in the daemon that could cause it to ignore
1547 SIGCHLD signals. This is not a major problem, because they are used only to
1548 wake it up if nothing else does.
1550 26. A malformed maildirsize file could cause Exim to calculate negative values
1551 for the mailbox size or file count. Odd effects could occur as a result.
1552 The maildirsize information is now recalculated if the size or filecount
1555 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1556 support for a long time. Removed HAVE_SYS_VFS_H.
1558 28. Installed the latest version of exipick from John Jetmore.
1560 29. In an address list, if the pattern was not a regular expression, an empty
1561 subject address (from a bounce message) matched only if the pattern was an
1562 empty string. Non-empty patterns were not even tested. This was the wrong
1563 because it is perfectly reasonable to use an empty address as part of a
1564 database query. An empty address is now tested by patterns that are
1565 lookups. However, all the other forms of pattern expect the subject to
1566 contain a local part and a domain, and therefore, for them, an empty
1567 address still always fails if the pattern is not itself empty.
1569 30. Exim went into a mad DNS loop when attempting to do a callout where the
1570 host was specified on an smtp transport, and looking it up yielded more
1571 than one IP address.
1573 31. Re-factored the code for checking spool and log partition space into a
1574 function that finds that data and another that does the check. The former
1575 is then used to implement four new variables: $spool_space, $log_space,
1576 $spool_inodes, and $log_inodes.
1578 32. The RFC2047 encoding function was originally intended for short strings
1579 such as real names; it was not keeping to the 75-character limit for
1580 encoded words that the RFC imposes. It now respects the limit, and
1581 generates multiple encoded words if necessary. To be on the safe side, I
1582 have increased the buffer size for the ${rfc2047: expansion operator from
1585 33. It is now permitted to omit both strings after an "if" condition; if the
1586 condition is true, the result is "true". As before, when the second string
1587 is omitted, a false condition yields an empty string. This makes it less
1588 cumbersome to write custom ACL and router conditions.
1590 34. Failure to deliver a bounce message always caused it to be frozen, even if
1591 there was an errors_to setting on the router. The errors_to setting is now
1594 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1595 canonical form (fully expanded) before being placed in
1596 $sender_host_address.
1598 36. The table in the code that translates DNS record types into text (T_A to
1599 "A" for instance) was missing entries for NS and CNAME. It is just possible
1600 that this could have caused confusion if both these types were looked up
1601 for the same domain, because the text type is used as part of Exim's
1602 per-process caching. But the chance of anyone hitting this buglet seems
1605 37. The dnsdb lookup has been extended in a number of ways.
1607 (1) There is a new type, "zns", which walks up the domain tree until it
1608 finds some nameserver records. It should be used with care.
1610 (2) There is a new type, "mxh", which is like "mx" except that it returns
1611 just the host names, not the priorities.
1613 (3) It is now possible to give a list of domains (or IP addresses) to be
1614 looked up. The behaviour when one of the lookups defers can be
1615 controlled by a keyword.
1617 (4) It is now possible to specify the separator character for use when
1618 multiple records are returned.
1620 38. The dnslists ACL condition has been extended: it is now possible to supply
1621 a list of IP addresses and/or domains to be looked up in a particular DNS
1624 39. Added log_selector=+queue_time_overall.
1626 40. When running the queue in the test harness, wait just a tad after forking a
1627 delivery process, to get repeatability of debugging output.
1629 41. Include certificate and key file names in error message when GnuTLS fails
1630 to set them up, because the GnuTLS error message doesn't include the name
1631 of the failing file when there is a problem reading it.
1633 42. Allow both -bf and -bF in the same test run.
1635 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1637 44. The "Exiscan patch" is now merged into the mainline Exim source.
1639 45. Sometimes the final signoff response after QUIT could fail to get
1640 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1641 < 0 before doing a fflush(). This bug looks as though it goes back to the
1642 introduction of TLS in release 3.20, but "sometimes" must have been rare
1643 because the tests only now provoked it.
1645 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1646 (this can affect the format of dates).
1648 47. exim_tidydb, when checking for the continued existence of a message for
1649 which it has found a message-specific retry record, was not finding
1650 messages that were in split spool directories. Consequently, it was
1651 deleting retry records that should have stayed in existence.
1653 48. Steve fixed some bugs in eximstats.
1655 49. The SPA authentication driver was not abandoning authentication and moving
1656 on to the next authenticator when an expansion was forced to fail,
1657 contradicting the general specification for all authenticators. Instead it
1658 was generating a temporary error. It now behaves as specified.
1660 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1661 (the order specifies the preference for clients). The order is now AES256,
1662 AES128, 3DES, ARCFOUR128.
1664 51. Small patch to Sieve code - explicitly set From: when generating an
1667 52. Exim crashed if a remote delivery caused a very long error message to be
1668 recorded - for instance if somebody sent an entire SpamAssassin report back
1669 as a large number of 550 error lines. This bug was coincidentally fixed by
1670 increasing the size of one of Exim's internal buffers (big_buffer) that
1671 happened as part of the Exiscan merge. However, to be on the safe side, I
1672 have made the code more robust (and fixed the comments that describe what
1675 53. Now that there can be additional text after "Completed" in log lines (if
1676 the queue_time_overall log selector is set), a one-byte patch to exigrep
1677 was needed to allow it to recognize "Completed" as not the last thing in
1680 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1681 patch that reportedly fixes this has been added. I am not expert enough to
1682 create a test for it. This is what the patch creator wrote:
1684 "I found a little strange behaviour of ldap code when working with
1685 Windows 2003 AD Domain, where users was placed in more than one
1686 Organization Units. When I tried to give exim partial DN, the exit code
1687 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1688 But simultaneously result of request was absolutely normal ldap result,
1689 so I produce this patch..."
1691 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1692 REFERENCE, so I have modified the code to exclude the patch when that macro
1695 55. Some experimental protocols are using DNS PTR records for new purposes. The
1696 keys for these records are domain names, not reversed IP addresses. The
1697 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1698 leaves it alone. Component reversal etc. now happens only for IP addresses.
1701 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1703 57. Double the size of the debug message buffer (to 2048) so that more of very
1704 long debug lines gets shown.
1706 58. The exicyclog utility now does better if the number of log files to keep
1707 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1709 59. Two changes related to the smtp_active_hostname option:
1711 (1) $smtp_active_hostname is now available as a variable.
1712 (2) The default for smtp_banner uses $smtp_active_hostname instead
1713 of $primary_hostname.
1715 60. The host_aton() function is supposed to be passed a string that is known
1716 to be a valid IP address. However, in the case of IPv6 addresses, it was
1717 not checking this. This is a hostage to fortune. Exim now panics and dies
1718 if the condition is not met. A case was found where this could be provoked
1719 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1720 components; fortuitously, this particular loophole had already been fixed
1721 by change 4.50/55 above.
1723 If there are any other similar loopholes, the new check in host_aton()
1724 itself should stop them being exploited. The report I received stated that
1725 data on the command line could provoke the exploit when Exim was running as
1726 exim, but did not say which command line option was involved. All I could
1727 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1728 running as the user.
1731 61. There was a buffer overflow vulnerability in the SPA authentication code
1732 (which came originally from the Samba project). I have added a test to the
1733 spa_base64_to_bits() function which I hope fixes it.
1736 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1737 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1739 63. The daemon start-up calls getloadavg() while still root for those OS that
1740 need the first call to be done as root, but it missed one case: when
1741 deliver_queue_load_max is set with deliver_drop_privilege. This is
1742 necessary for the benefit of the queue runner, because there is no re-exec
1743 when deliver_drop_privilege is set.
1745 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1746 This has been fixed.
1748 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1749 was in use, was not putting the data itself into the right store pool;
1750 consequently, it could be overwritten for a subsequent message in the same
1751 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1754 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1756 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1757 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1758 the function were treating the return as a boolean value, which happened to
1759 work because 0=false and not-0=true, but is not correct code.
1761 68. The host_aton() function was not handling scoped IPv6 addresses (those
1762 with, for example, "%eth0" on the end) correctly.
1764 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1765 negated items (that is, ~something) in unsigned ints. Some compilers
1766 apparently mutter when there is no cast.
1768 70. If an address verification called from an ACL failed, and did not produce a
1769 user-specific message (i.e. there was only a "system" message), nothing was
1770 put in $acl_verify_message. In this situation, it now puts the system
1773 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1774 session; change 4.31/43 added the unwanted input to the log line - except
1775 that it did not do this in the start of session case. It now does.
1777 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1778 This is wrong and can cause the other end to generate a synchronization
1779 error if it is another Exim or anything else that does the synchronization
1780 check. A QUIT command is no longer sent after a timeout.
1782 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1783 during host lookups.
1785 74. The defer_ok option of callout verification was not working if it was used
1786 when verifying addresses in header lines, that is, for this case:
1788 verify = header_sender/callout=defer_ok
1790 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1791 those file descriptors could be used for SMTP connections. If anything
1792 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1793 could be sent to the SMTP client, causing chaos. The daemon now opens
1794 stdin, stdout, and stderr to /dev/null when it puts itself into the
1797 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1798 log by default. The user can override this with suitable Perl magic.
1800 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1801 the log message when discard triggers, was not working for the DATA ACL or
1802 for the non-SMTP ACL.
1804 78. Error message wording change in sieve.c.
1806 79. If smtp_accept_max_per_host was set, the number of connections could be
1807 restricted to fewer than expected, because the daemon was trying to set up
1808 a new connection before checking whether the processes handling previous
1809 connections had finished. The check for completed processes is now done
1810 earlier. On busy systems, this bug wouldn't be noticed because something
1811 else would have woken the daemon, and it would have reaped the completed
1814 80. If a message was submitted locally by a user whose login name contained one
1815 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1816 It caused a spool format error. I have fixed the spool reading code. A
1817 related problem was that the "from" clause in the Received: line became
1818 illegal because of the space(s). It is now covered by ${quote_local_part.
1820 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1823 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1825 83. Patch from Sieve maintainer to fix unterminated string problem in
1826 "vacation" handling.
1828 84. Some minor changes to the Linux configuration files to help with other
1829 OS variants using glibc.
1831 85. One more patch for Sieve to update vacation handling to latest spec.
1834 ----------------------------------------------------
1835 See the note above about the 4.44 and 4.50 releases.
1836 ----------------------------------------------------
1842 1. Change 4.43/35 introduced a bug that caused file counts to be
1843 incorrectly computed when quota_filecount was set in an appendfile
1846 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1847 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1850 3. Give more explanation in the error message when the command for a transport
1851 filter fails to execute.
1853 4. There are several places where Exim runs a non-Exim command in a
1854 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1855 was being done only for the command run by the queryprogram router. It is
1856 now done for all such subprocesses. The other cases are: ${run, transport
1857 filters, and the commands run by the lmtp and pipe transports.
1859 5. Some older OS have a limit of 256 on the maximum number of file
1860 descriptors. Exim was using setrlimit() to set 1000 as a large value
1861 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1862 systems. I've change it so that if it can't get 1000, it tries for 256.
1864 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1865 was an oversight, and furthermore, ever since the addition of extra
1866 controls (e.g. 4.43/32), the checks on when to allow different forms of
1867 "control" were broken. There should now be diagnostics for all cases when a
1868 control that does not make sense is encountered.
1870 7. $recipients is now available in the predata ACL (oversight).
1872 8. Tidy the search cache before the fork to do a delivery from a message
1873 received from the command line. Otherwise the child will trigger a lookup
1874 failure and thereby defer the delivery if it tries to use (for example) a
1875 cached ldap connection that the parent has called unbind on.
1877 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1878 of $address_data from the recipient verification was clobbered by the
1879 sender verification.
1881 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1882 was its contents. (It was OK if the option was not defined at all.)
1884 11. A "Completed" log line is now written for messages that are removed from
1885 the spool by the -Mrm option.
1887 12. $host_address is now set to the target address during the checking of
1888 ignore_target_hosts.
1890 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1891 being passed; this would have caused $sender_host_name to have been used if
1892 matching the list had actually called for a host name (not very likely,
1893 since this list is usually IP addresses). A host name is now passed as
1896 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1897 code that specifies a non-restarting handler (typically sigaction() in
1898 modern systems) in an attempt to fix a rare and obscure crash bug.
1900 15. Narrowed the window for a race in the daemon that could cause it to ignore
1901 SIGCHLD signals. This is not a major problem, because they are used only to
1902 wake it up if nothing else does.
1904 16. A malformed maildirsize file could cause Exim to calculate negative values
1905 for the mailbox size or file count. Odd effects could occur as a result.
1906 The maildirsize information is now recalculated if the size or filecount
1909 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1910 support for a long time. Removed HAVE_SYS_VFS_H.
1912 18. Updated exipick to current release from John Jetmore.
1914 19. Allow an empty sender to be matched against a lookup in an address list.
1915 Previously the only cases considered were a regular expression, or an
1918 20. Exim went into a mad DNS lookup loop when doing a callout where the
1919 host was specified on the transport, if the DNS lookup yielded more than
1922 21. The RFC2047 encoding function was originally intended for short strings
1923 such as real names; it was not keeping to the 75-character limit for
1924 encoded words that the RFC imposes. It now respects the limit, and
1925 generates multiple encoded words if necessary. To be on the safe side, I
1926 have increased the buffer size for the ${rfc2047: expansion operator from
1929 22. Failure to deliver a bounce message always caused it to be frozen, even if
1930 there was an errors_to setting on the router. The errors_to setting is now
1933 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1934 canonical form (fully expanded) before being placed in
1935 $sender_host_address.
1937 24. Updated eximstats to version 1.33
1939 25. Include certificate and key file names in error message when GnuTLS fails
1940 to set them up, because the GnuTLS error message doesn't include the name
1941 of the failing file when there is a problem reading it.
1943 26. Expand error message when OpenSSL has problems setting up cert/key files.
1946 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1947 (this can affect the format of dates).
1949 28. exim_tidydb, when checking for the continued existence of a message for
1950 which it has found a message-specific retry record, was not finding
1951 messages that were in split spool directories. Consequently, it was
1952 deleting retry records that should have stayed in existence.
1954 29. eximstats updated to version 1.35
1955 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1956 1.35 - bugfix such that pie charts by volume are generated correctly
1958 30. The SPA authentication driver was not abandoning authentication and moving
1959 on to the next authenticator when an expansion was forced to fail,
1960 contradicting the general specification for all authenticators. Instead it
1961 was generating a temporary error. It now behaves as specified.
1963 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1964 (the order specifies the preference for clients). The order is now AES256,
1965 AES128, 3DES, ARCFOUR128.
1967 31. Small patch to Sieve code - explicitly set From: when generating an
1970 32. Exim crashed if a remote delivery caused a very long error message to be
1971 recorded - for instance if somebody sent an entire SpamAssassin report back
1972 as a large number of 550 error lines. This bug was coincidentally fixed by
1973 increasing the size of one of Exim's internal buffers (big_buffer) that
1974 happened as part of the Exiscan merge. However, to be on the safe side, I
1975 have made the code more robust (and fixed the comments that describe what
1978 33. Some experimental protocols are using DNS PTR records for new purposes. The
1979 keys for these records are domain names, not reversed IP addresses. The
1980 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1981 leaves it alone. Component reversal etc. now happens only for IP addresses.
1984 34. The host_aton() function is supposed to be passed a string that is known
1985 to be a valid IP address. However, in the case of IPv6 addresses, it was
1986 not checking this. This is a hostage to fortune. Exim now panics and dies
1987 if the condition is not met. A case was found where this could be provoked
1988 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1989 components; fortuitously, this particular loophole had already been fixed
1990 by change 4.50/55 or 4.44/33 above.
1992 If there are any other similar loopholes, the new check in host_aton()
1993 itself should stop them being exploited. The report I received stated that
1994 data on the command line could provoke the exploit when Exim was running as
1995 exim, but did not say which command line option was involved. All I could
1996 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1997 running as the user.
2000 35. There was a buffer overflow vulnerability in the SPA authentication code
2001 (which came originally from the Samba project). I have added a test to the
2002 spa_base64_to_bits() function which I hope fixes it.
2005 36. The daemon start-up calls getloadavg() while still root for those OS that
2006 need the first call to be done as root, but it missed one case: when
2007 deliver_queue_load_max is set with deliver_drop_privilege. This is
2008 necessary for the benefit of the queue runner, because there is no re-exec
2009 when deliver_drop_privilege is set.
2011 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2012 was in use, was not putting the data itself into the right store pool;
2013 consequently, it could be overwritten for a subsequent message in the same
2014 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2017 38. Sometimes the final signoff response after QUIT could fail to get
2018 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2019 < 0 before doing a fflush(). This bug looks as though it goes back to the
2020 introduction of TLS in release 3.20, but "sometimes" must have been rare
2021 because the tests only now provoked it.
2027 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2028 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2029 Now it returns an integer. A number of calls were still expecting a T/F
2030 return. Fortuitously, in all cases, the tests worked in OK situations,
2031 which is the norm. However, things would have gone wrong on any write
2032 failures on the smtp file descriptor. This function is used when sending
2033 messages over SMTP and also when doing verify callouts.
2035 2. When Exim is called to do synchronous delivery of a locally submitted
2036 message (the -odf or -odi options), it no longer closes stderr before doing
2039 3. Implemented the mua_wrapper option.
2041 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2043 5. Implemented the functions header_remove(), header_testname(),
2044 header_add_at_position(), and receive_remove_recipient(), and exported them
2047 6. If an ACL "warn" statement specified the addition of headers, Exim already
2048 inserted X-ACL-Warn: at the start if there was no header name. However, it
2049 was not making this test for the second and subsequent header lines if
2050 there were newlines in the string. This meant that an invalid header could
2051 be inserted if Exim was badly configured.
2053 7. Allow an ACL "warn" statement to add header lines at the start or after all
2054 the Received: headers, as well as at the end.
2056 8. Added the rcpt_4xx retry error code.
2058 9. Added postmaster_mailfrom=xxx to callout verification option.
2060 10. Added mailfrom=xxxx to the callout verification option, for verify=
2063 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2064 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2066 12. Inserted some casts to stop certain compilers warning when using pointer
2067 differences as field lengths or precisions in printf-type calls (mostly
2068 affecting debugging statements).
2070 13. Added optional readline() support for -be (dynamically loaded).
2072 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2073 same clock tick as a message's arrival, so that its received time was the
2074 same as the "first fail" time on the retry record, and that message
2075 remained on the queue past the ultimate address timeout, every queue runner
2076 would try a delivery (because it was past the ultimate address timeout) but
2077 after another failure, the ultimate address timeout, which should have then
2078 bounced the address, did not kick in. This was a "< instead of <=" error;
2079 in most cases the first failure would have been in the next clock tick
2080 after the received time, and all would be well.
2082 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2083 being recognized when the domain list was tested by the match_domain
2084 condition in an expansion string.
2086 16. Added the ${str2b64: operator.
2088 17. Exim was always calling setrlimit() to set a large limit for the number of
2089 processes, without checking whether the existing limit was already
2090 adequate. (It did check for the limit on file descriptors.) Furthermore,
2091 errors from getrlimit() and setrlimit() were being ignored. Now they are
2092 logged to the main and panic logs, but Exim does carry on, to try to do its
2093 job under whatever limits there are.
2095 18. Imported PCRE 5.0.
2097 19. Trivial typo in log message " temporarily refused connection" (the leading
2100 20. If the log selector return_path_on_delivery was set and an address was
2101 redirected to /dev/null, the delivery process crashed because it assumed
2102 that a return path would always be set for a "successful" delivery. In this
2103 case, the whole delivery is bypassed as an optimization, and therefore no
2106 21. Internal re-arrangement: the function for sending a challenge and reading
2107 a response while authentication was assuming a zero-terminated challenge
2108 string. It's now changed to take a pointer and a length, to allow for
2109 binary data in such strings.
2111 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2113 23. Exim was not respecting finduser_retries when seeking the login of the
2114 uid under which it was called; it was always trying 10 times. (The default
2115 setting of finduser_retries is zero.) Also, it was sleeping after the final
2116 failure, which is pointless.
2118 24. Implemented tls_on_connect_ports.
2120 25. Implemented acl_smtp_predata.
2122 26. If the domain in control=submission is set empty, Exim assumes that the
2123 authenticated id is a complete email address when it generates From: or
2124 Sender: header lines.
2126 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2127 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2128 chown and chgrp in /bin and hostname in /usr/bin.
2130 28. Exim was keeping the "process log" file open after each use, just as it
2131 does for the main log. This opens the possibility of it remaining open for
2132 long periods when the USR1 signal hits a daemon. Occasional processlog
2133 errors were reported, that could have been caused by this. Anyway, it seems
2134 much more sensible not to leave this file open at all, so that is what now
2137 29. The long-running daemon process does not normally write to the log once it
2138 has entered its main loop, and it closes the log before doing so. This is
2139 so that log files can straightforwardly be renamed and moved. However,
2140 there are a couple of unusual error situations where the daemon does write
2141 log entries, and I had neglected to close the log afterwards.
2143 30. The text of an SMTP error response that was received during a remote
2144 delivery was being truncated at 512 bytes. This is too short for some of
2145 the long messages that one sometimes sees. I've increased the limit to
2148 31. It is now possible to make retry rules that apply only when a message has a
2149 specific sender, in particular, an empty sender.
2151 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2152 it possible to be selective about when SMTP synchronization is enforced.
2154 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2156 32. Implemented hosts_connection_nolog.
2158 33. Added an ACL for QUIT.
2160 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2163 35. Added mailbox_size and mailbox_filecount to appendfile.
2165 36. Added control = no_multiline_responses to ACLs.
2167 37. There was a bug in the logic of the code that waits for the clock to tick
2168 in the case where the clock went backwards by a substantial amount such
2169 that the microsecond fraction of "now" was more than the microsecond
2170 fraction of "then" (but the whole seconds number was less).
2172 38. Added support for the libradius Radius client library this is found on
2173 FreeBSD (previously only the radiusclient library was supported).
2179 1. When certain lookups returned multiple values in the form name=value, the
2180 quoting of the values was not always being done properly. Specifically:
2181 (a) If the value started with a double quote, but contained no whitespace,
2183 (b) If the value contained whitespace other than a space character (i.e.
2184 tabs or newlines or carriage returns) it was not quoted.
2185 This fix has been applied to the mysql and pgsql lookups by writing a
2186 separate quoting function and calling it from the lookup code. The fix
2187 should probably also be applied to nisplus, ibase and oracle lookups, but
2188 since I cannot test any of those, I have not disturbed their existing code.
2190 2. A hit in the callout cache for a specific address caused a log line with no
2191 reason for rejecting RCPT. Now it says "Previous (cached) callout
2192 verification failure".
2194 3. There was an off-by-one bug in the queryprogram router. An over-long
2195 return line was truncated at 256 instead of 255 characters, thereby
2196 overflowing its buffer with the terminating zero. As well as fixing this, I
2197 have increased the buffer size to 1024 (and made a note to document this).
2199 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2200 when Exim is waiting for an SMTP response from a remote server, Exim
2201 restarts its select() call on the socket, thereby resetting its timeout.
2202 This is not a problem when such interrupts are rare. Somebody set up a cron
2203 job to run exiwhat every 2 minutes, which is less than the normal select()
2204 timeout (5 or 10 minutes). This meant that the select() timeout never
2205 kicked in because it was always reset. I have fixed this by comparing the
2206 time when an interrupt arrives with the time at the start of the first call
2207 to select(). If more time than the timeout has elapsed, the interrupt is
2208 treated as a timeout.
2210 5. Some internal re-factoring in preparation for the addition of Sieve
2211 extensions (by MH). In particular, the "personal" test is moved to a
2212 separate function, and given an option for scanning Cc: and Bcc: (which is
2213 not set for Exim filters).
2215 6. When Exim created an email address using the login of the caller as the
2216 local part (e.g. when creating a From: or Sender: header line), it was not
2217 quoting the local part when it contained special characters such as @.
2219 7. Installed new OpenBSD configuration files.
2221 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2222 try to make them clearer.
2224 9. Callout options, other than the timeout value, were being ignored when
2225 verifying sender addresses in header lines. For example, when using
2227 verify = header_sender/callout=no_cache
2229 the cache was (incorrectly) being used.
2231 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2232 only the Cygwin environment.
2234 11. When return_path_on_delivery was set as a log selector, if different remote
2235 addresses in the same message used different return paths and parallel
2236 remote delivery occurred, the wrong values would sometimes be logged.
2237 (Whenever a remote delivery process finished, the return path value from
2238 the most recently started remote delivery process was logged.)
2240 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2241 lines when AUTH and/or TLS are in use. This is the "received protocol"
2242 field. Exim used to use "asmtp" for authenticated SMTP, without any
2243 indication (in the protocol name) for TLS use. Now it follows the RFC and
2244 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2245 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2246 names appear in log lines as well as in Received: header lines.
2248 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2249 extensions, and comparison tests, and to fix some bugs.
2251 14. Changes to the "personal" filter test:
2253 (1) The test was buggy in that it was just doing the equivalent of
2254 "contains" tests on header lines. For example, if a user's address was
2255 anne@some.where, the "personal" test would incorrectly be true for
2257 To: susanne@some.where
2259 This test is now done by extracting each address from the header in turn,
2260 and checking the entire address. Other tests that are part of "personal"
2261 are now done using regular expressions (for example, to check local parts
2262 of addresses in From: header lines).
2264 (2) The list of non-personal local parts in From: addresses has been
2265 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2266 taken from the Sieve specification recommendations.
2268 (3) If the message contains any header line starting with "List-" it is
2269 treated as non-personal.
2271 (4) The test for "circular" in the Subject: header line has been removed
2272 because it now seems ill-conceived.
2274 15. Minor typos in src/EDITME comments corrected.
2276 16. Installed latest exipick from John Jetmore.
2278 17. If headers_add on a router specified a text string that was too long for
2279 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2280 of string_sprintf() is now avoided.
2282 18. $message_body_size was not set (it was always zero) when running the DATA
2283 ACL and the local_scan() function.
2285 19. For the "mail" command in an Exim filter, no default was being set for
2286 the once_repeat time, causing a random time value to be used if "once" was
2287 specified. (If the value happened to be <= 0, no repeat happened.) The
2288 default is now 0s, meaning "never repeat". The "vacation" command was OK
2289 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2290 (I found it when inspecting the code).
2292 20. There is now an overall timeout for performing a callout verification. It
2293 defaults to 4 times the callout timeout, which applies to individual SMTP
2294 commands during the callout. The overall timeout applies when there is more
2295 than one host that can be tried. The timeout is checked before trying the
2296 next host. This prevents very long delays if there are a large number of
2297 hosts and all are timing out (e.g. when the network connections are timing
2298 out). The value of the overall timeout can be changed by specifying an
2299 additional sub-option for "callout", called "maxwait". For example:
2301 verify = sender/callout=5s,maxwait=20s
2303 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2304 to the end before writing, but this should make it even safer).
2306 22. Exim was forgetting that it had advertised PIPELINING for the second and
2307 subsequent messages on an SMTP connection. It was also not resetting its
2308 memory on STARTTLS and an internal HELO.
2310 23. When Exim logs an SMTP synchronization error within a session, it now
2311 records whether PIPELINING has been advertised or not.
2313 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2314 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2315 rather than long int.
2317 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2319 26. Added the never_mail option to autoreply.
2325 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2326 crash if the getsockname() call failed; this can happen if a connection is
2327 closed very soon after it is established. The problem was simply in the
2328 order in which certain operations were done, causing Exim to try to write
2329 to the SMTP stream before it had set up the file descriptor. The bug has
2330 been fixed by making things happen in the correct order.
2336 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2337 before the connection was closed, thus losing the rejection response.
2339 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2340 some early Solaris releases, but causes trouble in current releases where
2341 socklen_t is defined.
2343 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2346 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2349 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2350 that releasing the top of it at the end releases what was used for sub-
2351 expansions (unless the block got too big). However, discard this block if
2352 the first thing is a variable or header, so that we can use its block when
2353 it is dynamic (useful for very large $message_headers, for example).
2355 6. Lookups now cache *every* query, not just the most recent. A new, separate
2356 store pool is used for this. It can be recovered when all lookup caches are
2357 flushed. Lookups now release memory at the end of their result strings.
2358 This has involved some general refactoring of the lookup sources.
2360 7. Some code has been added to the store_xxx() functions to reduce the amount
2361 of flapping under certain conditions.
2363 8. log_incoming_interface used to affect only the <= reception log lines. Now
2364 it causes the local interface and port to be added to several more SMTP log
2365 lines, for example "SMTP connection from", and rejection lines.
2367 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2369 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2371 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2372 could be overwritten at the end of the current message (or the start of a
2373 new message if it was set in a HELO ACL). The value is now preserved for
2374 the duration of the SMTP connection.
2376 12. If a transport had a headers_rewrite setting, and a matching header line
2377 contained an unqualified address, that address was qualified, even if it
2378 did not match any rewriting rules. The underlying bug was that the values
2379 of the flags that permit the existence of unqualified sender and recipient
2380 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2381 non-local messages, and by -bnq for local messages) were not being
2382 preserved with the message after it was received.
2384 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2385 "next input=" as part of the text comprising the host identity instead of
2386 the correct text. The code was using the same buffer for two different
2387 strings. However, depending on which order the printing function evaluated
2388 its arguments, the bug did not always show up. Under Linux, for example, my
2389 test suite worked just fine.
2391 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2392 doesn't work with some older versions of Perl. It has been changed to "my",
2393 which in any case is probably the better facility to use.
2395 15. A really picky compiler found some instances of statements for creating
2396 error messages that either had too many or two few arguments for the format
2399 16. The size of the buffer for calls to the DNS resolver has been increased
2400 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2401 for addresses that have a lot of PTR records. This alleviates a problem; it
2402 does not fully solve it.
2404 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2405 buffer now truncates the list and logs the incident, which is the same
2406 action as happens when Exim is looking up a host name and its aliases.
2407 Previously in this situation something unpredictable would happen;
2408 sometimes it was "internal error: store_reset failed".
2410 18. If a server dropped the connection unexpectedly when an Exim client was
2411 using GnuTLS and trying to read a response, the client delivery process
2412 crashed while trying to generate an error log message.
2414 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2415 string, for example:
2417 warn message = H1: something\nH2: something
2419 the text was added as a single header line from Exim's point of view
2420 though it ended up OK in the delivered message. However, searching for the
2421 second and subsequent header lines using $h_h2: did not work. This has been
2422 fixed. Similarly, if a system filter added multiple headers in this way,
2423 the routers could not see them.
2425 20. Expanded the error message when iplsearch is called with an invalid key to
2426 suggest using net-iplsearch in a host list.
2428 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2429 ACLs are no longer actually imposed (and a message to that effect is
2432 22. If a "gecos" field in a passwd entry contained escaped characters, in
2433 particular, if it contained a \" sequence, Exim got it wrong when building
2434 a From: or a Sender: header from that name. A second bug also caused
2435 incorrect handling when an unquoted " was present following a character
2436 that needed quoting.
2438 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2439 was not being matched caselessly.
2441 24. Arranged for all hyphens in the exim.8 source to be escaped with
2444 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2445 properly. Recipient callout cache records were still being keyed to include
2446 the sender, even when use_sender was set false. This led to far more
2447 callouts that were necessary. The sender is no longer included in the key
2448 when use_sender is false.
2450 26. Added "control = submission" modifier to ACLs.
2452 27. Added the ${base62d: operator to decode base 62 numbers.
2454 28. dnsdb lookups can now access SRV records.
2456 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2457 the configuration file.
2459 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2460 (-v) mode. This makes the output for a verbose queue run more intelligible.
2462 31. Added a use_postmaster feature to recipient callouts.
2464 32. Added the $body_zerocount variable, containing the number of binary zero
2465 bytes in the message body.
2467 33. The time of last modification of the "new" subdirectory is now used as the
2468 "mailbox time last read" when there is a quota error for a maildir
2471 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2473 35. Added +ignore_unknown as a special item in host lists.
2475 36. Code for decoding IPv6 addresses in host lists is now included, even if
2476 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2477 address was recognized as an IP address, but was then not correctly decoded
2478 into binary, causing unexpected and incorrect effects when compared with
2485 1. Very minor rewording of debugging text in manualroute to say "list of
2486 hosts" instead of "hostlist".
2488 2. If verify=header_syntax was set, and a header line with an unqualified
2489 address (no domain) and a large number of spaces between the end of the
2490 name and the colon was received, the reception process suffered a buffer
2491 overflow, and (when I tested it) crashed. This was caused by some obsolete
2492 code that should have been removed. The fix is to remove it!
2494 3. When running in the test harness, delay a bit after writing a bounce
2495 message to get a bit more predictability in the log output.
2497 4. Added a call to search_tidyup() just before forking a reception process. In
2498 theory, someone could use a lookup in the expansion of smtp_accept_max_
2499 per_host which, without the tidyup, could leave open a database connection.
2501 5. Added the variables $recipient_data and $sender_data which get set from a
2502 lookup success in an ACL "recipients" or "senders" condition, or a router
2503 "senders" option, similar to $domain_data and $local_part_data.
2505 6. Moved the writing of debug_print from before to after the "senders" test
2508 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2509 problems for message scanning, either using a data ACL, or using
2510 local_scan() because the Received: header was not generated till after they
2511 were called (in order to set the time as the time of reception completion).
2512 I have revised the way this works. The header is now generated after the
2513 body is received, but before the ACL or local_scan() are called. After they
2514 are run, the timestamp in the header is updated.
2520 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2521 before starting a queue runner without re-exec. This happened only when
2522 deliver_drop_privilege was set or when the Exim user was set to root. The
2523 effect of the bug was that timeouts during subsequent deliveries caused
2524 crashes instead of being properly handled. The handler is now left at its
2525 default (and expected) setting.
2527 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2528 message, again when deliver_drop_privilege is set or Exim is run as root.
2529 The bug described in (1) was not present in this case, but the tidying up
2530 of the other signals was missing. I have made the two cases consistent.
2532 3. The ignore_target_hosts setting on a manualroute router was being ignored
2533 for hosts that were looked up using the /MX notation.
2535 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2538 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2539 operated on the sender address. After changing the $sender_address to <>
2540 for the sender address verify, Exim was re-instated it as the original
2541 (before rewriting) address, but remembering that it had rewritten it, so it
2542 wasn't rewriting it again. This bug also had the effect of breaking the
2543 sender address verification caching when the sender address was rewritten.
2545 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2546 This has been changed so that if the ip literal address matches
2547 ignore_target_hosts, the router declines.
2549 7. Added expansion conditions match_domain, match_address, and match_local_
2550 part (NOT match_host).
2552 8. The placeholder for the Received: header didn't have a length field set.
2554 9. Added code to Exim itself and to exim_lock to test for a specific race
2555 condition that could lead to file corruption when using MBX delivery. The
2556 issue is with the lockfile that is created in /tmp. If this file is removed
2557 after a process has opened it but before that process has acquired a lock,
2558 there is the potential for a second process to recreate the file and also
2559 acquire a lock. This could lead to two Exim processes writing to the file
2560 at the same time. The added code performs the same test as UW imapd; it
2561 checks after acquiring the lock that its file descriptor still refers to
2562 the same named file.
2564 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2565 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2568 11. Added the smtp_active_hostname option. If used, this will typically be made
2569 to depend on the incoming interface address. Because $interface_address is
2570 not set up until the daemon has forked a reception process, error responses
2571 that can happen earlier (such as "too many connections") no longer contain
2574 12. If an expansion in a condition on a "warn" statement fails because a lookup
2575 defers, the "warn" statement is abandoned, and the next ACL statement is
2576 processed. Previously this caused the whole ACL to be aborted.
2578 13. Added the iplsearch lookup type.
2580 14. Added ident_timeout as a log selector.
2582 15. Added tls_certificate_verified as a log selector.
2584 16. Added a global option tls_require_ciphers (compare the smtp transport
2585 option of the same name). This controls incoming TLS connections.
2587 17. I finally figured out how to make tls_require_ciphers do a similar thing
2588 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2589 before starting the TLS session.
2591 18. Tabs are now shown as \t in -bP output.
2593 19. If the log selector return_path_on_delivery was set, Exim crashed when
2594 bouncing a message because it had too many Received: header lines.
2596 20. If two routers both had headers_remove settings, and the first one included
2597 a superfluous trailing colon, the final name in the first list and the
2598 first name in the second list were incorrectly joined into one item (with a
2599 colon in the middle).
2605 1. Added -C and -D options to the exinext utility, mainly to make it easier
2606 to include in the automated testing, but these could be helpful when
2607 multiple configurations are in use.
2609 2. The exinext utility was not formatting the output nicely when there was
2610 an alternate port involved in the retry record key, nor when there was a
2611 message id as well (for retries that were specific to a specific message
2612 and a specific host). It was also confused by IPv6 addresses, because of
2613 the additional colons they contain. I have fixed the IPv4 problem, and
2614 patched it up to do a reasonable job for IPv6.
2616 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2617 delivery, the log line now contains "pipelined" if PIPELINING was used.
2619 4. An SMTP transport process used to panic and die if the bind() call to set
2620 an explicit outgoing interface failed. This has been changed; it is now
2621 treated in the same way as a connect() failure.
2623 5. A reference to $sender_host_name in the part of a conditional expansion
2624 that was being skipped was still causing a DNS lookup. This no longer
2627 6. The def: expansion condition was not recognizing references to header lines
2628 that used bh_ and bheader_.
2630 7. Added the _cache feature to named lists.
2632 8. The code for checking quota_filecount in the appendfile transport was
2633 allowing one more file than it should have been.
2635 9. For compatibility with Sendmail, the command line option
2643 and sets the incoming protocol and host name (for trusted callers). The
2644 host name and its colon can be omitted when only the protocol is to be set.
2645 Note the Exim already has two private options, -pd and -ps, that refer to
2646 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2647 "s", but I don't think that's a major issue.
2649 10. A number of refactoring changes to the code, none of which should affect
2652 (a) The number of logging options was getting close to filling up the
2653 32-bit word that was used as a bit map. I have split them into two classes:
2654 those that are passed in the argument to log_write(), and those that are
2655 only ever tested independently outside of that function. These are now in
2656 separate 32-bit words, so there is plenty of room for expansion again.
2657 There is no change in the user interface or the logging behaviour.
2659 (b) When building, for example, log lines, the code previously used a
2660 macro that called string_cat() twice, in order to add two strings. This is
2661 not really sufficiently general. Furthermore, there was one instance where
2662 it was actually wrong because one of the argument was used twice, and in
2663 one call a function was used. (As it happened, calling the function twice
2664 did not affect the overall behaviour.) The macro has been replaced by a
2665 function that can join an arbitrary number of extra strings onto a growing
2668 (c) The code for expansion conditions now uses a table and a binary chop
2669 instead of a serial search (which was left over from when there were very
2670 few conditions). Also, it now recognizes conditions like "pam" even when
2671 the relevant support is not compiled in: a suitably worded error message is
2672 given if an attempt is made to use such a condition.
2674 11. Added ${time_interval:xxxxx}.
2676 12. A bug was causing one of the ddress fields not to be passed back correctly
2677 from remote delivery subprocesses. The field in question was not being
2678 subsequently used, so this caused to problems in practice.
2680 13. Added new log selectors queue_time and deliver_time.
2682 14. Might have fixed a bug in maildirsizefile handling that threw up
2683 "unexpected character" debug warnings, and recalculated the data
2684 unnecessarily. In any case, I expanded the warning message to give more
2687 15. Added the message "Restricted characters in address" to the statements in
2688 the default ACL that block characters like @ and % in local parts.
2690 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2691 Three changes have been made:
2693 (a) There was a serious bug; a negative response to MAIL caused the whole
2694 recipient domain to be cached as invalid, thereby blocking all messages
2695 to all local parts at the same domain, from all senders. This bug has
2696 been fixed. The domain is no longer cached after a negative response to
2697 MAIL if the sender used is not empty.
2699 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2702 (c) A new callout option, "use_sender" has been added for people who want
2703 the modified behaviour.
2709 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2712 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2713 indeed breaks things for older releases.
2715 3. Added additional logging to the case where there is a problem reading data
2716 from a filter that is running in a subprocess using a pipe, in order to
2717 try to track down a specific problem.
2719 4. Testing facility fudge: when running in the test harness and attempting
2720 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2721 getting "No route to host". Convert this to a timeout.
2723 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2726 6. Some OS don't have socklen_t but use size_t instead. This affects the
2727 fifth argument of getsockopt() amongst other things. This is now
2728 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2729 can be set for individual OS. I have set it for SunOS5, OSF1, and
2730 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2731 some earlier ones do not.
2733 7. Change 4.30/15 was not doing the test caselessly.
2735 8. The standard form for an IPv6 address literal was being rejected by address
2736 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2737 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2738 this, as well as the form without the "IPv6" on the front (but only when
2739 address literals are enabled, of course).
2741 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2743 10. Exim crashed if a message with an empty sender address specified by -f
2744 encountered a router with an errors_to setting. This could be provoked only
2745 by a command such as
2749 where an empty string was supplied; "<>" did not hit this bug.
2751 11. Installed PCRE release 4.5.
2753 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2754 remained set. It is now erased.
2756 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2757 times from message ids (which are base 36 rather than the normal 62).
2759 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2760 were being counted as actual protocol errors, and logged if the log
2761 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2762 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2763 and DATA following a set of rejected RCPTs do not count as protocol errors.
2764 In other words, Exim assumes they were pipelined, though this may not
2765 actually be the case. Of course, in all cases the client gets an
2766 appropriate error code.
2768 15. If a lookup fails in an ACL condition, a message about the failure may
2769 be available; it is used if testing the ACL cannot continue, because most
2770 such messages specify what the cause of the deferral is. However, some
2771 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2772 that caused an old message to be retained and used if a later statement
2773 caused a defer, replacing the real cause of the deferral.
2775 16. If an IP address had so many PTR records that the DNS lookup buffer
2776 was not large enough to hold them, Exim could crash while trying to process
2777 the truncated data. It now detects and logs this case.
2779 17. Further to 4.21/58, another change has been made: if (and only if) the
2780 first line of a message (the first header line) ends with CRLF, a bare LF
2781 in a subsequent header line has a space inserted after it, so as not to
2782 terminate the header.
2784 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2785 unnecessarily, used atoi() instead of strtol(), and didn't check the
2786 termination when getting file sizes from file names by regex.
2788 19. Completely re-implemented the support for maildirsize files, in the light
2789 of a number of problems with the previous contributed implementation
2790 (4.30/29). In particular:
2792 . If the quota is zero, the maildirsize file is maintained, but no quota is
2795 . If the maildir directory does not exist, it is created before any attempt
2796 to write a maildirsize file.
2798 . The quota value in the file is just a cache; if the quota is changed in
2799 the transport, the new value overrides.
2801 . A regular expression is available for excluding directories from the
2804 20. The autoreply transport checks the characters in options that define the
2805 message's headers; it allows continued headers, but it was checking with
2806 isspace() after an embedded newline instead of explicitly looking for a
2809 21. If all the "regular" hosts to which an address was routed had passed their
2810 expiry times, and had not reached their retry times, the address was
2811 bounced, even if fallback hosts were defined. Now Exim should go on to try
2814 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2815 equivalent code in the SMTP transport. Some hosts send humungous responses
2816 to HELO/EHLO, more than 1024 it seems.
2818 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2819 gives compiler warnings in some environments. I've now done it "properly",
2822 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2823 (because of problems with the built-in one) was declared to return uschar *
2824 instead of char *, causing compiler failure.
2826 25. Fixed a file descriptor leak when processing alias/forward files.
2828 26. Fixed a minor format string issue in dbfn.c.
2830 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2832 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2833 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2836 29. When a sender address is verified, it is cached, to save repeating the test
2837 when there is more than one recipient in a message. However, when the
2838 verification involves a callout, it is possible for different callout
2839 options to be set for different recipients. It is too complicated to keep
2840 track of this in the cache, so now Exim always runs a verification when a
2841 callout is required, relying on the callout cache for the optimization.
2842 The overhead is duplication of the address routing, but this should not be
2845 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2846 to be verified with callout=postmaster, and the main callout worked but the
2847 postmaster check failed, the verification correctly failed. However, if a
2848 subsequent RCPT command asked for sender verification *without* the
2849 postmaster check, incorrect caching caused this verification also to fail,
2852 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2853 it was not caching the DNS options (qualify_single, search_parents) that
2854 were used when the lookup failed. A subsequent lookup with different
2855 options therefore always gave the same answer, though there were cases
2856 where it should not have. (Example: a "domains = !$mx_any" option on a
2857 dnslookup router: the "domains" option is always processed without any
2858 widening, but the router might have qualify_single set.) Now Exim uses the
2859 cached value only when the same options are set.
2861 32. Added John Jetmore's "exipick" utility to the distribution.
2863 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2864 than a timeout (e.g. a certificate is required, and is not provided), an
2865 Exim server now closes the connection immediately. Previously it waited for
2866 the client to close - but if the client is SSL, it seems that they each
2867 wait for each other, leading to a delay before one of them times out.
2869 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2870 maintained 0.8.x compatibility because I don't think many are using it, and
2871 it is clearly obsolete.
2873 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2876 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2877 client certificate was expired. A simple patch fixes this, though I don't
2878 understand the full logic of why the verify callback is called multiple
2881 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2882 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2883 which causes problems with some clients (such as the Certicom SSL Plus
2884 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2885 disables the coutermeasure allowing Eudora to connect."
2887 38. Exim was not checking that a write() to a log file succeeded. This could
2888 lead to Bad Things if a log got too big, in particular if it hit a file
2889 size limit. Exim now panics and dies if it cannot write to a log file, just
2890 as it does if it cannot open a log file.
2892 39. Modified OS/Makefile-Linux so that it now contains
2894 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2896 The two -D definitions ensure that Exim is compiled with large file
2897 support, which makes it possible to handle log files that are bigger than
2900 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2901 instance) a domain was checked against a named list that involved a lookup,
2902 causing $domain_data to be set, then another domain was checked against the
2903 same list, then the first domain was re-checked, the value of $domain_data
2904 after the final check could be wrong. In particular, if the second check
2905 failed, it could be set empty. This bug probably also applied to
2908 41. The strip_trailing_dot option was not being applied to the address given
2909 with the -f command-line option.
2911 42. The code for reading a message's header from the spool was incrementing
2912 $received_count, but never initializing it. This meant that the value was
2913 incorrect (doubled) while delivering a message in the same process in which
2914 it was received. In the most common configuration of Exim, this never
2915 happens - a fresh exec is done - but it can happen when
2916 deliver_drop_privilege is set.
2918 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2919 it now includes up to 150 characters of the unexpected data in the log
2922 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2923 and building data strings. The size of both of these buffers was 10 000
2924 bytes - far larger than anybody would *ever* want, thought I. Needless to
2925 say, somebody hit the limit. I have increased the maximum line length to
2926 20 000 and the maximum data length of concatenated lines to 100 000. I have
2927 also fixed two bugs, because there was no checking on these buffers. Tsk,
2928 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2929 buffer is too small.
2931 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2932 lsearch lookups. Now it does.
2934 46. When parsing a route_list item in a manualroute router, a fixed-length
2935 buffer was used for the list of hosts. I made this 1024 bytes long,
2936 thinking that nobody would ever have a list of hosts that long. Wrong.
2937 Somebody had a whole pile of complicated expansion conditions, and the
2938 string was silently truncated, leading to an expansion error. It turns out
2939 that it is easier to change to an unlimited length (owing to other changes
2940 that have happened since this code was originally written) than to build
2941 structure for giving a limitation error. The length of the item that
2942 expands into the list of hosts is now unlimited.
2944 47. The lsearch lookup could not handle data where the length of text line was
2945 more than 4095 characters. Such lines were truncated, leading to shortened
2946 data being returned. It should now handle lines of any length.
2948 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2949 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2952 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2953 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2954 no longer happens in comment lines. A list of replacements is now placed
2955 at the head of all of the source files, except those whose only change is
2956 to replace PERL_COMMAND in the very first #! line.
2958 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2959 messages on the queue, with a bottom-up merge sort, using code contributed
2960 by Michael Haardt. This should make operations like -bp somewhat faster on
2961 large queues. It won't affect queue runners, except when queue_run_in_order
2964 51. Installed eximstats 1.31 in the distribution.
2966 52. Added support for SRV lookups to the dnslookup router.
2968 53. If an ACL referred to $message_body or $message_body_end, the value was not
2969 reset for any messages that followed in the same SMTP session.
2971 54. The store-handling optimization for building very long strings was not
2972 differentiating between the different store pools. I don't think this
2973 actually made any difference in practice, but I've tidied it.
2975 55. While running the routers to verify a sender address, $sender_address
2976 was still set to the sender address. This is wrong, because when routing to
2977 send a bounce to the sender, it would be empty. Therefore, I have changed
2978 it so that, while verifying a sender address, $sender_address is set to <>.
2979 (There is no change to what happens when verifying a recipient address.)
2981 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2982 target A or AAAA records (if not already returned) without resetting the
2983 qualify_single or search_parents options of the DNS resolver. These are
2984 inappropriate in this case because the targets of MX and SRV records must
2985 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2986 target that, when qualified, matched something in the local domain. These
2987 two options are now turned off when doing these lookups.
2989 57. It seems that at least some releases of Reiserfs (which does not have the
2990 concept of a fixed number of inodes) returns zero and not -1 for the
2991 number of available inodes. This interacted badly with check_spool_inodes,
2992 which assumed that -1 was the "no such thing" setting. What I have done is
2993 to check that the total number of inodes is greater than zero before doing
2994 the test of how many are available.
2996 58. When a "warn" ACL statement has a log_message modifier, the message is
2997 remembered, and not repeated. This is to avoid a lot of repetition when a
2998 message has many recipients that cause the same warning to be written.
2999 Howewer, Exim was preserving the list of already written lines for an
3000 entire SMTP session, which doesn't seem right. The memory is now reset if a
3001 new message is started.
3003 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3004 debugging output unless log_rewrite was also set.
3006 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3007 of (int)(handle) when we know that handle contains (void *)(-1).
3009 61. The Exim daemon panic-logs an error return when it closes the incoming
3010 connection. However "connection reset by peer" seems to be common, and
3011 isn't really an error worthy of noting specially, so that particular error
3014 62. When Exim is trying to find all the local interfaces, it used to panic and
3015 die if the ioctl to get the interface flags failed. However, it seems that
3016 on at least one OS (Solaris 9) it is possible to have an interface that is
3017 included in the list of interfaces, but for which you get a failure error
3018 for this call. This happens when the interface is not "plumbed" into a
3019 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3020 failure of the "get flags" call assumes that the interface is down.
3022 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3023 makes life easier for people who are doing arithmetic on fields extracted
3024 from dates, where you often get leading zeros that should not be
3025 interpreted as octal.
3027 64. Added qualify_domain to the redirect router, to override the global
3030 65. If a pathologically long header line contained very many addresses (the
3031 report of this problem mentioned 10 000) and each of them was rewritten,
3032 Exim could use up a very large amount of memory. (It kept on making new
3033 copies of the header line as it rewrote, and never released the old ones.)
3034 At the expense of a bit more processing, the header rewriting function has
3035 been changed so that it no longer eats memory in this way.
3037 66. The generation of the Received: header has been moved from the time that a
3038 message starts to be received, to the time that it finishes. The timestamp
3039 in the Received: header should now be very close to that of the <= log
3040 line. There are two side-effects of this change:
3042 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3043 logged header lines no longer include the local Received: line, because
3044 it has not yet been created. The same applies to a copy of the message
3045 that is returned to a non-SMTP sender when a message is rejected.
3047 (b) When a filter file is tested using -bf, no additional Received: header
3048 is added to the test message. After some thought, I decided that this
3051 This change does not affect the value of $received_for. It is still set
3052 after address rewriting, but before local_scan() is called.
3054 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3056 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3057 gave an unhelpful panic error message, and a defer error. I have managed to
3058 change this behaviour so that it now rejects any supplied certificate,
3059 which seems right, as the list of acceptable certificates is empty.
3061 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3062 gave an unhelpful defer error. I have not managed to make this reject any
3063 supplied certificates, but the error message it gives is "no certificate
3064 supplied", which is not helpful.
3066 70. exigrep's output now also includes lines that are not associated with any
3067 message, but which match the given pattern. Implemented by a patch from
3068 Martin Sluka, which also tidied up the Perl a bit.
3070 71. Recipient callout verification, like sender verification, was using <> in
3071 the MAIL FROM command. This isn't really the right thing, since the actual
3072 sender may affect whether the remote host accepts the recipient or not. I
3073 have changed it to use the actual sender in the callout; this means that
3074 the cache record is now keyed on a recipient/sender pair, not just the
3075 recipient address. There doesn't seem to be a real danger of callout loops,
3076 since a callout by the remote host to check the sender would use <>.
3077 [SEE ABOVE: changed after hitting problems.]
3079 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3080 temporary errors. However, in the case of such a code being given after
3081 the end of a data transmission (i.e. after ".") Exim was failing to write
3082 a retry record for the message. (Yes, there was some broken host that was
3083 actually sending 8xx at this point.)
3085 73. An unknown lookup type in a host list could cause Exim to panic-die when
3086 the list was checked. (An example that provoked this was putting <; in the
3087 middle of a list instead of at the start.) If this happened during a DATA
3088 ACL check, a -D file could be left lying around. This kind of configuration
3089 error no longer causes Exim to die; instead it causes a defer errror. The
3090 incident is still logged to the main and panic logs.
3092 74. Buglet left over from Exim 3 conversion. The message "too many messages
3093 in one connection" was written to the rejectlog but not the mainlog, except
3094 when address rewriting (yes!) was being logged.
3096 75. Added write_rejectlog option.
3098 76. When a system filter was run not as root (that is, when system_filter_user
3099 was set), the values of the $n variables were not being returned to the
3100 main process; thus, they were not subsequently available in the $sn
3103 77. Added +return_path_on_delivery log selector.
3105 78. A connection timeout was being treated differently from recipients deferred
3106 when testing hosts_max_try with a message that was older than the host's
3107 retry timeout. (The host should not be counted, thus allowing all hosts to
3108 be tried at least once before bouncing.) This may have been the cause of an
3109 occasionally reported bug whereby a message would remain on the queue
3110 longer than the retry timeout, but would be bounced if a delivery was
3111 forced. I say "may" because I never totally pinned down the problem;
3112 setting up timeout/retry tests is difficult. See also the next item.
3114 79. The ultimate address timeout was not being applied to errors that involved
3115 a combination of host plus message (for example, a timeout on a MAIL
3116 command). When an address resolved to a number of possible hosts, and they
3117 were not all tried for each delivery (e.g. because of hosts_max_try), a
3118 message could remain on the queue longer than the retry timeout.
3120 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3123 81. Fixed an obscure SMTP outgoing bug which required at least the following
3124 conditions: (a) there was another message waiting for the same server;
3125 (b) the server returned 5xx to all RCPT commands in the first message so
3126 that the message was not completed; (c) the server dropped the connection
3127 or gave a negative response to the RSET that Exim sends to abort the
3128 transaction. The observed case was a dropped connection after DATA that had
3129 been sent in pipelining mode. That is, the server had advertised PIPELINING
3130 but was not implementing it correctly. The effect of the bug was incorrect
3131 behaviour, such as trying another host, and this could lead to a crash.
3137 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3138 and daemon.c were passed as pointers to ints; they should have been
3139 pointers to socklen_t variables (which are typically unsigned ints).
3141 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3144 3. Fixed a really odd bug that affected only the testing scheme; patching a
3145 certain fixed string in the binary changed the value of another string that
3146 happened to be identical to the end of the original first string.
3148 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3149 name", it returns that address as the IP address. On some operating
3150 systems (e.g. Solaris), it also passes back the IP address string as the
3151 "host name". However, on others (e.g. Linux), it passes back an empty
3152 string. Exim wasn't checking for this, and was changing the host name to an
3153 empty string, assuming it had been canonicized.
3155 5. Although rare, it is permitted to have more than one PTR record for a given
3156 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3157 all the names associated with an address, because they do in Solaris.
3158 However, it seems that they do not in Linux for data that comes from the
3159 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3160 I found this out when I moved to a new Linux workstation and tried to run
3161 the Exim test suite.
3163 To get round this problem I have changed the code so that it now does its
3164 own call to the DNS to look up PTR records when searching for a host name.
3165 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3166 addresses that are only in /etc/hosts are still found.
3168 This behaviour is, however, controlled by an option called host_lookup_
3169 order, which defaults to "bydns:byaddr". If people want to use the other
3170 order, or indeed, just use one or the other means of lookup, they can
3171 specify it in this variable.
3173 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3174 some operating systems, this comes back from gethostbyaddr() as an empty
3175 string, and this is what Exim used to test for. However, it seems that in
3176 other systems, "." is yielded. Exim now tests for this case too.
3178 7. The values of check_spool_space and check_log_space are now held internally
3179 as a number of kilobytes instead of an absolute number of bytes. If a
3180 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3181 kilobyte. This means that much larger values can be stored.
3183 8. Exim monitor: an attempt to get the action menu when not actually pointing
3184 at a message produces an empty menu entitled "No message selected". This
3185 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3186 no entries in it ("Shell widget menu has zero width and/or height"). So I
3187 have added a single, blank menu entry in this case.
3189 9. Added ${quote_local_part.
3191 10. MIME decoding is now applied to the contents of Subject: header lines when
3194 11. Now that a reference to $sender_host_address automatically causes a reverse
3195 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3196 host lookup before query-style lookups in lists that might use this
3197 variable. This has therefore been abolished, and the "net-" prefix is no
3198 longer necessary for query-style lookups.
3200 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3201 have been a typo for LFLAGS, so it has been changed.
3203 13. The install script calls Exim with "-C /dev/null" in order to find the
3204 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3205 to be output. Howeve, since Exim outputs its version number before the
3206 error, it didn't break the script. It just looked ugly. I fixed this by
3207 always allowing "-C /dev/null" if the caller is root.
3209 14. Ignore overlarge ACL variable number when reading spool file - insurance
3210 against a later release with more variables having written the file.
3212 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3213 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3214 this, as well as the form without the "IPv6" on the front.
3216 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3217 OS/Makefile-Darwin file.
3219 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3220 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3222 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3223 variable. However, some OS use ints for inodes. I've added cast to long int
3224 to get rid of the compiler warning.
3226 19. I had forgotten to lock out "/../" in configuration file names when
3227 ALT_CONFIG_PREFIX was set.
3229 20. Routers used for verification do not need to specify transports. However,
3230 if such a router generated a host list, and callout was configured, Exim
3231 crashed, because it could not find a port number from the (non-existent)
3232 transport. It now assumes port 25 in this circumstance.
3234 21. Added the -t option to exigrep.
3236 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3237 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3238 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3239 them is set, LOOKUP_LSEARCH is forced.
3241 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3243 24. Added sender and host information to the "rejected by local_scan()" log
3244 line; previously there was no indication of these.
3246 25. Added .include_if_exists.
3248 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3249 Linux. It turns out that not all file systems support this. Apparently some
3250 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3251 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3252 not supported on the file descriptor, is now ignored when Exim is trying to
3253 sync a directory. This applies only to Linux.
3255 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3257 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3258 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3261 29. Added support for maildirsize files from supplied patch (modified a bit).
3263 30. The use of :fail: followed by an empty string could lead Exim to respond to
3264 sender verification failures with (e.g.):
3266 550 Verification failed for <xxx>
3267 550 Sender verify failed
3269 where the first response line was missing the '-' that indicates it is not
3270 the final line of the response.
3272 31. The loop for finding the name of the user that called Exim had a hardwired
3273 limit of 10; it now uses the value of finduser_retries, which is used for
3274 all other user lookups.
3276 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3279 33. Exim was neglecting to zero errno before one call of strtol() when
3280 expanding a string and expecting an integer value. On some systems this
3281 resulted in spurious "integer overflow" errors. Also, it was casting the
3282 result into an int without checking.
3284 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3285 did not work. The code looks as if it has *never* worked, though it appears
3286 to have been documented since at least releast 1.62. I have made it work.
3288 35. The "timeout_DNS" error in retry rules, also documented since at least
3289 1.62, also never worked. As it isn't clear exactly what this means, and
3290 clearly it isn't a major issue, I have abolished the feature by treating it
3291 as "timeout", and writing a warning to the main and panic logs.
3293 36. The display of retry rules for -brt wasn't always showing the error code
3296 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3297 timeout_connect_A, timeout_connect_MX.
3299 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3300 to the empty sender.
3302 39. The daemon was not analysing the content of -oX till after it had closed
3303 stderr and disconnected from the controlling terminal. This meant that any
3304 syntax errors were only noted on the panic log, and the return code from
3305 the command was 0. By re-arranging the code a little, I've made the
3306 decoding happen first, so such errors now appear on stderr, and the return
3307 code is 1. However, the actual setting up of the sockets still happens in
3308 the disconnected process, so errors there are still only recorded on the
3311 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3312 connections (as happens on some IP stacks) was logged at start up time as
3313 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3314 it from "IPv6 and IPv4", which means that two separate sockets are being
3317 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3318 says whether AF_INET or AF_INET6 was passed as an argument.
3320 42. Exiwhat output was messed up when time zones were included in log
3323 43. Exiwhat now gives more information about the daemon's listening ports,
3324 and whether -tls-on-connect was used.
3326 44. The "port" option of the smtp transport is now expanded.
3328 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3329 silently ignored. Now an error message is written to the main and panic
3332 46. There's a new ACL modifier called "logwrite" which writes to a log file
3333 as soon as it is encountered.
3335 47. Added $local_user_uid and $local_user_gid at routing time.
3337 48. Exim crashed when trying to verify a sender address that was being
3340 49. Exim was recognizing only a space character after ".include". It now also
3341 recognizes a tab character.
3343 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3344 extracting the relevant information from the specification. The man page no
3345 longer contains scrambled data for the -d option, and I've added a section
3346 at the front about calling Exim under different names.
3348 51. Added "extra_headers" argument to the "mail" command in filter files.
3350 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3353 53. Installed eximstats 1.29.
3355 54. Added transport_filter_timeout as a generic transport option.
3357 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3358 Cc: header lines. This was required by RFC 822, but it not required by RFC
3361 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3362 incoming messages that did not have them. Now it does so only if the
3363 message originates locally, that is, if there is no associated remote host
3364 address. When Resent- header lines are present, this applies to the Resent-
3365 lines rather than the non-Resent- lines.
3367 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3368 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3370 58. Messages for configuration errors now include the name of the main
3371 configuration file - useful now that there may be more than one file in a
3372 list (.included file names were always shown).
3374 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3375 for those rare installations that do not start the daemon as root or run it
3376 setuid root. I've cut out the call to initgroups() if the daemon is not
3379 60. The Exim user and group can now be bound into the binary as text strings
3380 that are looked up at the start of Exim's processing.
3382 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3384 62. Added $mailstore_basename variable.
3386 63. Installed patch to sieve.c from Michael Haardt.
3388 64. When Exim failed to open the panic log after failing to open the main log,
3389 the original message it was trying to log was written to stderr and debug
3390 output, but if they were not available (the usual case in production), it
3391 was lost. Now it is written to syslog before the two lines that record the
3392 failures to open the logs.
3394 65. Users' Exim filters run in subprocesses under the user's uid. It is
3395 possible for a "deliver" command or an alias in a "personal" command to
3396 provoke an address rewrite. If logging of address rewriting is configured,
3397 this fails because the process is not running as root or exim. There may be
3398 a better way of dealing with this, but for the moment (because 4.30 needs
3399 to be released), I have disabled address rewrite logging when running a
3400 filter in a non-root, non-exim process.
3406 1. The buildconfig auxiliary program wasn't quoting the value set for
3407 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3408 not defined. This bug was masked in 4.22 by the effect that was fixed in
3411 2. Some messages that were rejected after a message id was allocated were
3412 shown as "incomplete" by exigrep. It no longer does this for messages that
3413 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3415 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3416 have allow_domain_literals set, the ID did not get logged in the <= line.
3417 Domain literals are now always recognized in Message-ID: header lines.
3419 4. The first argument for a ${extract expansion item is the key name or field
3420 number. Leading and trailing spaces in this item were not being ignored,
3421 causing some misleading effects.
3423 5. When deliver_drop_privilege was set, single queue runner processes started
3424 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3425 same command in the process it spins off) were not dropping privilege.
3427 6. When the daemon running as "exim" started a queue runner, it always
3428 re-executed Exim in the spun-off process. This is a waste of effort when
3429 deliver_drop_privilege is set. The new process now just calls the
3430 queue-runner function directly.
3436 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3439 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3440 ignored. Though the use of -oP was forcing the writing of a pid file, it
3441 was always written to the default place.
3443 3. If the message "no IP address found for host xxxx" is generated during
3444 incoming verification, it is now followed by identification of the incoming
3445 connection (so you can more easily find what provoked it).
3447 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3449 5. Added some features to "harden" Exim a bit more against certain attacks:
3451 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3452 be put in Local/Makefile. This is like the never_users runtime option,
3453 but it cannot be overridden. The default setting is "root".
3455 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3456 prefix string with which any file named in a -C command line option
3459 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3460 is retained for -C and -D only if the caller of Exim is root. Without
3461 it, the exim user may also use -C and -D and retain privilege.
3463 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3464 command line option is disabled.
3466 6. Macro names set by the -D option must start with an upper case letter, just
3467 like macro names defined in the configuration file.
3469 7. Added "dereference=" facility to LDAP.
3471 8. Two instances of the typo "uknown" in the source files are fixed.
3473 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3474 the Configure-Makefile script screwed up while processing it.
3476 10. Incorporated PCRE 4.4.
3478 11. The SMTP synchronization check was not operating right at the start of an
3479 SMTP session. For example, it could not catch a HELO sent before the client
3480 waited for the greeting. There is now a check for outstanding input at the
3481 point when the greeting is written. Because of the duplex, asynchronous
3482 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3483 way, but not yet received, when the check is performed.
3485 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3486 on TCP/IP sockets, because this apparently causes some broken clients to
3489 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3490 unchanged) from the Cygwin maintainer.
3492 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3493 was supported instead of testing for mbx. Effectively a typo.
3495 15. The spa authenticator server code was not checking that the input it
3496 received was valid base64.
3498 16. The debug output line for the "set" modifier in ACLs was not showing the
3499 name of the variable that was being set.
3501 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3503 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3504 lookup on its own. Something else was needed to trigger the lookup. For
3505 example, a match in host_lookup or the need for a host name in a host list.
3506 Now, if $sender_host_name is referenced and the host name has not yet been
3507 looked up, a lookup is performed. If the lookup fails, the variable remains
3508 empty, and $host_lookup_failed is set to "1".
3510 19. Added "eqi" as a case-independent comparison operator.
3512 20. The saslauthd authentication condition could segfault if neither service
3513 nor realm was specified.
3515 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3516 error message that was logged was misleading, and incoming SMTP
3517 connections were dropped. The message is now more accurate, and temporary
3518 errors are given to SMTP connections.
3520 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3521 (except RSET) with a 421 error, until QUIT is received. However, it was
3522 failing to send a response to QUIT.
3524 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3525 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3526 ACL is now run after the helo_try_verify_hosts code.
3528 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3529 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3530 case-independent, so other case variants are also recognized). Apparently
3531 some systems use these upper case variants.
3533 25. If more than two messages were waiting for the same host, and a transport
3534 filter was specified for the transport, Exim sent two messages over the
3535 same TCP/IP connection, and then failed with "socket operation on non-
3536 socket" when it tried to send the third.
3538 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3540 27. The extern definition of crypt16() in expand.c was not being excluded when
3541 the OS had its own crypt16() function.
3543 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3544 as a preferred synonym for return_size_limit, both as an option and as an
3547 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3549 30. Changed the default configuration ACL to relax the local part checking rule
3550 for addresses that are not in any local domains. For these addresses,
3551 slashes and pipe symbols are allowed within local parts, but the sequence
3552 /../ is explicitly forbidden.
3554 31. SPA server authentication was not clearing the challenge buffer before
3557 32. log_message in a "warn" ACL statement was writing to the reject log as
3558 well as to the main log, which contradicts the documentation and doesn't
3559 seem right (because no rejection is happening). So I have stopped it.
3561 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3562 However, I am unable to do any testing of this.
3564 34. Fixed an infelicity in the appendfile transport. When checking directories
3565 for a mailbox, to see if any needed to be created, it was accidentally
3566 using path names with one or more superfluous leading slashes; tracing
3567 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3569 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3570 contents are added to the log line that is written for every discarded
3571 recipient. (Previously a log_message setting was ignored.)
3573 36. The ${quote: operator now quotes the string if it is empty.
3575 37. The install script runs exim in order to find its version number. If for
3576 some reason other than non-existence or emptiness, which it checks, it
3577 could not run './exim', it was installing it with an empty version number,
3578 i.e. as "exim-". This error state is now caught, and the installation is
3581 38. An argument was missing from the function that creates an error message
3582 when Exim fails to connect to the socket for saslauthd authentication.
3583 This could cause Exim to crash, or give a corrupted message.
3585 39. Added isip, isip4, and isip6 to ${if conditions.
3587 40. The ACL variables $acl_xx are now saved with the message, and can be
3588 accessed later in routers, transports, and filters.
3590 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3591 strings in the file are not string-expanded.
3593 42. If a MAIL command specified a SIZE value that was too large to fit into an
3594 int variable, the check against message_size_limit failed. Such values are
3595 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3596 day this will have to be increased, but I don't think I want to be around
3597 when emails are that large.
3604 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3605 iconv() is not standard in FreeBSD.
3607 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3608 IPv6 enabled. The observed symptom was a segmentation fault on return from
3609 the function os_common_find_running_interfaces() in src/os.c.
3611 3. In the check_special_case() function in daemon.c I had used "errno" as an
3612 argument name, which causes warnings on some systems. This was basically a
3613 typo, since it was named "eno" in the comments!
3615 4. The code that waits for the clock to tick (at a resolution of some fraction
3616 of a second) so as to ensure message-id uniqueness was always waiting for
3617 at least one whole tick, when it could have waited for less. [This is
3618 almost certainly not relevant at current processor speeds, where it is
3619 unlikely to ever wait at all. But we try to future-proof.]
3621 5. The function that sleeps for a time interval that includes fractions of a
3622 second contained a race. It did not block SIGALRM between setting the
3623 timer, and suspending (a couple of lines later). If the interval was short
3624 and the sigsuspend() was delayed until after it had expired, the suspension
3625 never ended. On busy systems this could lead to processes getting stuck for
3628 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3629 process, before it forks any delivery processes. The open lookup caching
3630 mechanism meant that the open file or database connection was passed into
3631 the delivery process. The problem was that delivery processes always tidy
3632 up cached lookup data. This could cause a problem for the next delivery
3633 process started by the queue runner, because the external queue runner
3634 process does not know about the closure. So the next delivery process
3635 still has data in the lookup cache. In the case of a file lookup, there was
3636 no problem because closing a file descriptor in a subprocess doesn't affect
3637 the parent. However, if the lookup was caching a connection to a database,
3638 the connection was closed, and the second delivery process was likely to
3639 see errors such as "PGSQL: query failed: server closed the connection
3640 unexpectedly". The problem has been fixed by closing all cached lookups
3641 in a queue runner before running a delivery process.
3643 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3644 seem to have the "const" qualifier which it has on other OS. I've
3647 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3648 *of the same type* (client or server) with the same public name that an
3649 error should be diagnosed.
3651 9. When Exim looked up a host name for an IP address, but failed to find the
3652 original IP address when looking up the host name (a safety check), it
3653 output the message "<ip address> does not match any IP for NULL", which was
3654 confusing, to say the least. The bug was that the host name should have
3655 appeared instead of "NULL".
3657 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3658 user that is built into the binary, and the -C or -D options is used, root
3659 privilege is dropped before the configuration file is read. In addition,
3660 logging is switched to stderr instead of the normal log files. If the
3661 configuration then re-defines the Exim user, the unprivileged environment
3662 is probably not what is expected, so Exim logs a panic warning message (but
3665 However, if deliver_drop_privilege is set, the unprivileged state may well
3666 be exactly what is intended, so the warning has been cut out in that case,
3667 and Exim is allowed to try to write to its normal log files.
3673 1. smtp_return_error_details was not giving details for temporary sender
3674 or receiver verification errors.
3676 2. Diagnose a configuration error if two authenticators have the same public
3679 3. Exim used not to create the message log file for a message until the first
3680 delivery attempt. This could be confusing when incoming messages were held
3681 for policy or load reasons. The message log file is now created at the time
3682 the message is received, and an initial "Received" line is written to it.
3684 4. The automatically generated man page for command line options had a minor
3685 bug that caused no ill effects; however, a more serious problem was that
3686 the procedure for building the man page automatically didn't always
3687 operate. Consequently, release 4.20 contains an out-of-date version. This
3688 shouldn't happen again.
3690 5. When building Exim with embedded Perl support, the script that builds the
3691 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3692 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3694 6. The freeze_tell option was not being used for messages that were frozen on
3695 arrival, either by an ACL or by local_scan().
3697 7. Added the smtp_incomplete_transaction log selector.
3699 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3700 was accepting AUTH without a new EHLO.
3702 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3703 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3706 10. Logging of TCP/IP connections (when configured) now happens in the main
3707 daemon process instead of the child process, so that the TCP/IP connection
3708 count is more accurate (but it can never be perfect).
3710 11. The use of "drop" in a nested ACL was not being handled correctly in the
3711 outer ACL. Now, if condition failure induced by the nested "drop" causes
3712 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3713 or "require"), the connection is dropped.
3715 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3716 that yield "discard" can now be used with an "accept" or a "discard" verb,
3717 but an error is generated for any others (because I can't see a useful way
3718 to define what should happen).
3720 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3721 are now processed in the same way as lines in the Exim configuration file.
3722 In particular, continuation lines are supported.
3724 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3726 15. Added -ti meaning -t -i.
3728 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3729 domains, and warn by logging if others are found.
3731 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3732 returned from GIFCONF when Exim is trying to find the list of interfaces on
3733 a host. The code in os.c has been modified to copy each ifreq to an aligned
3734 structure in all cases.
3736 Also, in some cases, the returned ifreq's were being copied to a 'struct
3737 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3738 means the last couple of bytes of an IPv6 address could be chopped if the
3739 ifreq contained only a normal sockaddr (14 bytes storage).
3741 18. Named domain lists were not supported in the hosts_treat_as_local option.
3742 An entry such as +xxxx was not recognized, and was treated as a literal
3745 19. Ensure that header lines added by a DATA ACL are included in the reject log
3746 if the ACL subsequently rejects the message.
3748 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3749 MD5 (which is deprecated).
3751 21. When testing a filter file using -bf, Exim was writing a message when it
3752 took the sender from a "From " line in the message, but it was not doing so
3753 when it took $return_path from a Return-Path: header line. It now does.
3755 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3756 with a valid header line field name (a series of printing characters
3757 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3759 23. Changed "disc" in the source to "disk" to conform to the documentation and
3760 the book and for uniformity.
3762 24. Ignore Sendmail's -Ooption=value command line item.
3764 25. When execve() failed while trying to run a command in a pipe transport,
3765 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3766 could be confused with a return value of 69 from the command itself. This
3767 has been changed to 127, the value the shell returns if it is asked to run
3768 a non-existent command. The wording for the related log line suggests a
3769 non-existent command as the problem.
3771 26. If received_header_text expands to an empty string, do not add a Received:
3772 header line to the message. (Well, it adds a token one on the spool, but
3773 marks it "old" so that it doesn't get used or transmitted.)
3775 27. Installed eximstats 1.28 (addition of -nt option).
3777 28. There was no check for failure on the call to getsockname() in the daemon
3778 code. This can fail if there is a shortage of resources on the system, with
3779 ENOMEM, for example. A temporary error is now given on failure.
3781 29. Contrary to the C standard, it seems that in some environments, the
3782 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3783 program. Exim now does this explicitly; it affects the formatting of
3784 timestamps using strftime().
3786 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3787 complaints. I've now initialized all the variables, to avoid this.
3789 32. Header lines added by a system filter were not being "seen" during
3790 transport-time rewrites.
3792 33. The info_callback() function passed to OpenSSL is set up with type void
3793 (*)(SSL *, int, int), as described somewhere. However, when calling the
3794 function (actually a macro) that sets it up, the type void(*)() is
3795 expected. I've put in a cast to prevent warnings from picky compilers.
3797 34. If a DNS black list lookup found a CNAME record, but there were no A
3798 records associated with the domain it pointed at, Exim crashed.
3800 35. If a DNS black list lookup returned more than one A record, Exim ignored
3801 all but the first. It now scans all returned addresses if a particular IP
3802 value is being sought. In this situation, the contents of the
3803 $dnslist_value variable are a list of all the addresses, separated by a
3806 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3807 to accept a host name and all its aliases if the forward lookup for any of
3808 them yielded the IP address of the incoming connection. Now it accepts only
3809 those names whose forward lookup yields the correct IP address. Any other
3810 names are discarded. This closes a loophole whereby a rogue DNS
3811 administrator could create reverse DNS records to break through a
3812 wildcarded host restriction in an ACL.
3814 37. If a user filter or a system filter that ran in a subprocess used any of
3815 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3816 the wrong values were passed to the pipe command ($thisaddress had the
3817 value of $0, $0 had the value of $1, etc). This bug was introduced by
3818 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3820 38. Improved the line breaking for long SMTP error messages from ACLs.
3821 Previously, if there was no break point between 40 and 75 characters, Exim
3822 left the rest of the message alone. Two changes have been made: (a) I've
3823 reduced the minimum length to 35 characters; (b) if it can't find a break
3824 point between 35 and 75 characters, it looks ahead and uses the first one
3825 that it finds. This may give the occasional overlong line, but at least the
3826 remaining text gets split now.
3828 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3829 file descriptors might be low, and that setting 1000 would always raise it.
3830 It turns out that in some environments, the limit is already over 1000 and
3831 that lowering it causes trouble. So now Exim takes care not to decrease it.
3833 40. When delivering a message, the value of $return_path is set to $sender_
3834 address at the start of routing (routers may change the value). By an
3835 oversight, this default was not being set up when an address was tested by
3836 -bt or -bv, which affected the outcome if any router or filter referred to
3839 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3840 log only when "message" or "log_message" are set. However, if one of the
3841 conditions was an address verification, or a call to a nested ACL, the
3842 messages generated by the underlying test were being passed through. This
3843 no longer happens. The underlying message is available in $acl_verify_
3844 message for both "message" and "log_message" expansions, so it can be
3845 passed through if needed.
3847 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3848 new expansion $bh_ to give the encoded byte string without charset
3849 translation. Translation happens only if iconv() is available; HAVE_ICONV
3850 indicates this at build time. HEADERS_CHARSET gives the charset to
3851 translate to; headers_charset can change it in the configuration, and
3852 "headers charset" can change it in an individual filter file.
3854 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3855 that creates RFC 2047 encoded "words" labels them as that charset instead
3856 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3857 expansion operator; (ii) when Exim creates a From: line for a local
3858 message; (iii) when a header line is rewritten to include a "phrase" part.
3860 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3861 buggy, causing it to skip the first lines of messages whose message ID
3862 ended in 'D'. This would not have bitten before Exim release 4.14, because
3863 message IDs were unlikely to end in 'D' before then. The effect was to have
3864 incorrect size information for certain domains.
3866 45. #include "config.h" was missing at the start of the crypt16.c module. This
3867 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3870 46. If there was a timeout during a "random" callout check, Exim treated it as
3871 a failure of the random address, and carried on sending RSET and the real
3872 address. If the delay was just some slowness somewhere, the response to the
3873 original RCPT would be taken as a response to RSET and so on, causing
3874 mayhem of various kinds.
3876 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3877 when I implemented it. It didn't allow for the fact that some option values
3878 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3879 the right test for positive values.
3881 48. Domain names in DNS records are case-independent. Exim always looks them up
3882 in lower case. Some resolvers return domain names in exactly the case they
3883 appear in the zone file, that is, they may contain uppercase letters. Not
3884 all resolvers do this - some return always lower case. Exim was treating a
3885 change of case by a resolver as a change of domain, similar to a widening
3886 of a domain abbreviation. This triggered its re-routing code and so it was
3887 trying to route what was effectively the same domain again. This normally
3888 caused routing to fail (because the router wouldn't handle the domain
3889 twice). Now Exim checks for this case specially, and just changes the
3890 casing of the domain that it ultimately uses when it transmits the message
3893 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3896 50. If a filter generated a file delivery with a non-absolute name (possible if
3897 no home directory exists for the router), the forbid_file option was not
3900 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3901 the existing equality matching.
3903 52. Exim was using ints instead of ino_t variables in some places where it was
3904 dealing with inode numbers.
3906 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3907 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3908 called TMPDIR, and if it finds it is different, it changes its value.
3910 54. The smtp_printf() function is now made available to local_scan() so
3911 additional output lines can be written before returning. There is also an
3912 smtp_fflush() function to enable the detection of a dropped connection.
3913 The variables smtp_input and smtp_batched_input are exported to
3916 55. Changed the default runtime configuration: the message "Unknown user"
3917 has been removed from the ACL, and instead placed on the localuser router,
3918 using the cannot_route_message feature. This means that any verification
3919 failures that generate their own messages won't get overridden. Similarly,
3920 the "Unrouteable address" message that was in the ACL for unverifiable
3921 relay addresses has also been removed.
3923 56. Added hosts_avoid_esmtp to the smtp transport.
3925 57. The exicyclog script was not checking for the esoteric option
3926 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3927 will work only if exicyclog is run under the appropriate euid.
3929 58. Following a discussion on the list, the rules by which Exim recognises line
3930 endings on incoming messages have been changed. The -dropcr and drop_cr
3931 options are now no-ops, retained only for backwards compatibility. The
3932 following line terminators are recognized: LF CRLF CR. However, special
3933 processing applies to CR:
3935 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3936 nor a local message in the state where . is a terminator.
3938 (ii) If a bare CR is encountered in a header line, an extra space is added
3939 after the line terminator so as not to end the header. The reasoning
3940 behind this is that bare CRs in header lines are most likely either
3941 to be mistakes, or people trying to play silly games.
3943 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3944 was being incorrectly given as 18 bytes larger than it should have been.
3945 This is a VOB (very old bug).
3947 60. This may never have affected anything current, but just in case it has:
3948 When the local host is found other than at the start of a list of hosts,
3949 the local host, those with the same MX, and any that follow, are discarded.
3950 When the list in question was part of a longer list of hosts, the following
3951 hosts (not currently being processed) were also being discarded. This no
3952 longer happens. I'm not sure if this situation could ever has previously
3955 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3958 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3959 header. This is something that is recommended in a new Internet Draft, and
3960 is something that is documented as being done by Sendmail. There are two
3961 possible values. For messages generated by the autoreply transport, Exim
3964 Auto-Submitted: auto-replied
3966 whereas for all other generated messages (e.g. bounces) it adds
3968 Auto-Submitted: auto-generated
3970 63. The "personal" condition in filters now includes a test for the
3971 Auto-Submitted: header. If it contains the string "auto-" the message it
3972 not considered personal.
3974 64. Added rcpt_include_affixes as a generic transport option.
3976 65. Added queue_only_override (default true).
3978 66. Added the syslog_duplication option.
3980 67. If what should have been the first header line of a message consisted of
3981 a space followed by a colon, Exim was mis-interpreting it as a header line.
3982 It isn't of course - it is syntactically invalid and should therefore be
3983 treated as the start of the message body. The misbehaviour could have
3984 caused a number of strange effects, including loss of data in subsequent
3985 header lines, and spool format errors.
3987 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3988 client host had authenticated. This control can now be exercised by an ACL
3989 for more flexibility.
3991 69. By default, callouts do not happen when testing with -bh. There is now a
3992 variant, -bhc, which does actually run the callout code, including
3993 consulting and updating the callout cache.
3995 70. Added support for saslauthd authentication, courtesy of Alexander
3998 71. If statvfs() failed on the spool or log directories while checking their
3999 size for availability, Exim confusingly gave the error "space shortage".
4000 Furthermore, in debugging mode it crashed with a floating point exception.
4001 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4002 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4003 serious problem, Exim now writes to the main and panic logs when this
4004 happens, with details of the failure. It then refuses to accept the
4005 incoming message, giving the message "spool directory problem" or "log
4006 directory problem" with a 421 code for SMTP messages.
4008 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4009 0, 1, and 2 exist, because some OS complain for execs without them (see
4010 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4011 descriptors. However, the code omitted to check that the open succeeded,
4012 causing mysterious errors if for some reason the permissions on /dev/null
4013 got screwed. Now Exim writes a message to the main and panic logs, and
4014 bombs out if it can't open /dev/null.
4016 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4017 interact so that it is all more flexible. It is supposed to remain
4018 backwards compatible. Also added extra_local_interfaces.
4020 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4021 to bomb out with an assertion failure - to the client this appears as a
4022 connection drop. This problem occurs in the part of the code that was taken
4023 from the Samba project. Fortunately, the assertion is in a very simple
4024 function, so I have fixed this by reproducing the function inline in the
4025 one place where it is called, and arranging for authentication to fail
4026 instead of killing the process with assert().
4028 75. The SPA client code was not working when the server requested OEM rather
4029 than Unicode encoding.
4031 76. Added code to make require_files with a specific uid setting more usable in
4032 the case where statting the file as root fails - usually a non-root-mounted
4033 NFS file system. When this happens and the failure is EACCES, Exim now
4034 forks a subprocess and does the per-uid checking as the relevant uid.
4036 77. Added process_log_path.
4038 78. If log_file_path was not explicitly set, a setting of check_log_space or
4039 check_log_inodes was ignored.
4041 79. If a space check for the spool or log partitions fails, the incident is now
4042 logged. Of course, in the latter case the data may get lost...
4044 80. Added the %p formatting code to string_format() so that it can be used to
4045 print addresses in debug_print(). Adjusted all the address printing in the
4046 debugging in store.c to use %p rather than %d.
4048 81. There was a concern that a line of code in smtp_in.c could overflow a
4049 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4050 initially expressed, the concern was not well-founded, because trailing
4051 spaces are removed early. However, if the trailing spaces were followed by
4052 a NULL, they did not get removed, so the overflow was possible. Two fixes
4055 (a) I re-wrote the offending code in a cleaner fashion.
4056 (b) If an incoming SMTP command contains a NULL character, it is rejected
4059 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4060 runs initgroups(), so that if the Exim user is in any additional groups,
4061 they will be used during message reception.
4067 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob