1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.356 2006/06/07 17:42:27 fanf2 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 with
24 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.
42 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
43 other effects) broke the use of negated acl sub-conditions.
45 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
48 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
49 "Deny" causes Exim to reject the incoming connection with a 554 error.
50 Unfortunately, if there is a major crisis, such as a disk failure,
51 tcp-wrappers gives "deny", whereas what one would like would be some
52 kind of temporary error. A kludge has been added to help with this.
53 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
54 554 error is used if errno is still zero or contains ENOENT (which occurs
55 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
58 PH/03 Add -lutil to the default FreeBSD LIBS setting.
60 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
61 errors. Otherwise a message that provokes a temporary error (when other
62 messages do not) can cause a whole host to time out.
64 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
65 addresses were routed directly to files or pipes from a redirect router.
66 File deliveries just didn't batch; pipe deliveries might have suffered
69 PH/06 A failure to get a lock for a hints database would erroneously always say
70 "Failed to get write lock", even when it was really a read lock.
72 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
73 of 0600. This has been changed to use the value of the lockfile_mode
74 option (which defaults to 0600).
76 PH/08 Applied small patch from the Sieve maintainer.
78 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
79 folder from quota calculations, a direct delivery into this folder messed
80 up the contents of the maildirsize file. This was because the regex was
81 used only to exclude .Trash (or whatever) when the size of the mailbox
82 was calculated. There was no check that a delivery was happening into an
83 excluded directory. This bug has been fixed by ignoring all quota
84 processing for deliveries into excluded directories.
86 PH/10 Added the maildirfolder_create_regex option to appendfile.
92 PH/01 The code for finding all the local interface addresses on a FreeBSD
93 system running IPv6 was broken. This may well have applied to all BSD
94 systems, as well as to others that have similar system calls. The broken
95 code found IPv4 interfaces correctly, but gave incorrect values for the
96 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
97 that it would not match correctly against @[] and not recognize the IPv6
100 PH/02 The ipliteral router was not recognizing addresses of the form user@
101 [ipv6:....] because it didn't know about the "ipv6:" prefix.
103 PH/03 Added disable_ipv6.
105 PH/04 Changed $reply_address to use the raw form of the headers instead of the
106 decoded form, because it is most often used to construct To: headers
107 lines in autoreplies, and the decoded form may well be syntactically
108 invalid. However, $reply_address has leading white space removed, and all
109 newlines turned into spaces so that the autoreply transport does not
112 PH/05 If group was specified without a user on a router, and no group or user
113 was specified on a transport, the group from the router was ignored.
115 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
116 for visible compile-time settings that can be used to change these
117 numbers, for those that want even more. Backwards compatibility with old
118 spool files has been maintained. However, going back to a previous Exim
119 release will lost any variables that are in spool files.
121 PH/07 Two small changes when running in the test harness: increase delay when
122 passing a TCP/IP connection to a new process, in case the original
123 process has to generate a bounce, and remove special handling of
124 127.0.0.2 (sic), which is no longer necessary.
126 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
127 be the same on different OS.
129 PH/09 Moved a debug statement in filter processing to avoid a race problem when
132 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
133 whether --show-vars was specified or not
135 JJ/02 exipick: Added support for new ACL variable spool format introduced
138 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
139 syntactically invalid From: or Reply-to: line, and a filter used this to
140 generate an autoreply, and therefore failed to obtain an address for the
141 autoreply, Exim could try to deliver to a non-existent relative file
142 name, causing unrelated and misleading errors. What now happens is that
143 it logs this as a hard delivery error, but does not attempt to create a
146 PH/11 The exinext utility has a -C option for testing purposes, but although
147 the given file was scanned by exinext itself; it wasn't being passed on
148 when Exim was called.
150 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
151 an end-of-file indication when reading a command response.
153 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
154 compiled. In many other places in Exim, IPv6 addresses are always
155 recognized, so I have changed this. It also means that IPv4 domain
156 literals of the form [IPV4:n.n.n.n] are now always recognized.
158 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
159 used if the router is not running as root, for example, when verifying at
160 ACL time, or when using -bh. The debugging output from this situation was
161 non-existent - all you got was a failure to exec. I have made two
164 (a) Failures to set uid/gid, the current directory, or a process leader
165 in a subprocess such as that created by queryprogram now generate
166 suitable debugging ouput when -d is set.
168 (b) The queryprogram router detects when it is not running as root,
169 outputs suitable debugging information if -d is set, and then runs
170 the subprocess without attempting to change uid/gid.
172 PH/15 Minor change to Makefile for building test_host (undocumented testing
175 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
176 additional section of a DNS packet that returns MX or SRV records.
177 Instead, it always explicitly searches for A/AAAA records. This avoids
178 major problems that occur when a DNS server includes only records of one
179 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
180 fixed another bug: if SRV records were looked up and the corresponding
181 address records were *not* found in the additional section, the port
182 values from the SRV records were lost.
184 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
185 using the correct key (the original address) when searching the retry
186 rules in order to find which one to use for generating the retry hint.
188 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
189 adding the default one. Similarly, if it contains a Reply-To: header, the
190 errors_reply_to option, if set, is not used.
192 PH/19 When calculating a retry time, Exim used to measure the "time since
193 failure" by looking at the "first failed" field in the retry record. Now
194 it does not use this if it is later than than the arrival time of the
195 message. Instead it uses the arrival time. This makes for better
196 behaviour in cases where some deliveries succeed, thus re-setting the
197 "first failed" field. An example is a quota failure for a huge message
198 when small messages continue to be delivered. Without this change, the
199 "time since failure" will always be short, possible causing more frequent
200 delivery attempts for the huge message than are intended.
201 [Note: This change was subsequently modified - see PH/04 for 4.62.]
203 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
204 $1, $2, $3) because the numerical variables can be reset during some
205 expansion items (e.g. "match"), thereby losing the authentication data.
207 PH/21 Make -bV show the size of off_t variables so that the test suite can
208 decide whether to run tests for quotas > 2G.
210 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
211 mailbox_size, and mailbox_filecount in the appendfile transport. If a
212 filecount value is greater than 2G or if a quota value is greater than 2G
213 on a system where the size of off_t is not greater than 4, a panic error
216 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
217 never match. The debug and -bh output now contains an explicit error
218 message indicating a malformed IPv4 address or mask.
220 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
221 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
224 PH/25 Do not write to syslog when running in the test harness. The only
225 occasion when this arises is a failure to open the main or panic logs
226 (for which there is an explicit test).
228 PH/26 Added the /no_tell option to "control=freeze".
230 PH/27 If a host name lookup failed very early in a connection, for example, if
231 the IP address matched host_lookup and the reverse lookup yielded a name
232 that did not have a forward lookup, an error message of the form "no IP
233 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
234 could be logged. Now it outputs the IP address instead of "NULL".
236 PH/28 An enabling patch from MH: add new function child_open_exim2() which
237 allows the sender and the authenticated sender to be set when
238 submitting a message from within Exim. Since child_open_exim() is
239 documented for local_scan(), the new function should be too.
241 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
242 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
243 results in an empty string is now treated as unset.
245 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
247 PH/31 Added sender_verify_fail logging option.
249 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
250 needed by RFC 822 but not by RFC 2822 was commented out. I have now
251 tidied the source and removed it altogether.
253 PH/33 When a queue run was abandoned because the load average was too high, a
254 log line was always written; now it is written only if the queue_run log
255 selector is set. In addition, the log line for abandonment now contains
256 information about the queue run such as the pid. This is always present
257 in "start" and "stop" lines but was omitted from the "abandon" line.
259 PH/34 Omit spaces between a header name and the colon in the error message that
260 is given when verify = headers_syntax fails (if there are lots of them,
261 the message gets confusing).
263 PH/35 Change the default for dns_check_names_pattern to allow slashes within
264 names, as there are now some PTR records that contain slashes. This check
265 is only to protect against broken name servers that fall over on strange
266 characters, so the fact that it applies to all lookups doesn't matter.
268 PH/36 Now that the new test suite is complete, we can remove some of the
269 special code in Exim that was needed for the old test suite. For example,
270 sorting DNS records because real resolvers return them in an arbitrary
271 order. The new test suite's fake resolver always returns records in the
274 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
275 bounces) except when queue_only is set, to avoid logging races between
276 the different processes.
278 PH/38 Panic-die if .include specifies a non-absolute path.
280 PH/39 A tweak to the "H" retry rule from its user.
282 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
283 a label. They prevented compilation on older perls.
285 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
286 a warning to be raised on newish perls.
288 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
289 on queue. Changes to match documented behaviour of showing count of
290 messages matching specified criteria.
292 PH/40 Changed the default ident timeout from 30s to 5s.
294 PH/41 Added support for the use of login_cap features, on those BSD systems
295 that have them, for controlling the resources used by pipe deliveries.
297 PH/42 The content-scanning code uses fopen() to create files in which to put
298 message data. Previously it was not paying any attention to the mode of
299 the files. Exim runs with umask(0) because the rest of the code creates
300 files with open(), and sets the required mode explicitly. Thus, these
301 files were ending up world-writeable. This was not a big issue, because,
302 being within the spool directory, they were not world-accessible. I have
303 created a function called modefopen, which takes an additional mode
304 argument. It sets umask(777), creates the file, chmods it to the required
305 mode, then resets the umask. All the relevant calls to fopen() in the
306 content scanning code have been changed to use this function.
308 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
309 to 24 hours. This avoids potential overflow problems when processing G
310 and H retry rules. I suspect nobody ever tinkers with this value.
312 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
314 PH/45 When the plaintext authenticator is running as a client, the server's
315 challenges are checked to ensure they are valid base64 strings. By
316 default, the authentication attempt is cancelled if an invalid string is
317 received. Setting client_ignore_invalid_base64 true ignores these errors.
318 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
319 they are received. Thus, the responses can be made to depend on the
320 challenges. If an invalid string is ignored, an empty string is placed in
323 PH/46 Messages that are created by the autoreply transport now contains a
324 References: header, in accordance with RFCs 2822 and 3834.
326 PH/47 Added authenticated_sender_force to the smtp transport.
328 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
330 PH/49 Installed latest patch from the Sieve maintainer.
332 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
333 was also set, the appendfile transport was unnecessarily scanning a
334 directory of message files (e.g. for maildir delivery) to find the count
335 of files (along with the size), even though it did not need this
336 information. It now does the scan only if it needs to find either the
337 size of the count of files.
339 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
341 PH/52 Two bugs concerned with error handling when the smtp transport is
344 (i) Exim was not creating retry information for temporary errors given
345 for individual recipients after the DATA command when the smtp transport
346 was used in LMTP mode. This meant that they could be retried too
347 frequently, and not timed out correctly.
349 (ii) Exim was setting the flag that allows error details to be returned
350 for LMTP errors on RCPT commands, but not for LMTP errors for individual
351 recipients that were returned after the DATA command.
353 PH/53 This is related to PH/52, but is more general: for any failing address,
354 when detailed error information was permitted to be returned to the
355 sender, but the error was temporary, then after the final timeout, only
356 "retry timeout exceeded" was returned. Now it returns the full error as
357 well as "retry timeout exceeded".
359 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
360 do this, and (what is worse) MTAs that accept it.
362 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
363 will now be deprecated.
365 PH/56 New os.c-cygwin from the Cygwin maintainer.
367 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
368 formats (previously only available in exim formats via -bpr, -bpru,
369 and -bpra. Now also available in native and exiqgrep formats)
371 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
372 with very large, slow to parse queues
374 JJ/08 exipick: added ! as generic prefix to negate any criteria format
376 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
378 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
379 responses to authentication challenges, though it was showing the
380 challenges; (ii) I've removed the CR characters from the debug output for
383 PH/58 Allow for the insertion of a newline as well as a space when a string
384 is turned into more than one encoded-word during RFC 2047 encoding. The
385 Sieve code now uses this.
387 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
388 data_4xx, lost_connection, tls_required.
390 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
391 message was being sent as an SMTP response.
393 PH/61 Add -l and -k options to exicyclog.
395 PH/62 When verifying, if an address was redirected to one new address, so that
396 verification continued, and the new address failed or deferred after
397 having set something in $address_data, the value of $address_data was not
398 passed back to the ACL. This was different to the case when no
399 redirection occurred. The value is now passed back in both cases.
401 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
402 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
403 use login_cap.h, so on its own it isn't the distinguishing feature. The
404 new name refers directly to the setclassresources() function.
406 PH/65 Added configuration files for NetBSD3.
408 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
410 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
411 is preferred over IPv4.
413 PH/68 The bounce_return_message and bounce_return_body options were not being
414 honoured for bounces generated during the reception of non-SMTP messages.
415 In particular, this applied to messages rejected by the ACL. This bug has
416 been fixed. However, if bounce_return_message is true and bounce_return_
417 body is false, the headers that are returned for a non-SMTP message
418 include only those that have been read before the error was detected.
419 (In the case of an ACL rejection, they have all been read.)
421 PH/69 The HTML version of the specification is now built in a directory called
422 spec_html instead of spec.html, because the latter looks like a path with
423 a MIME-type, and this confuses some software.
425 PH/70 Catch two compiler warnings in sieve.c.
427 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
428 function verify_get_ident() calls ip_connect() to connect a socket, but
429 if the "connect()" function timed out, ip_connect() used to close the
430 socket. However, verify_get_ident() also closes the socket later, and in
431 between Exim writes to the log, which may get opened at this point. When
432 the socket was closed in ip_connect(), the log could get the same file
433 descriptor number as the socket. This naturally causes chaos. The fix is
434 not to close the socket in ip_connect(); the socket should be closed by
435 the function that creates it. There was only one place in the code where
436 this was missing, in the iplookup router, which I don't think anybody now
437 uses, but I've fixed it anyway.
439 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
440 well as to direct DNS lookups. Otherwise the handling of names in host
441 lists is inconsistent and therefore confusing.
447 PH/01 Two changes to the default runtime configuration:
449 (1) Move the checks for relay_from_hosts and authenticated clients from
450 after to before the (commented out) DNS black list checks.
452 (2) Add control=submission to the relay_from_hosts and authenticated
453 clients checks, on the grounds that messages accepted by these
454 statements are most likely to be submissions.
456 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
458 (1) Generate an error if the third argument for the ${prvs expansion is
461 (2) Treat a missing third argument of ${prvscheck as if it were an empty
464 (3) Reset the variables that are obtained from the first argument of
465 ${prvscheck and used in the second argument before leaving the code,
466 because their memory is reclaimed, so using them afterwards may do
469 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
470 one (it's much easier than Tom thought :-).
472 (5) Because of (4), we can now allow for the use of $prvscheck_result
473 inside the third argument.
475 PH/03 For some reason, the default setting of PATH when running a command from
476 a pipe transport was just "/usr/bin". I have changed it to
479 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
480 anything to be listed in the output from -bV.
482 PH/05 When a filter generated an autoreply, the entire To: header line was
483 quoted in the delivery log line, like this:
485 => >A.N.Other <ano@some.domain> <original@ddress> ...
487 This has been changed so that it extracts the operative address. There
488 may be more than one such address. If so, they are comma-separated, like
491 => >ano@some.domain,ona@other.domain <original@ddress> ...
493 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
494 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
495 not being looked up in the rDNS to get a host name, Exim was showing the
496 IP address twice in Received: lines, even though the IP addresses were
497 identical. For example:
499 Received: from [1.2.3.4] (helo=[1.2.3.4])
501 However, if the real host name was known, it was omitting the HELO data
502 if it matched the actual IP address. This has been tidied up so that it
503 doesn't show the same IP address twice.
505 PH/07 When both +timestamp and +memory debugging was on, the value given by
506 $tod_xxx expansions could be wrong, because the tod_stamp() function was
507 called by the debug printing, thereby overwriting the timestamp buffer.
508 Debugging no longer uses the tod_stamp() function when +timestamp is set.
510 PH/08 When the original message was included in an autoreply transport, it
511 always said "this is a copy of the message, including all the headers",
512 even if body_only or headers_only was set. It now gives an appropriate
515 PH/09 Applied a patch from the Sieve maintainer which:
517 o fixes some comments
518 o adds the (disabled) notify extension core
519 o adds some debug output for the result of if/elsif tests
520 o points to the current vacation draft in the documentation
521 and documents the missing references header update
525 o fixes a bug in processing the envelope test (when testing
526 multiple envelope elements, the last element determinted the
529 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
530 Electronic Mail") by including:
532 Auto-submitted: auto-generated
534 in the messages that it generates (bounce messages and others, such as
535 warnings). In the case of bounce messages for non-SMTP mesages, there was
536 also a typo: it was using "Auto_submitted" (underscore instead of
537 hyphen). Since every message generated by Exim is necessarily in response
538 to another message, thes have all been changed to:
540 Auto-Submitted: auto-replied
542 in accordance with these statements in the RFC:
544 The auto-replied keyword:
546 - SHOULD be used on messages sent in direct response to another
547 message by an automatic process,
549 - MUST NOT be used on manually-generated messages,
551 - MAY be used on Delivery Status Notifications (DSNs) and Message
552 Disposition Notifications (MDNs),
554 - MUST NOT be used on messages generated by automatic or periodic
555 processes, except for messages which are automatic responses to
558 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
559 to the default Received: header definition.
561 PH/12 Added log selector acl_warn_skipped (default on).
563 PH/13 After a successful wildlsearch lookup, discard the values of numeric
564 variables because (a) they are in the wrong storage pool and (b) even if
565 they were copied, it wouldn't work properly because of the caching.
567 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
568 checking when decoding. Apparently there are clients that generate
569 overlong encoded strings. Why am I not surprised?
571 PH/15 If the first argument of "${if match_address" was not empty, but did not
572 contain an "@" character, Exim crashed. Now it writes a panic log message
573 and treats the condition as false.
575 PH/16 In autoreply, treat an empty string for "once" the same as unset.
577 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
578 extension "envelope-auth". The code is finished and in agreement with
579 other implementations, but there is no documentation so far and in fact,
580 nobody wrote the draft yet. This extension is currently #undef'ed, thus
581 not changing the active code.
583 Print executed "if" and "elsif" statements when debugging is used. This
584 helps a great deal to understand what a filter does.
586 Document more things not specified clearly in RFC3028. I had all this
587 sorted out, when out of a sudden new issues came to my mind. Oops."
589 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
592 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
593 canonical form (as documented). However, after a host name lookup from
594 the IP address, check_host() was doing a simple string comparison with
595 addresses acquired from the DNS when checking that the found name did
596 have the original IP as one of its addresses. Since any found IPv6
597 addresses are likely to be in abbreviated form, the comparison could
598 fail. Luckily, there already exists a function for doing the comparison
599 by converting both addresses to binary, so now that is used instead of
602 PH/20 There was another similar case to PH/19, when a complete host name was
603 given in a host list; looking up its IP address could give an abbreviated
604 form, whereas the current host's name might or might not be abbreviated.
605 The same fix has been applied.
611 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
612 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
615 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
616 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
618 PH/03 Typo: missing ".o" in src/pcre/Makefile.
620 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
621 header line, restrict the check to what is listed in RFCs 2369 and 2929.
622 Also, for "Auto-Submitted", treat anything other than "no" as
623 non-personal, in accordance with RFC 3834. (Previously it treated
624 anything starting "auto-" as non-personal.)
626 TF/01 The control=submission/name=... option had a problem with syntax
627 errors if the name included a slash character. The /name= option
628 now slurps the rest of the string, so it can include any characters
629 but it must come last in the list of options (after /sender_retain
632 PH/05 Some modifications to the interface to the fake nameserver for the new
640 TK/01 Added the "success_on_redirect" address verification option. See
641 NewStuff for rationale and an example.
643 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
645 PH/02 Patch to exigrep to allow it to work on syslog lines.
647 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
648 fread() to skip over the body file's header line, because in Cygwin the
649 header line is locked and is inaccessible.
651 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
652 co-exist for some time) to make it clear that it is the Exim ID that is
653 referenced, not the Message-ID: header line.
655 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
656 string_format() function, because snprintf() does not exist on all
659 PH/06 The use of forbid_filter_existstest now also locks out the use of the
660 ${stat: expansion item.
662 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
663 protocol synchronization error", to keep the pedants happy.
665 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
666 well as for IRIX systems, when gcc is being used. See the host.c source
669 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
671 PH/10 Named domain lists were not working if used in a queue_smtp_domains
674 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
675 transport and to the smtp transport in LMTP mode.
677 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
679 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
680 run a filter in a subprocess. This could lead to confusion in subsequent
681 lookups in the parent process. There should also be a search_tidyup() at
682 the end of the subprocess.
684 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
685 only if the host matched helo_try_verify_hosts, which caused the
686 verification to occur when the EHLO/HELO command was issued. The ACL just
687 tested the remembered result. Now, if a previous verification attempt has
688 not happened, "verify = helo" does it there and then.
690 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
692 TK/03 Fix log output including CR from clamd.
694 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
695 exist provoked a memory error which could cause a segfault.
697 PH/15 Installed PCRE 6.2
699 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
701 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
702 of the problem. Specifically, suggested +O2 rather than +O1 for the
705 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
707 PH/20 If a delivery was routed to a non-standard port by means of an SRV
708 record, the port was not correctly logged when the outgoing_port log
709 selector was set (it logged the transort's default port).
711 PH/21 Added support for host-specific ports to manualroute, queryprogram,
712 fallback_hosts, and "hosts" in the smtp transport.
714 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
715 host errors such as "Connection refused".
717 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
718 authentication with radiusclient 0.4.9:
720 - Error returned from rc_read_config was caught wrongly
721 - Username/password not passed on to radius server due to wrong length.
723 The presumption is that some radiusclient API changes for 4.51/PH/17
724 were not taken care of correctly. The code is still untested by me (my
725 Linux distribution still has 0.3.2 of radiusclient), but it was
726 contributed by a Radius user.
728 PH/24 When doing a callout, the value of $domain wasn't set correctly when
729 expanding the "port" option of the smtp transport.
731 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
732 while reading a MIME header. Thanks to Tom Hughes for a patch.
734 PH/24 Include config.h inside local_scan.h so that configuration settings are
737 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
738 that in an ACL for RCPT (for example), you can examine exactly what was
741 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
742 commands, but it was not correctly comparing the address with the actual
743 client host address. Thus, it would show the EHLO address in Received:
744 header lines when this was not necessary.
746 PH/27 Added the % operator to ${eval:}.
748 PH/28 Exim tries to create and chdir to its spool directory when it starts;
749 it should be ignoring failures (because with -C, for example, it has lost
750 privilege). It wasn't ignoring creation failures other than "already
753 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
756 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
757 because an input line was too long, either on its own, or by virtue of
758 too many continuations, the temporary file was not being removed, and the
759 return code was incorrect.
761 PH/31 Missing "BOOL" in function definition in filtertest.c.
763 PH/32 Applied Sieve patches from the maintainer.
765 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
767 PH/33 Added "verify = not_blind".
769 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
770 Local/Makefile (with some defaults set). These are used in built scripts
771 such as exicyclog, but they have never been used in the exim_install
772 script (though there are many overriding facilities there). I have
773 arranged that the exim_install script now takes note of these two
776 PH/35 Installed configuration files for Dragonfly.
778 PH/36 When a locally submitted message by a trusted user did not contain a
779 From: header, and the sender address was obtained from -f or from an SMTP
780 MAIL command, and the trusted user did not use -F to supply a sender
781 name, $originator_name was incorrectly used when constructing a From:
782 header. Furthermore, $originator_name was used for submission mode
783 messages from external hosts without From: headers in a similar way,
784 which is clearly wrong.
786 PH/37 Added control=suppress_local_fixups.
788 PH/38 When log_selector = +received_sender was set, and the addition of the
789 sender made the log line's construction buffer exactly full, or one byte
790 less than full, an overflow happened when the terminating "\n" was
793 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
794 when the result of a list match is failure because a DNS lookup failed.
796 PH/40 RM_COMMAND is now used in the building process.
798 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
799 the "build-* directories that it finds.
801 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
802 address in a domain literal was a prefix of an interface address.
804 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
805 when verifying a sender address, unless rewrite_headers is false.
807 PH/44 Wrote a long comment about why errors_to addresses are verified as
808 recipients, not senders.
810 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
811 the ratelimit ACL was added.
813 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
815 PH/46 Added extra information about PostgreSQL errors to the error string.
817 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
818 avoiding the need to install special zones in a real server. This is
819 backwards compatible; if it can't find the fake resolver, it drops back.
820 Thus, both old and new test suites can be run.
822 TF/02 Added util/ratelimit.pl
824 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
825 clock is set back in time.
827 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
828 Candler <B.Candler@pobox.com>.
830 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
831 OK for addresses that are the result of redirections.
833 PH/48 A number of further additions for the benefit of the new test suite,
834 including a fake gethostbyname() that interfaces to the fake DNS resolver
837 TF/06 The fix for widen_domains has also been applied to qualify_single and
838 search_parents which are the other dnslookup options that can cause
841 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
844 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
846 TF/07 Exim produced the error message "an SRV record indicated no SMTP
847 service" if it encountered an MX record with an empty target hostname.
848 The message is now "an MX or SRV record indicated no SMTP service".
850 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
851 if the DNS of the sending site is misconfigured. This is quite a
852 common situation. This change restores the behaviour of treating a
853 helo verification defer as a failure.
855 PH/51 If self=fail was set on a router, the bounce message did not include the
856 actual error message.
862 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
864 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
865 command itself ended in error, the underlying message about the transport
866 filter timeout was being overwritten with the pipe command error. Now the
867 underlying error message should be appended to the second error message.
869 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
871 PH/02 Reduce the amount of output that "make" produces by default. Full output
872 can still be requested.
874 PH/03 The warning log line about a condition test deferring for a "warn" verb
875 was being output only once per connection, rather than after each
876 occurrence (because it was using the same function as for successful
877 "warn" verbs). This seems wrong, so I have changed it.
879 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
880 it should not have, which might have caused a crash in the right
881 circumstances, but probably never did.
883 PH/04 Installed a modified version of Tony Finch's patch to make submission
884 mode fix the return path as well as the Sender: header line, and to
885 add a /name= option so that you can make the user's friendly name appear
888 TF/03 Added the control = fakedefer ACL modifier.
890 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
891 Mark Lowes for thorough testing.
893 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
895 TK/03 Merged latest SRS patch from Miles Wilton.
897 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
898 with the definition in sysexits.h (which is #included earlier).
899 Fortunately, Exim does not actually use EX_OK. The code used to try to
900 preserve the sysexits.h value, by assumimg that macro definitions were
901 scanned for macro replacements. I have been disabused of this notion,
902 so now the code just undefines EX_OK before #including unistd.h.
904 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
905 in the smtp transport. When a block could not be written in a single
906 write() function, the timeout was being re-applied to each part-write.
907 This seems wrong - if the receiver was accepting one byte at a time it
908 would take for ever. The timeout is now adjusted when this happens. It
909 doesn't have to be particularly precise.
911 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
912 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
914 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
915 without a domain if the check to <postmaster@domain> fails.
917 SC/01 Eximstats: added -xls and the ability to specify output files
918 (patch written by Frank Heydlauf).
920 SC/02 Eximstats: use FileHandles for outputing results.
922 SC/03 Eximstats: allow any combination of xls, txt, and html output.
924 SC/04 Eximstats: fixed display of large numbers with -nvr option
926 SC/05 Eximstats: fixed merging of reports with empty tables.
928 SC/06 Eximstats: added the -include_original_destination flag
930 SC/07 Eximstats: removed tabs and trailing whitespace.
932 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
934 TK/06 MBOX spool code: Add real "From " MBOX separator line
935 so the .eml file is really in mbox format (even though
936 most programs do not really care). Patch from Alex Miller.
938 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
939 The latter is generated from $received_to and is only set if the
940 message has one envelope recipient. SA can use these headers,
941 obviously out-of-the-box. Patch from Alex Miller.
943 PH/08 The ${def test on a variable was returning false if the variable's
944 value was "0", contrary to what the specification has always said!
945 The result should be true unless the variable is empty.
947 PH/09 The syntax error of a character other than { following "${if
948 def:variable_name" (after optional whitespace) was not being diagnosed.
949 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
950 accidental colon was present, for example, could give incorrect results.
952 PH/10 Tidied the code in a number of places where the st_size field of a stat()
953 result is used (not including appendfile, where other changes are about
956 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
957 This involved changing a lot of size variables from int to off_t. It
958 should work with maildirs and everything.
960 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
961 spamd dying while we are connected to it.
963 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
966 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
967 function for BDB 4.3.
969 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
971 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
972 the sources for PCRE have been re-arranged and more files are now
975 PH/15 The code I had for printing potentially long long variables in PH/11
976 above was not the best (it lost precision). The length of off_t variables
977 is now inspected at build time, and an appropriate printing format (%ld
978 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
979 to be "long long int" or "long int". This is needed for the internal
980 formatting function string_vformat().
982 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
983 the configuration file to be ":syslog", then the script "guesses" where
984 the logs files are, rather than using the compiled in default. In our
985 case the guess is not the same as the compiled default, so the script
986 suddenly stopped working when I started to use syslog. The patch checks
987 to see if log_file_path is "". If so, it attempts to read it from exim
988 with no configuration file to get the compiled in version, before it
989 falls back to the previous guessing code."
991 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
992 implementing BATV in an Exim configuration. See NewStuff for the gory
995 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
996 Makefile that are specific to HP-UX.
998 PH/18 If the "use_postmaster" option was set for a recipient callout together
999 with the "random" option, the postmaster address was used as the MAIL
1000 FROM address for the random test, but not for the subsequent recipient
1001 test. It is now used for both.
1003 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1004 patch removes a few documentation additions to RFC 3028, because the
1005 latest draft now contains them. It adds the new en;ascii-case comparator
1006 and a new error check for 8bit text in MIME parts. Comparator and
1007 require names are now matched exactly. I enabled the subaddress
1008 extension, but it is not well tested yet (read: it works for me)."
1010 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1011 rework some of the code of TK/09 above to avoid the hardwired use of
1012 "%lld" and "long long". Replaced the call to snprintf() with a call to
1015 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1016 records point to non-existent hosts", "retry timeout exceeded", and
1017 "retry time not reached for any host after a long failure period".
1019 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1020 experimental DomainKeys support:
1022 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1023 (2) On an error such as an illegally used "control", the wrong name for
1024 the control was given.
1026 These problems did NOT occur unless DomainKeys support was compiled.
1028 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1030 PH/24 Added ${if match_ip condition.
1032 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1033 databases so that it will be absolutely obvious if a crash occurs in the
1034 DB library. This is a regular occurrence (often caused by mis-matched
1037 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1038 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1039 picked up on a user's system that detects such things. There doesn't seem
1040 to be a gcc warning option for this - only an attribute that has to be
1041 put on the function's prototype. It seems that in Fedora Core 4 they have
1042 set this on a number of new functions. No doubt there will be more in due
1045 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1046 specify a transport. However, if an address that was verified by such a
1047 router was the subject of a callout, Exim crashed because it tried to
1048 read the rcpt_include_affixes from the non-existent transport. Now it
1049 just assumes that the setting of that option is false. This bug was
1050 introduced by 4.51/PH/31.
1052 PH/28 Changed -d+all to exclude +memory, because that information is very
1053 rarely of interest, but it makes the output a lot bigger. People tend to
1054 do -d+all out of habit.
1056 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1057 code in os-type was giving problems when libc.so lives in lib64, like on
1060 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1061 aren't the modern standard, and it seems that some systems' include files
1062 don't always have them. Exim was already checking for some of the newer
1063 ones like T_AAAA, and defining it itself. I've added checks for all the
1064 record types that Exim uses.
1066 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1067 not automatically generating a new one, as it is supposed to. This
1068 prevented TLS from working. If the file did exist, but contained invalid
1069 data, a new version was generated, as expected. It was only the case of a
1070 non-existent file that was broken.
1072 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1073 with a change in libdomainkeys > 0.64.
1075 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1076 from DNS. If the selector record carries the flag, it now has
1077 precedence over the domain-wide flag.
1079 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1081 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1082 the use of a port name that isn't defined in /etc/services) occurred, the
1083 message was deferred as in a normal delivery, and thus remained on the
1084 spool, instead of being failed because of the mua_wrapper setting. This
1085 is now fixed, and I tidied up some of the mua_wrapper messages at the
1088 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1089 the messages in a hash of arrays rather than using individual hashes.
1090 This is a bit cleaner and results in dramatic memory savings, albeit
1091 at a slight CPU cost.
1093 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1094 as requested by Marc Sherman.
1096 SC/10 Eximstats: added histograms for user specified patterns as requested
1099 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1101 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1102 fopen() in the content-scanning modules that did not already have it.
1108 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1109 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1111 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1113 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1115 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1116 to test Sieve filters that use "vacation".
1118 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1119 that changes the way the GnuTLS parameters are stored in the cache file.
1120 The new format can be generated externally. For backward compatibility,
1121 if the data in the cache doesn't make sense, Exim assumes it has read an
1122 old-format file, and it generates new data and writes a new file. This
1123 means that you can't go back to an older release without removing the
1126 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1127 work if there are any delivery delays because "one_time" forces the
1128 parent to be marked "delivered", so its unseen clone is never tried
1129 again. For this reason, Exim now forbids the simultaneous setting of
1132 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1133 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1134 entry if you want to know the details. The fix, however, neglected to
1135 consider the case when local delivery batching is involved. The test for
1136 "previously delivered" was not happening when checking to see if an
1137 address could be batched with a previous (undelivered) one; under
1138 certain circumstances this could lead to multiple deliveries to the same
1141 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1142 in its include files, and this causes problems building Exim.
1144 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1145 header_syntax) but Exim was just ignoring anything given after a slash.
1146 In particular, this caused confusion with an attempt to use "verify =
1147 reverse_host_lookup/defer_ok". An error is now given when options are
1148 supplied for verify items that do not have them. (Maybe reverse_host_
1149 lookup should have a defer_ok option, but that's a different point.)
1151 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1152 defined by RFC 821) to 2048, because there were problems with some AUTH
1153 commands, and RFC 1869 says the size should be increased for extended
1154 SMTP commands that take arguments.
1156 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1159 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1160 "unknown" error; now it says that the functionality isn't in the binary.
1162 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1163 an address' error message when a string expansion fails (syntax or
1164 whatever). Otherwise the password may appear in the log. Following change
1165 PH/42 below, there is no longer a chance of it appearing in a bounce
1168 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1170 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1171 of its addresses was ever tried. (Bugzilla bug #2.)
1173 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1174 the result incorrectly in the debug output. (It correctly added a newline
1175 to what was transported.)
1177 TF/01 Added $received_time.
1179 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1180 commented out examples of how to interface to a virus scanner and to
1181 SpamAssassin. Also added commented examples of av_scanner and
1182 spamd_address settings.
1184 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1185 and controls are allowed in which ACLs. There were a couple of minor
1186 errors. Some of the entries in the conditions table (which is a table of
1187 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1188 negation of where the condition IS allowed.
1190 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1192 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1193 header file does not have a version number, so I've had to invent a new
1194 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1195 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1196 radiusclient), but it was contributed by a Radius user.
1198 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1199 files or directories, for OpenSSL.
1201 PH/19 When an Exim process that is running as root has to create an Exim log
1202 file, it does so in a subprocess that runs as exim:exim so as to get the
1203 ownership right at creation (otherwise, other Exim processes might see
1204 the file with the wrong ownership). There was no test for failure of this
1205 fork() call, which would lead to the process getting stuck as it waited
1206 for a non-existent subprocess. Forks do occasionally fail when resources
1207 run out. I reviewed all the other calls to fork(); they all seem to check
1210 PH/20 When checking for unexpected SMTP input at connect time (before writing
1211 the banner), Exim was not dealing correctly with a non-positive return
1212 from the read() function. If the client had disconnected by this time,
1213 the result was a log entry for a synchronization error with an empty
1214 string after "input=" when read() returned zero. If read() returned -1
1215 (an event I could not check), uninitialized data bytes were printed.
1216 There were reports of junk text (parts of files, etc) appearing after
1219 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1221 PH/22 Added support for macro redefinition, and (re)definition in between
1222 driver and ACL definitions.
1224 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1225 forgetting to use the resulting value; it was using the unexpanded value.
1227 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1228 hadn't been configured. The fix is from Juergen Kreileder, who
1229 understands it better than I do:
1231 "Here's what I see happening with three configured cyrus_sasl
1232 authenticators configured (plain, login, cram-md5):
1234 On startup auth_cyrus_sasl_init() gets called for each of these.
1235 This means three calls to sasl_listmech() without a specified mech_list.
1236 => SASL tests which mechs of all available mechs actually work
1237 => three warnings about OTP not working
1238 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1240 With the patch, sasl_listmech() also gets called three times. But now
1241 SASL's mech_list option is set to the server_mech specified in the the
1242 authenticator. Or in other words, the answer from sasl_listmech()
1243 gets limited to just the mech you're testing for (which is different
1245 => the return list contains just 'plain' or 'login', 'cram-md5' or
1246 nothing depending on the value of ob->server_mech.
1248 I've just tested the patch: Authentication still works fine,
1249 unavailable mechs specified in the exim configuration are still
1250 caught, and the auth.log warnings about OTP are gone."
1252 PH/25 When debugging is enabled, the contents of the command line are added
1253 to the debugging output, even when log_selector=+arguments is not
1256 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1257 answer is "GNU", and only if the return is "GNU/something" is the answer
1260 PH/27 $acl_verify_message is now set immediately after the failure of a
1261 verification in an ACL, and so is available in subsequent modifiers. In
1262 particular, the message can be preserved by coding like this:
1264 warn !verify = sender
1265 set acl_m0 = $acl_verify_message
1267 Previously, $acl_verify_message was set only while expanding "message"
1268 and "log_message" when a very denied access.
1270 PH/28 Modified OS/os.c-Linux with
1272 -#ifndef OS_LOAD_AVERAGE
1273 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1275 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1276 nomenclature these days.)
1278 PH/29 Installed patch from the Sieve maintainer that adds the options
1279 sieve_useraddress and sieve_subaddress to the redirect router.
1281 PH/30 In these circumstances:
1282 . Two addresses routed to the same list of hosts;
1283 . First host does not offer TLS;
1284 . First host accepts first address;
1285 . First host gives temporary error to second address;
1286 . Second host offers TLS and a TLS session is established;
1287 . Second host accepts second address.
1288 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1289 and peerdn, if requested) that were in fact used only for the second
1292 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1293 attention to any local part prefix or suffix that was matched by the
1294 router that accepted the address. It now behaves in the same way as it
1295 does for delivery: the affixes are removed from the local part unless
1296 rcpt_include_affixes is set on the transport.
1298 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1299 timeout during the DATA phase of an incoming message.
1301 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1302 applied a patch sanctioned by the Sieve maintainer.
1304 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1305 the uid or gid is negative. A case of a negative gid caused this to be
1306 noticed. The fix allows for either to be negative.
1308 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1309 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1312 PH/36 The change PH/12 above was broken. Fixed it.
1314 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1315 the grounds that routing the same address twice would always produce the
1316 same answer. This might have been true once, but it is certainly no
1317 longer true now. Routing a child address may depend on the previous
1318 routing that produced that child. Some complicated redirection strategies
1319 went wrong when messages had multiple recipients, and made Exim's
1320 behaviour dependent on the order in which the addresses were given.
1322 I have moved the duplicate checking until after the routing is complete.
1323 Exim scans the addresses that are assigned to local and remote
1324 transports, and removes any duplicates. This means that more work will be
1325 done, as duplicates will always all be routed, but duplicates are
1326 presumably rare, so I don't expect this is of any significance.
1328 For deliveries to pipes, files, and autoreplies, the duplicate checking
1329 still happens during the routing process, since they are not going to be
1332 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1333 It corrects a timeout issue with spamd. This is Ian's comment: "The
1334 background is that sometimes spamd either never reads data from a
1335 connection it has accepted, or it never writes response data. The exiscan
1336 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1337 blindly assumes that writes won't block so it may never time out."
1339 PH/39 Allow G after quota size as well as K and M.
1341 PH/40 The value set for $authenticated_id in an authenticator may not contain
1342 binary zeroes or newlines because the value is written to log lines and
1343 to spool files. There was no check on this. Now the value is run through
1344 the string_printing() function so that such characters are converted to
1345 printable escape sequences.
1347 PH/41 $message_linecount is a new variable that contains the total number of
1348 lines in the message. Compare $body_linecount, which is the count for the
1351 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1352 bounce and delay warning messages, except in certain special cases, which
1355 (a) An SMTP error message from a remote host;
1356 (b) A message specified in a :fail: redirection;
1357 (c) A message specified in a "fail" command in a system filter;
1358 (d) A message specified in a FAIL return from the queryprogram router;
1359 (e) A message specified by the cannot_route_message router option.
1361 In these cases only, Exim does include the error details in bounce and
1362 warning messages. There are also a few cases where bland messages such
1363 as "unrouteable address" or "local delivery error" are given.
1365 PH/43 $value is now also set for the "else" part of a ${run expansion.
1367 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1368 being worked on, but at least Exim now implements the latest version to
1371 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1372 process to complete was treated as a delivery failure, a timeout while
1373 writing the message to the pipe was logged, but erroneously treated as a
1374 successful delivery. Such timeouts include transport filter timeouts. For
1375 consistency with the overall process timeout, these timeouts are now
1376 treated as errors, giving rise to delivery failures by default. However,
1377 there is now a new Boolean option for the pipe transport called
1378 timeout_defer, which, if set TRUE, converts the failures into defers for
1379 both kinds of timeout. A transport filter timeout is now identified in
1382 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1383 systems where "make" and "gmake" are different, calling "gmake" at top
1384 level broke things. I've arranged for the value of $(MAKE) to be passed
1385 from the Makefile to this script so that it can call the same version of
1389 A note about Exim versions 4.44 and 4.50
1390 ----------------------------------------
1392 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1393 changes of various kinds. As a consequence, a big documentation update was
1394 needed. This delayed the release for rather longer than seemed good, especially
1395 in the light of a couple of (minor) security issues. Therefore, the changes
1396 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1397 release. So 4.44 and 4.50 are in effect two different branches that both start
1400 I have left the 4.50 change log unchanged; it contains all the changes since
1401 4.43. The change log for 4.44 is below; many of its items are identical to
1402 those for 4.50. This seems to be the most sensible way to preserve the
1403 historical information.
1409 1. Minor wording change to the doc/README.SIEVE file.
1411 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1412 computation of the current number of files was incorrect.
1414 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1415 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1418 4. Give more explanation in the error message when the command for a transport
1419 filter fails to execute.
1421 5. There are several places where Exim runs a non-Exim command in a
1422 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1423 was being done only for the command run by the queryprogram router. It is
1424 now done for all such subprocesses. The other cases are: ${run, transport
1425 filters, and the commands run by the lmtp and pipe transports.
1427 6. Added CONFIGURE_GROUP build-time option.
1429 7. Some older OS have a limit of 256 on the maximum number of file
1430 descriptors. Exim was using setrlimit() to set 1000 as a large value
1431 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1432 systems. I've change it so that if it can't get 1000, it tries for 256.
1434 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1435 was an oversight, and furthermore, ever since the addition of extra
1436 controls (e.g. 4.43/32), the checks on when to allow different forms of
1437 "control" were broken. There should now be diagnostics for all cases when a
1438 control that does not make sense is encountered.
1440 9. Added the /retain_sender option to "control=submission".
1442 10. $recipients is now available in the predata ACL (oversight).
1444 11. Tidy the search cache before the fork to do a delivery from a message
1445 received from the command line. Otherwise the child will trigger a lookup
1446 failure and thereby defer the delivery if it tries to use (for example) a
1447 cached ldap connection that the parent has called unbind on.
1449 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1450 of $address_data from the recipient verification was clobbered by the
1451 sender verification.
1453 13. The value of address_data from a sender verification is now available in
1454 $sender_address_data in subsequent conditions in the ACL statement.
1456 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1458 15. Added a new option "connect=<time>" to callout options, to set a different
1461 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1462 was its contents. (It was OK if the option was not defined at all.)
1464 17. A "Completed" log line is now written for messages that are removed from
1465 the spool by the -Mrm option.
1467 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1468 information about exactly what failed.
1470 19. Added -dd to debug only the daemon process.
1472 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1473 handles timeouts, both on the server side and network timeouts. Renamed the
1474 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1476 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1477 It is now set to "smtps".
1479 22. $host_address is now set to the target address during the checking of
1480 ignore_target_hosts.
1482 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1483 being passed; this would have caused $sender_host_name to have been used if
1484 matching the list had actually called for a host name (not very likely,
1485 since this list is usually IP addresses). A host name is now passed as
1488 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1489 code that specifies a non-restarting handler (typically sigaction() in
1490 modern systems) in an attempt to fix a rare and obscure crash bug.
1492 25. Narrowed the window for a race in the daemon that could cause it to ignore
1493 SIGCHLD signals. This is not a major problem, because they are used only to
1494 wake it up if nothing else does.
1496 26. A malformed maildirsize file could cause Exim to calculate negative values
1497 for the mailbox size or file count. Odd effects could occur as a result.
1498 The maildirsize information is now recalculated if the size or filecount
1501 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1502 support for a long time. Removed HAVE_SYS_VFS_H.
1504 28. Installed the latest version of exipick from John Jetmore.
1506 29. In an address list, if the pattern was not a regular expression, an empty
1507 subject address (from a bounce message) matched only if the pattern was an
1508 empty string. Non-empty patterns were not even tested. This was the wrong
1509 because it is perfectly reasonable to use an empty address as part of a
1510 database query. An empty address is now tested by patterns that are
1511 lookups. However, all the other forms of pattern expect the subject to
1512 contain a local part and a domain, and therefore, for them, an empty
1513 address still always fails if the pattern is not itself empty.
1515 30. Exim went into a mad DNS loop when attempting to do a callout where the
1516 host was specified on an smtp transport, and looking it up yielded more
1517 than one IP address.
1519 31. Re-factored the code for checking spool and log partition space into a
1520 function that finds that data and another that does the check. The former
1521 is then used to implement four new variables: $spool_space, $log_space,
1522 $spool_inodes, and $log_inodes.
1524 32. The RFC2047 encoding function was originally intended for short strings
1525 such as real names; it was not keeping to the 75-character limit for
1526 encoded words that the RFC imposes. It now respects the limit, and
1527 generates multiple encoded words if necessary. To be on the safe side, I
1528 have increased the buffer size for the ${rfc2047: expansion operator from
1531 33. It is now permitted to omit both strings after an "if" condition; if the
1532 condition is true, the result is "true". As before, when the second string
1533 is omitted, a false condition yields an empty string. This makes it less
1534 cumbersome to write custom ACL and router conditions.
1536 34. Failure to deliver a bounce message always caused it to be frozen, even if
1537 there was an errors_to setting on the router. The errors_to setting is now
1540 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1541 canonical form (fully expanded) before being placed in
1542 $sender_host_address.
1544 36. The table in the code that translates DNS record types into text (T_A to
1545 "A" for instance) was missing entries for NS and CNAME. It is just possible
1546 that this could have caused confusion if both these types were looked up
1547 for the same domain, because the text type is used as part of Exim's
1548 per-process caching. But the chance of anyone hitting this buglet seems
1551 37. The dnsdb lookup has been extended in a number of ways.
1553 (1) There is a new type, "zns", which walks up the domain tree until it
1554 finds some nameserver records. It should be used with care.
1556 (2) There is a new type, "mxh", which is like "mx" except that it returns
1557 just the host names, not the priorities.
1559 (3) It is now possible to give a list of domains (or IP addresses) to be
1560 looked up. The behaviour when one of the lookups defers can be
1561 controlled by a keyword.
1563 (4) It is now possible to specify the separator character for use when
1564 multiple records are returned.
1566 38. The dnslists ACL condition has been extended: it is now possible to supply
1567 a list of IP addresses and/or domains to be looked up in a particular DNS
1570 39. Added log_selector=+queue_time_overall.
1572 40. When running the queue in the test harness, wait just a tad after forking a
1573 delivery process, to get repeatability of debugging output.
1575 41. Include certificate and key file names in error message when GnuTLS fails
1576 to set them up, because the GnuTLS error message doesn't include the name
1577 of the failing file when there is a problem reading it.
1579 42. Allow both -bf and -bF in the same test run.
1581 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1583 44. The "Exiscan patch" is now merged into the mainline Exim source.
1585 45. Sometimes the final signoff response after QUIT could fail to get
1586 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1587 < 0 before doing a fflush(). This bug looks as though it goes back to the
1588 introduction of TLS in release 3.20, but "sometimes" must have been rare
1589 because the tests only now provoked it.
1591 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1592 (this can affect the format of dates).
1594 47. exim_tidydb, when checking for the continued existence of a message for
1595 which it has found a message-specific retry record, was not finding
1596 messages that were in split spool directories. Consequently, it was
1597 deleting retry records that should have stayed in existence.
1599 48. Steve fixed some bugs in eximstats.
1601 49. The SPA authentication driver was not abandoning authentication and moving
1602 on to the next authenticator when an expansion was forced to fail,
1603 contradicting the general specification for all authenticators. Instead it
1604 was generating a temporary error. It now behaves as specified.
1606 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1607 (the order specifies the preference for clients). The order is now AES256,
1608 AES128, 3DES, ARCFOUR128.
1610 51. Small patch to Sieve code - explicitly set From: when generating an
1613 52. Exim crashed if a remote delivery caused a very long error message to be
1614 recorded - for instance if somebody sent an entire SpamAssassin report back
1615 as a large number of 550 error lines. This bug was coincidentally fixed by
1616 increasing the size of one of Exim's internal buffers (big_buffer) that
1617 happened as part of the Exiscan merge. However, to be on the safe side, I
1618 have made the code more robust (and fixed the comments that describe what
1621 53. Now that there can be additional text after "Completed" in log lines (if
1622 the queue_time_overall log selector is set), a one-byte patch to exigrep
1623 was needed to allow it to recognize "Completed" as not the last thing in
1626 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1627 patch that reportedly fixes this has been added. I am not expert enough to
1628 create a test for it. This is what the patch creator wrote:
1630 "I found a little strange behaviour of ldap code when working with
1631 Windows 2003 AD Domain, where users was placed in more than one
1632 Organization Units. When I tried to give exim partial DN, the exit code
1633 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1634 But simultaneously result of request was absolutely normal ldap result,
1635 so I produce this patch..."
1637 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1638 REFERENCE, so I have modified the code to exclude the patch when that macro
1641 55. Some experimental protocols are using DNS PTR records for new purposes. The
1642 keys for these records are domain names, not reversed IP addresses. The
1643 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1644 leaves it alone. Component reversal etc. now happens only for IP addresses.
1647 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1649 57. Double the size of the debug message buffer (to 2048) so that more of very
1650 long debug lines gets shown.
1652 58. The exicyclog utility now does better if the number of log files to keep
1653 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1655 59. Two changes related to the smtp_active_hostname option:
1657 (1) $smtp_active_hostname is now available as a variable.
1658 (2) The default for smtp_banner uses $smtp_active_hostname instead
1659 of $primary_hostname.
1661 60. The host_aton() function is supposed to be passed a string that is known
1662 to be a valid IP address. However, in the case of IPv6 addresses, it was
1663 not checking this. This is a hostage to fortune. Exim now panics and dies
1664 if the condition is not met. A case was found where this could be provoked
1665 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1666 components; fortuitously, this particular loophole had already been fixed
1667 by change 4.50/55 above.
1669 If there are any other similar loopholes, the new check in host_aton()
1670 itself should stop them being exploited. The report I received stated that
1671 data on the command line could provoke the exploit when Exim was running as
1672 exim, but did not say which command line option was involved. All I could
1673 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1674 running as the user.
1677 61. There was a buffer overflow vulnerability in the SPA authentication code
1678 (which came originally from the Samba project). I have added a test to the
1679 spa_base64_to_bits() function which I hope fixes it.
1682 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1683 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1685 63. The daemon start-up calls getloadavg() while still root for those OS that
1686 need the first call to be done as root, but it missed one case: when
1687 deliver_queue_load_max is set with deliver_drop_privilege. This is
1688 necessary for the benefit of the queue runner, because there is no re-exec
1689 when deliver_drop_privilege is set.
1691 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1692 This has been fixed.
1694 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1695 was in use, was not putting the data itself into the right store pool;
1696 consequently, it could be overwritten for a subsequent message in the same
1697 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1700 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1702 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1703 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1704 the function were treating the return as a boolean value, which happened to
1705 work because 0=false and not-0=true, but is not correct code.
1707 68. The host_aton() function was not handling scoped IPv6 addresses (those
1708 with, for example, "%eth0" on the end) correctly.
1710 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1711 negated items (that is, ~something) in unsigned ints. Some compilers
1712 apparently mutter when there is no cast.
1714 70. If an address verification called from an ACL failed, and did not produce a
1715 user-specific message (i.e. there was only a "system" message), nothing was
1716 put in $acl_verify_message. In this situation, it now puts the system
1719 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1720 session; change 4.31/43 added the unwanted input to the log line - except
1721 that it did not do this in the start of session case. It now does.
1723 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1724 This is wrong and can cause the other end to generate a synchronization
1725 error if it is another Exim or anything else that does the synchronization
1726 check. A QUIT command is no longer sent after a timeout.
1728 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1729 during host lookups.
1731 74. The defer_ok option of callout verification was not working if it was used
1732 when verifying addresses in header lines, that is, for this case:
1734 verify = header_sender/callout=defer_ok
1736 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1737 those file descriptors could be used for SMTP connections. If anything
1738 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1739 could be sent to the SMTP client, causing chaos. The daemon now opens
1740 stdin, stdout, and stderr to /dev/null when it puts itself into the
1743 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1744 log by default. The user can override this with suitable Perl magic.
1746 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1747 the log message when discard triggers, was not working for the DATA ACL or
1748 for the non-SMTP ACL.
1750 78. Error message wording change in sieve.c.
1752 79. If smtp_accept_max_per_host was set, the number of connections could be
1753 restricted to fewer than expected, because the daemon was trying to set up
1754 a new connection before checking whether the processes handling previous
1755 connections had finished. The check for completed processes is now done
1756 earlier. On busy systems, this bug wouldn't be noticed because something
1757 else would have woken the daemon, and it would have reaped the completed
1760 80. If a message was submitted locally by a user whose login name contained one
1761 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1762 It caused a spool format error. I have fixed the spool reading code. A
1763 related problem was that the "from" clause in the Received: line became
1764 illegal because of the space(s). It is now covered by ${quote_local_part.
1766 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1769 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1771 83. Patch from Sieve maintainer to fix unterminated string problem in
1772 "vacation" handling.
1774 84. Some minor changes to the Linux configuration files to help with other
1775 OS variants using glibc.
1777 85. One more patch for Sieve to update vacation handling to latest spec.
1780 ----------------------------------------------------
1781 See the note above about the 4.44 and 4.50 releases.
1782 ----------------------------------------------------
1788 1. Change 4.43/35 introduced a bug that caused file counts to be
1789 incorrectly computed when quota_filecount was set in an appendfile
1792 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1793 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1796 3. Give more explanation in the error message when the command for a transport
1797 filter fails to execute.
1799 4. There are several places where Exim runs a non-Exim command in a
1800 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1801 was being done only for the command run by the queryprogram router. It is
1802 now done for all such subprocesses. The other cases are: ${run, transport
1803 filters, and the commands run by the lmtp and pipe transports.
1805 5. Some older OS have a limit of 256 on the maximum number of file
1806 descriptors. Exim was using setrlimit() to set 1000 as a large value
1807 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1808 systems. I've change it so that if it can't get 1000, it tries for 256.
1810 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1811 was an oversight, and furthermore, ever since the addition of extra
1812 controls (e.g. 4.43/32), the checks on when to allow different forms of
1813 "control" were broken. There should now be diagnostics for all cases when a
1814 control that does not make sense is encountered.
1816 7. $recipients is now available in the predata ACL (oversight).
1818 8. Tidy the search cache before the fork to do a delivery from a message
1819 received from the command line. Otherwise the child will trigger a lookup
1820 failure and thereby defer the delivery if it tries to use (for example) a
1821 cached ldap connection that the parent has called unbind on.
1823 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1824 of $address_data from the recipient verification was clobbered by the
1825 sender verification.
1827 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1828 was its contents. (It was OK if the option was not defined at all.)
1830 11. A "Completed" log line is now written for messages that are removed from
1831 the spool by the -Mrm option.
1833 12. $host_address is now set to the target address during the checking of
1834 ignore_target_hosts.
1836 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1837 being passed; this would have caused $sender_host_name to have been used if
1838 matching the list had actually called for a host name (not very likely,
1839 since this list is usually IP addresses). A host name is now passed as
1842 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1843 code that specifies a non-restarting handler (typically sigaction() in
1844 modern systems) in an attempt to fix a rare and obscure crash bug.
1846 15. Narrowed the window for a race in the daemon that could cause it to ignore
1847 SIGCHLD signals. This is not a major problem, because they are used only to
1848 wake it up if nothing else does.
1850 16. A malformed maildirsize file could cause Exim to calculate negative values
1851 for the mailbox size or file count. Odd effects could occur as a result.
1852 The maildirsize information is now recalculated if the size or filecount
1855 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1856 support for a long time. Removed HAVE_SYS_VFS_H.
1858 18. Updated exipick to current release from John Jetmore.
1860 19. Allow an empty sender to be matched against a lookup in an address list.
1861 Previously the only cases considered were a regular expression, or an
1864 20. Exim went into a mad DNS lookup loop when doing a callout where the
1865 host was specified on the transport, if the DNS lookup yielded more than
1868 21. The RFC2047 encoding function was originally intended for short strings
1869 such as real names; it was not keeping to the 75-character limit for
1870 encoded words that the RFC imposes. It now respects the limit, and
1871 generates multiple encoded words if necessary. To be on the safe side, I
1872 have increased the buffer size for the ${rfc2047: expansion operator from
1875 22. Failure to deliver a bounce message always caused it to be frozen, even if
1876 there was an errors_to setting on the router. The errors_to setting is now
1879 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1880 canonical form (fully expanded) before being placed in
1881 $sender_host_address.
1883 24. Updated eximstats to version 1.33
1885 25. Include certificate and key file names in error message when GnuTLS fails
1886 to set them up, because the GnuTLS error message doesn't include the name
1887 of the failing file when there is a problem reading it.
1889 26. Expand error message when OpenSSL has problems setting up cert/key files.
1892 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1893 (this can affect the format of dates).
1895 28. exim_tidydb, when checking for the continued existence of a message for
1896 which it has found a message-specific retry record, was not finding
1897 messages that were in split spool directories. Consequently, it was
1898 deleting retry records that should have stayed in existence.
1900 29. eximstats updated to version 1.35
1901 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1902 1.35 - bugfix such that pie charts by volume are generated correctly
1904 30. The SPA authentication driver was not abandoning authentication and moving
1905 on to the next authenticator when an expansion was forced to fail,
1906 contradicting the general specification for all authenticators. Instead it
1907 was generating a temporary error. It now behaves as specified.
1909 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1910 (the order specifies the preference for clients). The order is now AES256,
1911 AES128, 3DES, ARCFOUR128.
1913 31. Small patch to Sieve code - explicitly set From: when generating an
1916 32. Exim crashed if a remote delivery caused a very long error message to be
1917 recorded - for instance if somebody sent an entire SpamAssassin report back
1918 as a large number of 550 error lines. This bug was coincidentally fixed by
1919 increasing the size of one of Exim's internal buffers (big_buffer) that
1920 happened as part of the Exiscan merge. However, to be on the safe side, I
1921 have made the code more robust (and fixed the comments that describe what
1924 33. Some experimental protocols are using DNS PTR records for new purposes. The
1925 keys for these records are domain names, not reversed IP addresses. The
1926 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1927 leaves it alone. Component reversal etc. now happens only for IP addresses.
1930 34. The host_aton() function is supposed to be passed a string that is known
1931 to be a valid IP address. However, in the case of IPv6 addresses, it was
1932 not checking this. This is a hostage to fortune. Exim now panics and dies
1933 if the condition is not met. A case was found where this could be provoked
1934 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1935 components; fortuitously, this particular loophole had already been fixed
1936 by change 4.50/55 or 4.44/33 above.
1938 If there are any other similar loopholes, the new check in host_aton()
1939 itself should stop them being exploited. The report I received stated that
1940 data on the command line could provoke the exploit when Exim was running as
1941 exim, but did not say which command line option was involved. All I could
1942 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1943 running as the user.
1946 35. There was a buffer overflow vulnerability in the SPA authentication code
1947 (which came originally from the Samba project). I have added a test to the
1948 spa_base64_to_bits() function which I hope fixes it.
1951 36. The daemon start-up calls getloadavg() while still root for those OS that
1952 need the first call to be done as root, but it missed one case: when
1953 deliver_queue_load_max is set with deliver_drop_privilege. This is
1954 necessary for the benefit of the queue runner, because there is no re-exec
1955 when deliver_drop_privilege is set.
1957 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1958 was in use, was not putting the data itself into the right store pool;
1959 consequently, it could be overwritten for a subsequent message in the same
1960 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1963 38. Sometimes the final signoff response after QUIT could fail to get
1964 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1965 < 0 before doing a fflush(). This bug looks as though it goes back to the
1966 introduction of TLS in release 3.20, but "sometimes" must have been rare
1967 because the tests only now provoked it.
1973 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1974 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1975 Now it returns an integer. A number of calls were still expecting a T/F
1976 return. Fortuitously, in all cases, the tests worked in OK situations,
1977 which is the norm. However, things would have gone wrong on any write
1978 failures on the smtp file descriptor. This function is used when sending
1979 messages over SMTP and also when doing verify callouts.
1981 2. When Exim is called to do synchronous delivery of a locally submitted
1982 message (the -odf or -odi options), it no longer closes stderr before doing
1985 3. Implemented the mua_wrapper option.
1987 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1989 5. Implemented the functions header_remove(), header_testname(),
1990 header_add_at_position(), and receive_remove_recipient(), and exported them
1993 6. If an ACL "warn" statement specified the addition of headers, Exim already
1994 inserted X-ACL-Warn: at the start if there was no header name. However, it
1995 was not making this test for the second and subsequent header lines if
1996 there were newlines in the string. This meant that an invalid header could
1997 be inserted if Exim was badly configured.
1999 7. Allow an ACL "warn" statement to add header lines at the start or after all
2000 the Received: headers, as well as at the end.
2002 8. Added the rcpt_4xx retry error code.
2004 9. Added postmaster_mailfrom=xxx to callout verification option.
2006 10. Added mailfrom=xxxx to the callout verification option, for verify=
2009 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2010 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2012 12. Inserted some casts to stop certain compilers warning when using pointer
2013 differences as field lengths or precisions in printf-type calls (mostly
2014 affecting debugging statements).
2016 13. Added optional readline() support for -be (dynamically loaded).
2018 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2019 same clock tick as a message's arrival, so that its received time was the
2020 same as the "first fail" time on the retry record, and that message
2021 remained on the queue past the ultimate address timeout, every queue runner
2022 would try a delivery (because it was past the ultimate address timeout) but
2023 after another failure, the ultimate address timeout, which should have then
2024 bounced the address, did not kick in. This was a "< instead of <=" error;
2025 in most cases the first failure would have been in the next clock tick
2026 after the received time, and all would be well.
2028 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2029 being recognized when the domain list was tested by the match_domain
2030 condition in an expansion string.
2032 16. Added the ${str2b64: operator.
2034 17. Exim was always calling setrlimit() to set a large limit for the number of
2035 processes, without checking whether the existing limit was already
2036 adequate. (It did check for the limit on file descriptors.) Furthermore,
2037 errors from getrlimit() and setrlimit() were being ignored. Now they are
2038 logged to the main and panic logs, but Exim does carry on, to try to do its
2039 job under whatever limits there are.
2041 18. Imported PCRE 5.0.
2043 19. Trivial typo in log message " temporarily refused connection" (the leading
2046 20. If the log selector return_path_on_delivery was set and an address was
2047 redirected to /dev/null, the delivery process crashed because it assumed
2048 that a return path would always be set for a "successful" delivery. In this
2049 case, the whole delivery is bypassed as an optimization, and therefore no
2052 21. Internal re-arrangement: the function for sending a challenge and reading
2053 a response while authentication was assuming a zero-terminated challenge
2054 string. It's now changed to take a pointer and a length, to allow for
2055 binary data in such strings.
2057 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2059 23. Exim was not respecting finduser_retries when seeking the login of the
2060 uid under which it was called; it was always trying 10 times. (The default
2061 setting of finduser_retries is zero.) Also, it was sleeping after the final
2062 failure, which is pointless.
2064 24. Implemented tls_on_connect_ports.
2066 25. Implemented acl_smtp_predata.
2068 26. If the domain in control=submission is set empty, Exim assumes that the
2069 authenticated id is a complete email address when it generates From: or
2070 Sender: header lines.
2072 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2073 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2074 chown and chgrp in /bin and hostname in /usr/bin.
2076 28. Exim was keeping the "process log" file open after each use, just as it
2077 does for the main log. This opens the possibility of it remaining open for
2078 long periods when the USR1 signal hits a daemon. Occasional processlog
2079 errors were reported, that could have been caused by this. Anyway, it seems
2080 much more sensible not to leave this file open at all, so that is what now
2083 29. The long-running daemon process does not normally write to the log once it
2084 has entered its main loop, and it closes the log before doing so. This is
2085 so that log files can straightforwardly be renamed and moved. However,
2086 there are a couple of unusual error situations where the daemon does write
2087 log entries, and I had neglected to close the log afterwards.
2089 30. The text of an SMTP error response that was received during a remote
2090 delivery was being truncated at 512 bytes. This is too short for some of
2091 the long messages that one sometimes sees. I've increased the limit to
2094 31. It is now possible to make retry rules that apply only when a message has a
2095 specific sender, in particular, an empty sender.
2097 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2098 it possible to be selective about when SMTP synchronization is enforced.
2100 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2102 32. Implemented hosts_connection_nolog.
2104 33. Added an ACL for QUIT.
2106 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2109 35. Added mailbox_size and mailbox_filecount to appendfile.
2111 36. Added control = no_multiline_responses to ACLs.
2113 37. There was a bug in the logic of the code that waits for the clock to tick
2114 in the case where the clock went backwards by a substantial amount such
2115 that the microsecond fraction of "now" was more than the microsecond
2116 fraction of "then" (but the whole seconds number was less).
2118 38. Added support for the libradius Radius client library this is found on
2119 FreeBSD (previously only the radiusclient library was supported).
2125 1. When certain lookups returned multiple values in the form name=value, the
2126 quoting of the values was not always being done properly. Specifically:
2127 (a) If the value started with a double quote, but contained no whitespace,
2129 (b) If the value contained whitespace other than a space character (i.e.
2130 tabs or newlines or carriage returns) it was not quoted.
2131 This fix has been applied to the mysql and pgsql lookups by writing a
2132 separate quoting function and calling it from the lookup code. The fix
2133 should probably also be applied to nisplus, ibase and oracle lookups, but
2134 since I cannot test any of those, I have not disturbed their existing code.
2136 2. A hit in the callout cache for a specific address caused a log line with no
2137 reason for rejecting RCPT. Now it says "Previous (cached) callout
2138 verification failure".
2140 3. There was an off-by-one bug in the queryprogram router. An over-long
2141 return line was truncated at 256 instead of 255 characters, thereby
2142 overflowing its buffer with the terminating zero. As well as fixing this, I
2143 have increased the buffer size to 1024 (and made a note to document this).
2145 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2146 when Exim is waiting for an SMTP response from a remote server, Exim
2147 restarts its select() call on the socket, thereby resetting its timeout.
2148 This is not a problem when such interrupts are rare. Somebody set up a cron
2149 job to run exiwhat every 2 minutes, which is less than the normal select()
2150 timeout (5 or 10 minutes). This meant that the select() timeout never
2151 kicked in because it was always reset. I have fixed this by comparing the
2152 time when an interrupt arrives with the time at the start of the first call
2153 to select(). If more time than the timeout has elapsed, the interrupt is
2154 treated as a timeout.
2156 5. Some internal re-factoring in preparation for the addition of Sieve
2157 extensions (by MH). In particular, the "personal" test is moved to a
2158 separate function, and given an option for scanning Cc: and Bcc: (which is
2159 not set for Exim filters).
2161 6. When Exim created an email address using the login of the caller as the
2162 local part (e.g. when creating a From: or Sender: header line), it was not
2163 quoting the local part when it contained special characters such as @.
2165 7. Installed new OpenBSD configuration files.
2167 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2168 try to make them clearer.
2170 9. Callout options, other than the timeout value, were being ignored when
2171 verifying sender addresses in header lines. For example, when using
2173 verify = header_sender/callout=no_cache
2175 the cache was (incorrectly) being used.
2177 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2178 only the Cygwin environment.
2180 11. When return_path_on_delivery was set as a log selector, if different remote
2181 addresses in the same message used different return paths and parallel
2182 remote delivery occurred, the wrong values would sometimes be logged.
2183 (Whenever a remote delivery process finished, the return path value from
2184 the most recently started remote delivery process was logged.)
2186 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2187 lines when AUTH and/or TLS are in use. This is the "received protocol"
2188 field. Exim used to use "asmtp" for authenticated SMTP, without any
2189 indication (in the protocol name) for TLS use. Now it follows the RFC and
2190 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2191 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2192 names appear in log lines as well as in Received: header lines.
2194 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2195 extensions, and comparison tests, and to fix some bugs.
2197 14. Changes to the "personal" filter test:
2199 (1) The test was buggy in that it was just doing the equivalent of
2200 "contains" tests on header lines. For example, if a user's address was
2201 anne@some.where, the "personal" test would incorrectly be true for
2203 To: susanne@some.where
2205 This test is now done by extracting each address from the header in turn,
2206 and checking the entire address. Other tests that are part of "personal"
2207 are now done using regular expressions (for example, to check local parts
2208 of addresses in From: header lines).
2210 (2) The list of non-personal local parts in From: addresses has been
2211 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2212 taken from the Sieve specification recommendations.
2214 (3) If the message contains any header line starting with "List-" it is
2215 treated as non-personal.
2217 (4) The test for "circular" in the Subject: header line has been removed
2218 because it now seems ill-conceived.
2220 15. Minor typos in src/EDITME comments corrected.
2222 16. Installed latest exipick from John Jetmore.
2224 17. If headers_add on a router specified a text string that was too long for
2225 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2226 of string_sprintf() is now avoided.
2228 18. $message_body_size was not set (it was always zero) when running the DATA
2229 ACL and the local_scan() function.
2231 19. For the "mail" command in an Exim filter, no default was being set for
2232 the once_repeat time, causing a random time value to be used if "once" was
2233 specified. (If the value happened to be <= 0, no repeat happened.) The
2234 default is now 0s, meaning "never repeat". The "vacation" command was OK
2235 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2236 (I found it when inspecting the code).
2238 20. There is now an overall timeout for performing a callout verification. It
2239 defaults to 4 times the callout timeout, which applies to individual SMTP
2240 commands during the callout. The overall timeout applies when there is more
2241 than one host that can be tried. The timeout is checked before trying the
2242 next host. This prevents very long delays if there are a large number of
2243 hosts and all are timing out (e.g. when the network connections are timing
2244 out). The value of the overall timeout can be changed by specifying an
2245 additional sub-option for "callout", called "maxwait". For example:
2247 verify = sender/callout=5s,maxwait=20s
2249 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2250 to the end before writing, but this should make it even safer).
2252 22. Exim was forgetting that it had advertised PIPELINING for the second and
2253 subsequent messages on an SMTP connection. It was also not resetting its
2254 memory on STARTTLS and an internal HELO.
2256 23. When Exim logs an SMTP synchronization error within a session, it now
2257 records whether PIPELINING has been advertised or not.
2259 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2260 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2261 rather than long int.
2263 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2265 26. Added the never_mail option to autoreply.
2271 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2272 crash if the getsockname() call failed; this can happen if a connection is
2273 closed very soon after it is established. The problem was simply in the
2274 order in which certain operations were done, causing Exim to try to write
2275 to the SMTP stream before it had set up the file descriptor. The bug has
2276 been fixed by making things happen in the correct order.
2282 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2283 before the connection was closed, thus losing the rejection response.
2285 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2286 some early Solaris releases, but causes trouble in current releases where
2287 socklen_t is defined.
2289 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2292 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2295 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2296 that releasing the top of it at the end releases what was used for sub-
2297 expansions (unless the block got too big). However, discard this block if
2298 the first thing is a variable or header, so that we can use its block when
2299 it is dynamic (useful for very large $message_headers, for example).
2301 6. Lookups now cache *every* query, not just the most recent. A new, separate
2302 store pool is used for this. It can be recovered when all lookup caches are
2303 flushed. Lookups now release memory at the end of their result strings.
2304 This has involved some general refactoring of the lookup sources.
2306 7. Some code has been added to the store_xxx() functions to reduce the amount
2307 of flapping under certain conditions.
2309 8. log_incoming_interface used to affect only the <= reception log lines. Now
2310 it causes the local interface and port to be added to several more SMTP log
2311 lines, for example "SMTP connection from", and rejection lines.
2313 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2315 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2317 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2318 could be overwritten at the end of the current message (or the start of a
2319 new message if it was set in a HELO ACL). The value is now preserved for
2320 the duration of the SMTP connection.
2322 12. If a transport had a headers_rewrite setting, and a matching header line
2323 contained an unqualified address, that address was qualified, even if it
2324 did not match any rewriting rules. The underlying bug was that the values
2325 of the flags that permit the existence of unqualified sender and recipient
2326 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2327 non-local messages, and by -bnq for local messages) were not being
2328 preserved with the message after it was received.
2330 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2331 "next input=" as part of the text comprising the host identity instead of
2332 the correct text. The code was using the same buffer for two different
2333 strings. However, depending on which order the printing function evaluated
2334 its arguments, the bug did not always show up. Under Linux, for example, my
2335 test suite worked just fine.
2337 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2338 doesn't work with some older versions of Perl. It has been changed to "my",
2339 which in any case is probably the better facility to use.
2341 15. A really picky compiler found some instances of statements for creating
2342 error messages that either had too many or two few arguments for the format
2345 16. The size of the buffer for calls to the DNS resolver has been increased
2346 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2347 for addresses that have a lot of PTR records. This alleviates a problem; it
2348 does not fully solve it.
2350 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2351 buffer now truncates the list and logs the incident, which is the same
2352 action as happens when Exim is looking up a host name and its aliases.
2353 Previously in this situation something unpredictable would happen;
2354 sometimes it was "internal error: store_reset failed".
2356 18. If a server dropped the connection unexpectedly when an Exim client was
2357 using GnuTLS and trying to read a response, the client delivery process
2358 crashed while trying to generate an error log message.
2360 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2361 string, for example:
2363 warn message = H1: something\nH2: something
2365 the text was added as a single header line from Exim's point of view
2366 though it ended up OK in the delivered message. However, searching for the
2367 second and subsequent header lines using $h_h2: did not work. This has been
2368 fixed. Similarly, if a system filter added multiple headers in this way,
2369 the routers could not see them.
2371 20. Expanded the error message when iplsearch is called with an invalid key to
2372 suggest using net-iplsearch in a host list.
2374 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2375 ACLs are no longer actually imposed (and a message to that effect is
2378 22. If a "gecos" field in a passwd entry contained escaped characters, in
2379 particular, if it contained a \" sequence, Exim got it wrong when building
2380 a From: or a Sender: header from that name. A second bug also caused
2381 incorrect handling when an unquoted " was present following a character
2382 that needed quoting.
2384 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2385 was not being matched caselessly.
2387 24. Arranged for all hyphens in the exim.8 source to be escaped with
2390 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2391 properly. Recipient callout cache records were still being keyed to include
2392 the sender, even when use_sender was set false. This led to far more
2393 callouts that were necessary. The sender is no longer included in the key
2394 when use_sender is false.
2396 26. Added "control = submission" modifier to ACLs.
2398 27. Added the ${base62d: operator to decode base 62 numbers.
2400 28. dnsdb lookups can now access SRV records.
2402 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2403 the configuration file.
2405 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2406 (-v) mode. This makes the output for a verbose queue run more intelligible.
2408 31. Added a use_postmaster feature to recipient callouts.
2410 32. Added the $body_zerocount variable, containing the number of binary zero
2411 bytes in the message body.
2413 33. The time of last modification of the "new" subdirectory is now used as the
2414 "mailbox time last read" when there is a quota error for a maildir
2417 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2419 35. Added +ignore_unknown as a special item in host lists.
2421 36. Code for decoding IPv6 addresses in host lists is now included, even if
2422 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2423 address was recognized as an IP address, but was then not correctly decoded
2424 into binary, causing unexpected and incorrect effects when compared with
2431 1. Very minor rewording of debugging text in manualroute to say "list of
2432 hosts" instead of "hostlist".
2434 2. If verify=header_syntax was set, and a header line with an unqualified
2435 address (no domain) and a large number of spaces between the end of the
2436 name and the colon was received, the reception process suffered a buffer
2437 overflow, and (when I tested it) crashed. This was caused by some obsolete
2438 code that should have been removed. The fix is to remove it!
2440 3. When running in the test harness, delay a bit after writing a bounce
2441 message to get a bit more predictability in the log output.
2443 4. Added a call to search_tidyup() just before forking a reception process. In
2444 theory, someone could use a lookup in the expansion of smtp_accept_max_
2445 per_host which, without the tidyup, could leave open a database connection.
2447 5. Added the variables $recipient_data and $sender_data which get set from a
2448 lookup success in an ACL "recipients" or "senders" condition, or a router
2449 "senders" option, similar to $domain_data and $local_part_data.
2451 6. Moved the writing of debug_print from before to after the "senders" test
2454 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2455 problems for message scanning, either using a data ACL, or using
2456 local_scan() because the Received: header was not generated till after they
2457 were called (in order to set the time as the time of reception completion).
2458 I have revised the way this works. The header is now generated after the
2459 body is received, but before the ACL or local_scan() are called. After they
2460 are run, the timestamp in the header is updated.
2466 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2467 before starting a queue runner without re-exec. This happened only when
2468 deliver_drop_privilege was set or when the Exim user was set to root. The
2469 effect of the bug was that timeouts during subsequent deliveries caused
2470 crashes instead of being properly handled. The handler is now left at its
2471 default (and expected) setting.
2473 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2474 message, again when deliver_drop_privilege is set or Exim is run as root.
2475 The bug described in (1) was not present in this case, but the tidying up
2476 of the other signals was missing. I have made the two cases consistent.
2478 3. The ignore_target_hosts setting on a manualroute router was being ignored
2479 for hosts that were looked up using the /MX notation.
2481 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2484 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2485 operated on the sender address. After changing the $sender_address to <>
2486 for the sender address verify, Exim was re-instated it as the original
2487 (before rewriting) address, but remembering that it had rewritten it, so it
2488 wasn't rewriting it again. This bug also had the effect of breaking the
2489 sender address verification caching when the sender address was rewritten.
2491 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2492 This has been changed so that if the ip literal address matches
2493 ignore_target_hosts, the router declines.
2495 7. Added expansion conditions match_domain, match_address, and match_local_
2496 part (NOT match_host).
2498 8. The placeholder for the Received: header didn't have a length field set.
2500 9. Added code to Exim itself and to exim_lock to test for a specific race
2501 condition that could lead to file corruption when using MBX delivery. The
2502 issue is with the lockfile that is created in /tmp. If this file is removed
2503 after a process has opened it but before that process has acquired a lock,
2504 there is the potential for a second process to recreate the file and also
2505 acquire a lock. This could lead to two Exim processes writing to the file
2506 at the same time. The added code performs the same test as UW imapd; it
2507 checks after acquiring the lock that its file descriptor still refers to
2508 the same named file.
2510 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2511 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2514 11. Added the smtp_active_hostname option. If used, this will typically be made
2515 to depend on the incoming interface address. Because $interface_address is
2516 not set up until the daemon has forked a reception process, error responses
2517 that can happen earlier (such as "too many connections") no longer contain
2520 12. If an expansion in a condition on a "warn" statement fails because a lookup
2521 defers, the "warn" statement is abandoned, and the next ACL statement is
2522 processed. Previously this caused the whole ACL to be aborted.
2524 13. Added the iplsearch lookup type.
2526 14. Added ident_timeout as a log selector.
2528 15. Added tls_certificate_verified as a log selector.
2530 16. Added a global option tls_require_ciphers (compare the smtp transport
2531 option of the same name). This controls incoming TLS connections.
2533 17. I finally figured out how to make tls_require_ciphers do a similar thing
2534 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2535 before starting the TLS session.
2537 18. Tabs are now shown as \t in -bP output.
2539 19. If the log selector return_path_on_delivery was set, Exim crashed when
2540 bouncing a message because it had too many Received: header lines.
2542 20. If two routers both had headers_remove settings, and the first one included
2543 a superfluous trailing colon, the final name in the first list and the
2544 first name in the second list were incorrectly joined into one item (with a
2545 colon in the middle).
2551 1. Added -C and -D options to the exinext utility, mainly to make it easier
2552 to include in the automated testing, but these could be helpful when
2553 multiple configurations are in use.
2555 2. The exinext utility was not formatting the output nicely when there was
2556 an alternate port involved in the retry record key, nor when there was a
2557 message id as well (for retries that were specific to a specific message
2558 and a specific host). It was also confused by IPv6 addresses, because of
2559 the additional colons they contain. I have fixed the IPv4 problem, and
2560 patched it up to do a reasonable job for IPv6.
2562 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2563 delivery, the log line now contains "pipelined" if PIPELINING was used.
2565 4. An SMTP transport process used to panic and die if the bind() call to set
2566 an explicit outgoing interface failed. This has been changed; it is now
2567 treated in the same way as a connect() failure.
2569 5. A reference to $sender_host_name in the part of a conditional expansion
2570 that was being skipped was still causing a DNS lookup. This no longer
2573 6. The def: expansion condition was not recognizing references to header lines
2574 that used bh_ and bheader_.
2576 7. Added the _cache feature to named lists.
2578 8. The code for checking quota_filecount in the appendfile transport was
2579 allowing one more file than it should have been.
2581 9. For compatibility with Sendmail, the command line option
2589 and sets the incoming protocol and host name (for trusted callers). The
2590 host name and its colon can be omitted when only the protocol is to be set.
2591 Note the Exim already has two private options, -pd and -ps, that refer to
2592 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2593 "s", but I don't think that's a major issue.
2595 10. A number of refactoring changes to the code, none of which should affect
2598 (a) The number of logging options was getting close to filling up the
2599 32-bit word that was used as a bit map. I have split them into two classes:
2600 those that are passed in the argument to log_write(), and those that are
2601 only ever tested independently outside of that function. These are now in
2602 separate 32-bit words, so there is plenty of room for expansion again.
2603 There is no change in the user interface or the logging behaviour.
2605 (b) When building, for example, log lines, the code previously used a
2606 macro that called string_cat() twice, in order to add two strings. This is
2607 not really sufficiently general. Furthermore, there was one instance where
2608 it was actually wrong because one of the argument was used twice, and in
2609 one call a function was used. (As it happened, calling the function twice
2610 did not affect the overall behaviour.) The macro has been replaced by a
2611 function that can join an arbitrary number of extra strings onto a growing
2614 (c) The code for expansion conditions now uses a table and a binary chop
2615 instead of a serial search (which was left over from when there were very
2616 few conditions). Also, it now recognizes conditions like "pam" even when
2617 the relevant support is not compiled in: a suitably worded error message is
2618 given if an attempt is made to use such a condition.
2620 11. Added ${time_interval:xxxxx}.
2622 12. A bug was causing one of the ddress fields not to be passed back correctly
2623 from remote delivery subprocesses. The field in question was not being
2624 subsequently used, so this caused to problems in practice.
2626 13. Added new log selectors queue_time and deliver_time.
2628 14. Might have fixed a bug in maildirsizefile handling that threw up
2629 "unexpected character" debug warnings, and recalculated the data
2630 unnecessarily. In any case, I expanded the warning message to give more
2633 15. Added the message "Restricted characters in address" to the statements in
2634 the default ACL that block characters like @ and % in local parts.
2636 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2637 Three changes have been made:
2639 (a) There was a serious bug; a negative response to MAIL caused the whole
2640 recipient domain to be cached as invalid, thereby blocking all messages
2641 to all local parts at the same domain, from all senders. This bug has
2642 been fixed. The domain is no longer cached after a negative response to
2643 MAIL if the sender used is not empty.
2645 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2648 (c) A new callout option, "use_sender" has been added for people who want
2649 the modified behaviour.
2655 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2658 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2659 indeed breaks things for older releases.
2661 3. Added additional logging to the case where there is a problem reading data
2662 from a filter that is running in a subprocess using a pipe, in order to
2663 try to track down a specific problem.
2665 4. Testing facility fudge: when running in the test harness and attempting
2666 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2667 getting "No route to host". Convert this to a timeout.
2669 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2672 6. Some OS don't have socklen_t but use size_t instead. This affects the
2673 fifth argument of getsockopt() amongst other things. This is now
2674 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2675 can be set for individual OS. I have set it for SunOS5, OSF1, and
2676 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2677 some earlier ones do not.
2679 7. Change 4.30/15 was not doing the test caselessly.
2681 8. The standard form for an IPv6 address literal was being rejected by address
2682 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2683 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2684 this, as well as the form without the "IPv6" on the front (but only when
2685 address literals are enabled, of course).
2687 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2689 10. Exim crashed if a message with an empty sender address specified by -f
2690 encountered a router with an errors_to setting. This could be provoked only
2691 by a command such as
2695 where an empty string was supplied; "<>" did not hit this bug.
2697 11. Installed PCRE release 4.5.
2699 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2700 remained set. It is now erased.
2702 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2703 times from message ids (which are base 36 rather than the normal 62).
2705 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2706 were being counted as actual protocol errors, and logged if the log
2707 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2708 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2709 and DATA following a set of rejected RCPTs do not count as protocol errors.
2710 In other words, Exim assumes they were pipelined, though this may not
2711 actually be the case. Of course, in all cases the client gets an
2712 appropriate error code.
2714 15. If a lookup fails in an ACL condition, a message about the failure may
2715 be available; it is used if testing the ACL cannot continue, because most
2716 such messages specify what the cause of the deferral is. However, some
2717 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2718 that caused an old message to be retained and used if a later statement
2719 caused a defer, replacing the real cause of the deferral.
2721 16. If an IP address had so many PTR records that the DNS lookup buffer
2722 was not large enough to hold them, Exim could crash while trying to process
2723 the truncated data. It now detects and logs this case.
2725 17. Further to 4.21/58, another change has been made: if (and only if) the
2726 first line of a message (the first header line) ends with CRLF, a bare LF
2727 in a subsequent header line has a space inserted after it, so as not to
2728 terminate the header.
2730 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2731 unnecessarily, used atoi() instead of strtol(), and didn't check the
2732 termination when getting file sizes from file names by regex.
2734 19. Completely re-implemented the support for maildirsize files, in the light
2735 of a number of problems with the previous contributed implementation
2736 (4.30/29). In particular:
2738 . If the quota is zero, the maildirsize file is maintained, but no quota is
2741 . If the maildir directory does not exist, it is created before any attempt
2742 to write a maildirsize file.
2744 . The quota value in the file is just a cache; if the quota is changed in
2745 the transport, the new value overrides.
2747 . A regular expression is available for excluding directories from the
2750 20. The autoreply transport checks the characters in options that define the
2751 message's headers; it allows continued headers, but it was checking with
2752 isspace() after an embedded newline instead of explicitly looking for a
2755 21. If all the "regular" hosts to which an address was routed had passed their
2756 expiry times, and had not reached their retry times, the address was
2757 bounced, even if fallback hosts were defined. Now Exim should go on to try
2760 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2761 equivalent code in the SMTP transport. Some hosts send humungous responses
2762 to HELO/EHLO, more than 1024 it seems.
2764 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2765 gives compiler warnings in some environments. I've now done it "properly",
2768 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2769 (because of problems with the built-in one) was declared to return uschar *
2770 instead of char *, causing compiler failure.
2772 25. Fixed a file descriptor leak when processing alias/forward files.
2774 26. Fixed a minor format string issue in dbfn.c.
2776 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2778 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2779 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2782 29. When a sender address is verified, it is cached, to save repeating the test
2783 when there is more than one recipient in a message. However, when the
2784 verification involves a callout, it is possible for different callout
2785 options to be set for different recipients. It is too complicated to keep
2786 track of this in the cache, so now Exim always runs a verification when a
2787 callout is required, relying on the callout cache for the optimization.
2788 The overhead is duplication of the address routing, but this should not be
2791 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2792 to be verified with callout=postmaster, and the main callout worked but the
2793 postmaster check failed, the verification correctly failed. However, if a
2794 subsequent RCPT command asked for sender verification *without* the
2795 postmaster check, incorrect caching caused this verification also to fail,
2798 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2799 it was not caching the DNS options (qualify_single, search_parents) that
2800 were used when the lookup failed. A subsequent lookup with different
2801 options therefore always gave the same answer, though there were cases
2802 where it should not have. (Example: a "domains = !$mx_any" option on a
2803 dnslookup router: the "domains" option is always processed without any
2804 widening, but the router might have qualify_single set.) Now Exim uses the
2805 cached value only when the same options are set.
2807 32. Added John Jetmore's "exipick" utility to the distribution.
2809 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2810 than a timeout (e.g. a certificate is required, and is not provided), an
2811 Exim server now closes the connection immediately. Previously it waited for
2812 the client to close - but if the client is SSL, it seems that they each
2813 wait for each other, leading to a delay before one of them times out.
2815 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2816 maintained 0.8.x compatibility because I don't think many are using it, and
2817 it is clearly obsolete.
2819 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2822 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2823 client certificate was expired. A simple patch fixes this, though I don't
2824 understand the full logic of why the verify callback is called multiple
2827 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2828 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2829 which causes problems with some clients (such as the Certicom SSL Plus
2830 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2831 disables the coutermeasure allowing Eudora to connect."
2833 38. Exim was not checking that a write() to a log file succeeded. This could
2834 lead to Bad Things if a log got too big, in particular if it hit a file
2835 size limit. Exim now panics and dies if it cannot write to a log file, just
2836 as it does if it cannot open a log file.
2838 39. Modified OS/Makefile-Linux so that it now contains
2840 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2842 The two -D definitions ensure that Exim is compiled with large file
2843 support, which makes it possible to handle log files that are bigger than
2846 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2847 instance) a domain was checked against a named list that involved a lookup,
2848 causing $domain_data to be set, then another domain was checked against the
2849 same list, then the first domain was re-checked, the value of $domain_data
2850 after the final check could be wrong. In particular, if the second check
2851 failed, it could be set empty. This bug probably also applied to
2854 41. The strip_trailing_dot option was not being applied to the address given
2855 with the -f command-line option.
2857 42. The code for reading a message's header from the spool was incrementing
2858 $received_count, but never initializing it. This meant that the value was
2859 incorrect (doubled) while delivering a message in the same process in which
2860 it was received. In the most common configuration of Exim, this never
2861 happens - a fresh exec is done - but it can happen when
2862 deliver_drop_privilege is set.
2864 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2865 it now includes up to 150 characters of the unexpected data in the log
2868 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2869 and building data strings. The size of both of these buffers was 10 000
2870 bytes - far larger than anybody would *ever* want, thought I. Needless to
2871 say, somebody hit the limit. I have increased the maximum line length to
2872 20 000 and the maximum data length of concatenated lines to 100 000. I have
2873 also fixed two bugs, because there was no checking on these buffers. Tsk,
2874 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2875 buffer is too small.
2877 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2878 lsearch lookups. Now it does.
2880 46. When parsing a route_list item in a manualroute router, a fixed-length
2881 buffer was used for the list of hosts. I made this 1024 bytes long,
2882 thinking that nobody would ever have a list of hosts that long. Wrong.
2883 Somebody had a whole pile of complicated expansion conditions, and the
2884 string was silently truncated, leading to an expansion error. It turns out
2885 that it is easier to change to an unlimited length (owing to other changes
2886 that have happened since this code was originally written) than to build
2887 structure for giving a limitation error. The length of the item that
2888 expands into the list of hosts is now unlimited.
2890 47. The lsearch lookup could not handle data where the length of text line was
2891 more than 4095 characters. Such lines were truncated, leading to shortened
2892 data being returned. It should now handle lines of any length.
2894 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2895 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2898 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2899 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2900 no longer happens in comment lines. A list of replacements is now placed
2901 at the head of all of the source files, except those whose only change is
2902 to replace PERL_COMMAND in the very first #! line.
2904 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2905 messages on the queue, with a bottom-up merge sort, using code contributed
2906 by Michael Haardt. This should make operations like -bp somewhat faster on
2907 large queues. It won't affect queue runners, except when queue_run_in_order
2910 51. Installed eximstats 1.31 in the distribution.
2912 52. Added support for SRV lookups to the dnslookup router.
2914 53. If an ACL referred to $message_body or $message_body_end, the value was not
2915 reset for any messages that followed in the same SMTP session.
2917 54. The store-handling optimization for building very long strings was not
2918 differentiating between the different store pools. I don't think this
2919 actually made any difference in practice, but I've tidied it.
2921 55. While running the routers to verify a sender address, $sender_address
2922 was still set to the sender address. This is wrong, because when routing to
2923 send a bounce to the sender, it would be empty. Therefore, I have changed
2924 it so that, while verifying a sender address, $sender_address is set to <>.
2925 (There is no change to what happens when verifying a recipient address.)
2927 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2928 target A or AAAA records (if not already returned) without resetting the
2929 qualify_single or search_parents options of the DNS resolver. These are
2930 inappropriate in this case because the targets of MX and SRV records must
2931 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2932 target that, when qualified, matched something in the local domain. These
2933 two options are now turned off when doing these lookups.
2935 57. It seems that at least some releases of Reiserfs (which does not have the
2936 concept of a fixed number of inodes) returns zero and not -1 for the
2937 number of available inodes. This interacted badly with check_spool_inodes,
2938 which assumed that -1 was the "no such thing" setting. What I have done is
2939 to check that the total number of inodes is greater than zero before doing
2940 the test of how many are available.
2942 58. When a "warn" ACL statement has a log_message modifier, the message is
2943 remembered, and not repeated. This is to avoid a lot of repetition when a
2944 message has many recipients that cause the same warning to be written.
2945 Howewer, Exim was preserving the list of already written lines for an
2946 entire SMTP session, which doesn't seem right. The memory is now reset if a
2947 new message is started.
2949 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2950 debugging output unless log_rewrite was also set.
2952 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2953 of (int)(handle) when we know that handle contains (void *)(-1).
2955 61. The Exim daemon panic-logs an error return when it closes the incoming
2956 connection. However "connection reset by peer" seems to be common, and
2957 isn't really an error worthy of noting specially, so that particular error
2960 62. When Exim is trying to find all the local interfaces, it used to panic and
2961 die if the ioctl to get the interface flags failed. However, it seems that
2962 on at least one OS (Solaris 9) it is possible to have an interface that is
2963 included in the list of interfaces, but for which you get a failure error
2964 for this call. This happens when the interface is not "plumbed" into a
2965 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2966 failure of the "get flags" call assumes that the interface is down.
2968 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2969 makes life easier for people who are doing arithmetic on fields extracted
2970 from dates, where you often get leading zeros that should not be
2971 interpreted as octal.
2973 64. Added qualify_domain to the redirect router, to override the global
2976 65. If a pathologically long header line contained very many addresses (the
2977 report of this problem mentioned 10 000) and each of them was rewritten,
2978 Exim could use up a very large amount of memory. (It kept on making new
2979 copies of the header line as it rewrote, and never released the old ones.)
2980 At the expense of a bit more processing, the header rewriting function has
2981 been changed so that it no longer eats memory in this way.
2983 66. The generation of the Received: header has been moved from the time that a
2984 message starts to be received, to the time that it finishes. The timestamp
2985 in the Received: header should now be very close to that of the <= log
2986 line. There are two side-effects of this change:
2988 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2989 logged header lines no longer include the local Received: line, because
2990 it has not yet been created. The same applies to a copy of the message
2991 that is returned to a non-SMTP sender when a message is rejected.
2993 (b) When a filter file is tested using -bf, no additional Received: header
2994 is added to the test message. After some thought, I decided that this
2997 This change does not affect the value of $received_for. It is still set
2998 after address rewriting, but before local_scan() is called.
3000 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3002 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3003 gave an unhelpful panic error message, and a defer error. I have managed to
3004 change this behaviour so that it now rejects any supplied certificate,
3005 which seems right, as the list of acceptable certificates is empty.
3007 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3008 gave an unhelpful defer error. I have not managed to make this reject any
3009 supplied certificates, but the error message it gives is "no certificate
3010 supplied", which is not helpful.
3012 70. exigrep's output now also includes lines that are not associated with any
3013 message, but which match the given pattern. Implemented by a patch from
3014 Martin Sluka, which also tidied up the Perl a bit.
3016 71. Recipient callout verification, like sender verification, was using <> in
3017 the MAIL FROM command. This isn't really the right thing, since the actual
3018 sender may affect whether the remote host accepts the recipient or not. I
3019 have changed it to use the actual sender in the callout; this means that
3020 the cache record is now keyed on a recipient/sender pair, not just the
3021 recipient address. There doesn't seem to be a real danger of callout loops,
3022 since a callout by the remote host to check the sender would use <>.
3023 [SEE ABOVE: changed after hitting problems.]
3025 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3026 temporary errors. However, in the case of such a code being given after
3027 the end of a data transmission (i.e. after ".") Exim was failing to write
3028 a retry record for the message. (Yes, there was some broken host that was
3029 actually sending 8xx at this point.)
3031 73. An unknown lookup type in a host list could cause Exim to panic-die when
3032 the list was checked. (An example that provoked this was putting <; in the
3033 middle of a list instead of at the start.) If this happened during a DATA
3034 ACL check, a -D file could be left lying around. This kind of configuration
3035 error no longer causes Exim to die; instead it causes a defer errror. The
3036 incident is still logged to the main and panic logs.
3038 74. Buglet left over from Exim 3 conversion. The message "too many messages
3039 in one connection" was written to the rejectlog but not the mainlog, except
3040 when address rewriting (yes!) was being logged.
3042 75. Added write_rejectlog option.
3044 76. When a system filter was run not as root (that is, when system_filter_user
3045 was set), the values of the $n variables were not being returned to the
3046 main process; thus, they were not subsequently available in the $sn
3049 77. Added +return_path_on_delivery log selector.
3051 78. A connection timeout was being treated differently from recipients deferred
3052 when testing hosts_max_try with a message that was older than the host's
3053 retry timeout. (The host should not be counted, thus allowing all hosts to
3054 be tried at least once before bouncing.) This may have been the cause of an
3055 occasionally reported bug whereby a message would remain on the queue
3056 longer than the retry timeout, but would be bounced if a delivery was
3057 forced. I say "may" because I never totally pinned down the problem;
3058 setting up timeout/retry tests is difficult. See also the next item.
3060 79. The ultimate address timeout was not being applied to errors that involved
3061 a combination of host plus message (for example, a timeout on a MAIL
3062 command). When an address resolved to a number of possible hosts, and they
3063 were not all tried for each delivery (e.g. because of hosts_max_try), a
3064 message could remain on the queue longer than the retry timeout.
3066 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3069 81. Fixed an obscure SMTP outgoing bug which required at least the following
3070 conditions: (a) there was another message waiting for the same server;
3071 (b) the server returned 5xx to all RCPT commands in the first message so
3072 that the message was not completed; (c) the server dropped the connection
3073 or gave a negative response to the RSET that Exim sends to abort the
3074 transaction. The observed case was a dropped connection after DATA that had
3075 been sent in pipelining mode. That is, the server had advertised PIPELINING
3076 but was not implementing it correctly. The effect of the bug was incorrect
3077 behaviour, such as trying another host, and this could lead to a crash.
3083 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3084 and daemon.c were passed as pointers to ints; they should have been
3085 pointers to socklen_t variables (which are typically unsigned ints).
3087 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3090 3. Fixed a really odd bug that affected only the testing scheme; patching a
3091 certain fixed string in the binary changed the value of another string that
3092 happened to be identical to the end of the original first string.
3094 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3095 name", it returns that address as the IP address. On some operating
3096 systems (e.g. Solaris), it also passes back the IP address string as the
3097 "host name". However, on others (e.g. Linux), it passes back an empty
3098 string. Exim wasn't checking for this, and was changing the host name to an
3099 empty string, assuming it had been canonicized.
3101 5. Although rare, it is permitted to have more than one PTR record for a given
3102 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3103 all the names associated with an address, because they do in Solaris.
3104 However, it seems that they do not in Linux for data that comes from the
3105 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3106 I found this out when I moved to a new Linux workstation and tried to run
3107 the Exim test suite.
3109 To get round this problem I have changed the code so that it now does its
3110 own call to the DNS to look up PTR records when searching for a host name.
3111 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3112 addresses that are only in /etc/hosts are still found.
3114 This behaviour is, however, controlled by an option called host_lookup_
3115 order, which defaults to "bydns:byaddr". If people want to use the other
3116 order, or indeed, just use one or the other means of lookup, they can
3117 specify it in this variable.
3119 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3120 some operating systems, this comes back from gethostbyaddr() as an empty
3121 string, and this is what Exim used to test for. However, it seems that in
3122 other systems, "." is yielded. Exim now tests for this case too.
3124 7. The values of check_spool_space and check_log_space are now held internally
3125 as a number of kilobytes instead of an absolute number of bytes. If a
3126 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3127 kilobyte. This means that much larger values can be stored.
3129 8. Exim monitor: an attempt to get the action menu when not actually pointing
3130 at a message produces an empty menu entitled "No message selected". This
3131 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3132 no entries in it ("Shell widget menu has zero width and/or height"). So I
3133 have added a single, blank menu entry in this case.
3135 9. Added ${quote_local_part.
3137 10. MIME decoding is now applied to the contents of Subject: header lines when
3140 11. Now that a reference to $sender_host_address automatically causes a reverse
3141 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3142 host lookup before query-style lookups in lists that might use this
3143 variable. This has therefore been abolished, and the "net-" prefix is no
3144 longer necessary for query-style lookups.
3146 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3147 have been a typo for LFLAGS, so it has been changed.
3149 13. The install script calls Exim with "-C /dev/null" in order to find the
3150 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3151 to be output. Howeve, since Exim outputs its version number before the
3152 error, it didn't break the script. It just looked ugly. I fixed this by
3153 always allowing "-C /dev/null" if the caller is root.
3155 14. Ignore overlarge ACL variable number when reading spool file - insurance
3156 against a later release with more variables having written the file.
3158 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3159 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3160 this, as well as the form without the "IPv6" on the front.
3162 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3163 OS/Makefile-Darwin file.
3165 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3166 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3168 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3169 variable. However, some OS use ints for inodes. I've added cast to long int
3170 to get rid of the compiler warning.
3172 19. I had forgotten to lock out "/../" in configuration file names when
3173 ALT_CONFIG_PREFIX was set.
3175 20. Routers used for verification do not need to specify transports. However,
3176 if such a router generated a host list, and callout was configured, Exim
3177 crashed, because it could not find a port number from the (non-existent)
3178 transport. It now assumes port 25 in this circumstance.
3180 21. Added the -t option to exigrep.
3182 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3183 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3184 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3185 them is set, LOOKUP_LSEARCH is forced.
3187 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3189 24. Added sender and host information to the "rejected by local_scan()" log
3190 line; previously there was no indication of these.
3192 25. Added .include_if_exists.
3194 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3195 Linux. It turns out that not all file systems support this. Apparently some
3196 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3197 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3198 not supported on the file descriptor, is now ignored when Exim is trying to
3199 sync a directory. This applies only to Linux.
3201 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3203 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3204 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3207 29. Added support for maildirsize files from supplied patch (modified a bit).
3209 30. The use of :fail: followed by an empty string could lead Exim to respond to
3210 sender verification failures with (e.g.):
3212 550 Verification failed for <xxx>
3213 550 Sender verify failed
3215 where the first response line was missing the '-' that indicates it is not
3216 the final line of the response.
3218 31. The loop for finding the name of the user that called Exim had a hardwired
3219 limit of 10; it now uses the value of finduser_retries, which is used for
3220 all other user lookups.
3222 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3225 33. Exim was neglecting to zero errno before one call of strtol() when
3226 expanding a string and expecting an integer value. On some systems this
3227 resulted in spurious "integer overflow" errors. Also, it was casting the
3228 result into an int without checking.
3230 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3231 did not work. The code looks as if it has *never* worked, though it appears
3232 to have been documented since at least releast 1.62. I have made it work.
3234 35. The "timeout_DNS" error in retry rules, also documented since at least
3235 1.62, also never worked. As it isn't clear exactly what this means, and
3236 clearly it isn't a major issue, I have abolished the feature by treating it
3237 as "timeout", and writing a warning to the main and panic logs.
3239 36. The display of retry rules for -brt wasn't always showing the error code
3242 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3243 timeout_connect_A, timeout_connect_MX.
3245 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3246 to the empty sender.
3248 39. The daemon was not analysing the content of -oX till after it had closed
3249 stderr and disconnected from the controlling terminal. This meant that any
3250 syntax errors were only noted on the panic log, and the return code from
3251 the command was 0. By re-arranging the code a little, I've made the
3252 decoding happen first, so such errors now appear on stderr, and the return
3253 code is 1. However, the actual setting up of the sockets still happens in
3254 the disconnected process, so errors there are still only recorded on the
3257 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3258 connections (as happens on some IP stacks) was logged at start up time as
3259 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3260 it from "IPv6 and IPv4", which means that two separate sockets are being
3263 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3264 says whether AF_INET or AF_INET6 was passed as an argument.
3266 42. Exiwhat output was messed up when time zones were included in log
3269 43. Exiwhat now gives more information about the daemon's listening ports,
3270 and whether -tls-on-connect was used.
3272 44. The "port" option of the smtp transport is now expanded.
3274 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3275 silently ignored. Now an error message is written to the main and panic
3278 46. There's a new ACL modifier called "logwrite" which writes to a log file
3279 as soon as it is encountered.
3281 47. Added $local_user_uid and $local_user_gid at routing time.
3283 48. Exim crashed when trying to verify a sender address that was being
3286 49. Exim was recognizing only a space character after ".include". It now also
3287 recognizes a tab character.
3289 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3290 extracting the relevant information from the specification. The man page no
3291 longer contains scrambled data for the -d option, and I've added a section
3292 at the front about calling Exim under different names.
3294 51. Added "extra_headers" argument to the "mail" command in filter files.
3296 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3299 53. Installed eximstats 1.29.
3301 54. Added transport_filter_timeout as a generic transport option.
3303 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3304 Cc: header lines. This was required by RFC 822, but it not required by RFC
3307 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3308 incoming messages that did not have them. Now it does so only if the
3309 message originates locally, that is, if there is no associated remote host
3310 address. When Resent- header lines are present, this applies to the Resent-
3311 lines rather than the non-Resent- lines.
3313 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3314 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3316 58. Messages for configuration errors now include the name of the main
3317 configuration file - useful now that there may be more than one file in a
3318 list (.included file names were always shown).
3320 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3321 for those rare installations that do not start the daemon as root or run it
3322 setuid root. I've cut out the call to initgroups() if the daemon is not
3325 60. The Exim user and group can now be bound into the binary as text strings
3326 that are looked up at the start of Exim's processing.
3328 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3330 62. Added $mailstore_basename variable.
3332 63. Installed patch to sieve.c from Michael Haardt.
3334 64. When Exim failed to open the panic log after failing to open the main log,
3335 the original message it was trying to log was written to stderr and debug
3336 output, but if they were not available (the usual case in production), it
3337 was lost. Now it is written to syslog before the two lines that record the
3338 failures to open the logs.
3340 65. Users' Exim filters run in subprocesses under the user's uid. It is
3341 possible for a "deliver" command or an alias in a "personal" command to
3342 provoke an address rewrite. If logging of address rewriting is configured,
3343 this fails because the process is not running as root or exim. There may be
3344 a better way of dealing with this, but for the moment (because 4.30 needs
3345 to be released), I have disabled address rewrite logging when running a
3346 filter in a non-root, non-exim process.
3352 1. The buildconfig auxiliary program wasn't quoting the value set for
3353 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3354 not defined. This bug was masked in 4.22 by the effect that was fixed in
3357 2. Some messages that were rejected after a message id was allocated were
3358 shown as "incomplete" by exigrep. It no longer does this for messages that
3359 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3361 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3362 have allow_domain_literals set, the ID did not get logged in the <= line.
3363 Domain literals are now always recognized in Message-ID: header lines.
3365 4. The first argument for a ${extract expansion item is the key name or field
3366 number. Leading and trailing spaces in this item were not being ignored,
3367 causing some misleading effects.
3369 5. When deliver_drop_privilege was set, single queue runner processes started
3370 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3371 same command in the process it spins off) were not dropping privilege.
3373 6. When the daemon running as "exim" started a queue runner, it always
3374 re-executed Exim in the spun-off process. This is a waste of effort when
3375 deliver_drop_privilege is set. The new process now just calls the
3376 queue-runner function directly.
3382 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3385 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3386 ignored. Though the use of -oP was forcing the writing of a pid file, it
3387 was always written to the default place.
3389 3. If the message "no IP address found for host xxxx" is generated during
3390 incoming verification, it is now followed by identification of the incoming
3391 connection (so you can more easily find what provoked it).
3393 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3395 5. Added some features to "harden" Exim a bit more against certain attacks:
3397 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3398 be put in Local/Makefile. This is like the never_users runtime option,
3399 but it cannot be overridden. The default setting is "root".
3401 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3402 prefix string with which any file named in a -C command line option
3405 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3406 is retained for -C and -D only if the caller of Exim is root. Without
3407 it, the exim user may also use -C and -D and retain privilege.
3409 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3410 command line option is disabled.
3412 6. Macro names set by the -D option must start with an upper case letter, just
3413 like macro names defined in the configuration file.
3415 7. Added "dereference=" facility to LDAP.
3417 8. Two instances of the typo "uknown" in the source files are fixed.
3419 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3420 the Configure-Makefile script screwed up while processing it.
3422 10. Incorporated PCRE 4.4.
3424 11. The SMTP synchronization check was not operating right at the start of an
3425 SMTP session. For example, it could not catch a HELO sent before the client
3426 waited for the greeting. There is now a check for outstanding input at the
3427 point when the greeting is written. Because of the duplex, asynchronous
3428 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3429 way, but not yet received, when the check is performed.
3431 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3432 on TCP/IP sockets, because this apparently causes some broken clients to
3435 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3436 unchanged) from the Cygwin maintainer.
3438 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3439 was supported instead of testing for mbx. Effectively a typo.
3441 15. The spa authenticator server code was not checking that the input it
3442 received was valid base64.
3444 16. The debug output line for the "set" modifier in ACLs was not showing the
3445 name of the variable that was being set.
3447 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3449 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3450 lookup on its own. Something else was needed to trigger the lookup. For
3451 example, a match in host_lookup or the need for a host name in a host list.
3452 Now, if $sender_host_name is referenced and the host name has not yet been
3453 looked up, a lookup is performed. If the lookup fails, the variable remains
3454 empty, and $host_lookup_failed is set to "1".
3456 19. Added "eqi" as a case-independent comparison operator.
3458 20. The saslauthd authentication condition could segfault if neither service
3459 nor realm was specified.
3461 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3462 error message that was logged was misleading, and incoming SMTP
3463 connections were dropped. The message is now more accurate, and temporary
3464 errors are given to SMTP connections.
3466 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3467 (except RSET) with a 421 error, until QUIT is received. However, it was
3468 failing to send a response to QUIT.
3470 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3471 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3472 ACL is now run after the helo_try_verify_hosts code.
3474 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3475 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3476 case-independent, so other case variants are also recognized). Apparently
3477 some systems use these upper case variants.
3479 25. If more than two messages were waiting for the same host, and a transport
3480 filter was specified for the transport, Exim sent two messages over the
3481 same TCP/IP connection, and then failed with "socket operation on non-
3482 socket" when it tried to send the third.
3484 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3486 27. The extern definition of crypt16() in expand.c was not being excluded when
3487 the OS had its own crypt16() function.
3489 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3490 as a preferred synonym for return_size_limit, both as an option and as an
3493 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3495 30. Changed the default configuration ACL to relax the local part checking rule
3496 for addresses that are not in any local domains. For these addresses,
3497 slashes and pipe symbols are allowed within local parts, but the sequence
3498 /../ is explicitly forbidden.
3500 31. SPA server authentication was not clearing the challenge buffer before
3503 32. log_message in a "warn" ACL statement was writing to the reject log as
3504 well as to the main log, which contradicts the documentation and doesn't
3505 seem right (because no rejection is happening). So I have stopped it.
3507 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3508 However, I am unable to do any testing of this.
3510 34. Fixed an infelicity in the appendfile transport. When checking directories
3511 for a mailbox, to see if any needed to be created, it was accidentally
3512 using path names with one or more superfluous leading slashes; tracing
3513 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3515 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3516 contents are added to the log line that is written for every discarded
3517 recipient. (Previously a log_message setting was ignored.)
3519 36. The ${quote: operator now quotes the string if it is empty.
3521 37. The install script runs exim in order to find its version number. If for
3522 some reason other than non-existence or emptiness, which it checks, it
3523 could not run './exim', it was installing it with an empty version number,
3524 i.e. as "exim-". This error state is now caught, and the installation is
3527 38. An argument was missing from the function that creates an error message
3528 when Exim fails to connect to the socket for saslauthd authentication.
3529 This could cause Exim to crash, or give a corrupted message.
3531 39. Added isip, isip4, and isip6 to ${if conditions.
3533 40. The ACL variables $acl_xx are now saved with the message, and can be
3534 accessed later in routers, transports, and filters.
3536 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3537 strings in the file are not string-expanded.
3539 42. If a MAIL command specified a SIZE value that was too large to fit into an
3540 int variable, the check against message_size_limit failed. Such values are
3541 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3542 day this will have to be increased, but I don't think I want to be around
3543 when emails are that large.
3550 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3551 iconv() is not standard in FreeBSD.
3553 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3554 IPv6 enabled. The observed symptom was a segmentation fault on return from
3555 the function os_common_find_running_interfaces() in src/os.c.
3557 3. In the check_special_case() function in daemon.c I had used "errno" as an
3558 argument name, which causes warnings on some systems. This was basically a
3559 typo, since it was named "eno" in the comments!
3561 4. The code that waits for the clock to tick (at a resolution of some fraction
3562 of a second) so as to ensure message-id uniqueness was always waiting for
3563 at least one whole tick, when it could have waited for less. [This is
3564 almost certainly not relevant at current processor speeds, where it is
3565 unlikely to ever wait at all. But we try to future-proof.]
3567 5. The function that sleeps for a time interval that includes fractions of a
3568 second contained a race. It did not block SIGALRM between setting the
3569 timer, and suspending (a couple of lines later). If the interval was short
3570 and the sigsuspend() was delayed until after it had expired, the suspension
3571 never ended. On busy systems this could lead to processes getting stuck for
3574 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3575 process, before it forks any delivery processes. The open lookup caching
3576 mechanism meant that the open file or database connection was passed into
3577 the delivery process. The problem was that delivery processes always tidy
3578 up cached lookup data. This could cause a problem for the next delivery
3579 process started by the queue runner, because the external queue runner
3580 process does not know about the closure. So the next delivery process
3581 still has data in the lookup cache. In the case of a file lookup, there was
3582 no problem because closing a file descriptor in a subprocess doesn't affect
3583 the parent. However, if the lookup was caching a connection to a database,
3584 the connection was closed, and the second delivery process was likely to
3585 see errors such as "PGSQL: query failed: server closed the connection
3586 unexpectedly". The problem has been fixed by closing all cached lookups
3587 in a queue runner before running a delivery process.
3589 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3590 seem to have the "const" qualifier which it has on other OS. I've
3593 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3594 *of the same type* (client or server) with the same public name that an
3595 error should be diagnosed.
3597 9. When Exim looked up a host name for an IP address, but failed to find the
3598 original IP address when looking up the host name (a safety check), it
3599 output the message "<ip address> does not match any IP for NULL", which was
3600 confusing, to say the least. The bug was that the host name should have
3601 appeared instead of "NULL".
3603 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3604 user that is built into the binary, and the -C or -D options is used, root
3605 privilege is dropped before the configuration file is read. In addition,
3606 logging is switched to stderr instead of the normal log files. If the
3607 configuration then re-defines the Exim user, the unprivileged environment
3608 is probably not what is expected, so Exim logs a panic warning message (but
3611 However, if deliver_drop_privilege is set, the unprivileged state may well
3612 be exactly what is intended, so the warning has been cut out in that case,
3613 and Exim is allowed to try to write to its normal log files.
3619 1. smtp_return_error_details was not giving details for temporary sender
3620 or receiver verification errors.
3622 2. Diagnose a configuration error if two authenticators have the same public
3625 3. Exim used not to create the message log file for a message until the first
3626 delivery attempt. This could be confusing when incoming messages were held
3627 for policy or load reasons. The message log file is now created at the time
3628 the message is received, and an initial "Received" line is written to it.
3630 4. The automatically generated man page for command line options had a minor
3631 bug that caused no ill effects; however, a more serious problem was that
3632 the procedure for building the man page automatically didn't always
3633 operate. Consequently, release 4.20 contains an out-of-date version. This
3634 shouldn't happen again.
3636 5. When building Exim with embedded Perl support, the script that builds the
3637 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3638 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3640 6. The freeze_tell option was not being used for messages that were frozen on
3641 arrival, either by an ACL or by local_scan().
3643 7. Added the smtp_incomplete_transaction log selector.
3645 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3646 was accepting AUTH without a new EHLO.
3648 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3649 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3652 10. Logging of TCP/IP connections (when configured) now happens in the main
3653 daemon process instead of the child process, so that the TCP/IP connection
3654 count is more accurate (but it can never be perfect).
3656 11. The use of "drop" in a nested ACL was not being handled correctly in the
3657 outer ACL. Now, if condition failure induced by the nested "drop" causes
3658 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3659 or "require"), the connection is dropped.
3661 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3662 that yield "discard" can now be used with an "accept" or a "discard" verb,
3663 but an error is generated for any others (because I can't see a useful way
3664 to define what should happen).
3666 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3667 are now processed in the same way as lines in the Exim configuration file.
3668 In particular, continuation lines are supported.
3670 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3672 15. Added -ti meaning -t -i.
3674 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3675 domains, and warn by logging if others are found.
3677 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3678 returned from GIFCONF when Exim is trying to find the list of interfaces on
3679 a host. The code in os.c has been modified to copy each ifreq to an aligned
3680 structure in all cases.
3682 Also, in some cases, the returned ifreq's were being copied to a 'struct
3683 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3684 means the last couple of bytes of an IPv6 address could be chopped if the
3685 ifreq contained only a normal sockaddr (14 bytes storage).
3687 18. Named domain lists were not supported in the hosts_treat_as_local option.
3688 An entry such as +xxxx was not recognized, and was treated as a literal
3691 19. Ensure that header lines added by a DATA ACL are included in the reject log
3692 if the ACL subsequently rejects the message.
3694 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3695 MD5 (which is deprecated).
3697 21. When testing a filter file using -bf, Exim was writing a message when it
3698 took the sender from a "From " line in the message, but it was not doing so
3699 when it took $return_path from a Return-Path: header line. It now does.
3701 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3702 with a valid header line field name (a series of printing characters
3703 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3705 23. Changed "disc" in the source to "disk" to conform to the documentation and
3706 the book and for uniformity.
3708 24. Ignore Sendmail's -Ooption=value command line item.
3710 25. When execve() failed while trying to run a command in a pipe transport,
3711 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3712 could be confused with a return value of 69 from the command itself. This
3713 has been changed to 127, the value the shell returns if it is asked to run
3714 a non-existent command. The wording for the related log line suggests a
3715 non-existent command as the problem.
3717 26. If received_header_text expands to an empty string, do not add a Received:
3718 header line to the message. (Well, it adds a token one on the spool, but
3719 marks it "old" so that it doesn't get used or transmitted.)
3721 27. Installed eximstats 1.28 (addition of -nt option).
3723 28. There was no check for failure on the call to getsockname() in the daemon
3724 code. This can fail if there is a shortage of resources on the system, with
3725 ENOMEM, for example. A temporary error is now given on failure.
3727 29. Contrary to the C standard, it seems that in some environments, the
3728 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3729 program. Exim now does this explicitly; it affects the formatting of
3730 timestamps using strftime().
3732 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3733 complaints. I've now initialized all the variables, to avoid this.
3735 32. Header lines added by a system filter were not being "seen" during
3736 transport-time rewrites.
3738 33. The info_callback() function passed to OpenSSL is set up with type void
3739 (*)(SSL *, int, int), as described somewhere. However, when calling the
3740 function (actually a macro) that sets it up, the type void(*)() is
3741 expected. I've put in a cast to prevent warnings from picky compilers.
3743 34. If a DNS black list lookup found a CNAME record, but there were no A
3744 records associated with the domain it pointed at, Exim crashed.
3746 35. If a DNS black list lookup returned more than one A record, Exim ignored
3747 all but the first. It now scans all returned addresses if a particular IP
3748 value is being sought. In this situation, the contents of the
3749 $dnslist_value variable are a list of all the addresses, separated by a
3752 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3753 to accept a host name and all its aliases if the forward lookup for any of
3754 them yielded the IP address of the incoming connection. Now it accepts only
3755 those names whose forward lookup yields the correct IP address. Any other
3756 names are discarded. This closes a loophole whereby a rogue DNS
3757 administrator could create reverse DNS records to break through a
3758 wildcarded host restriction in an ACL.
3760 37. If a user filter or a system filter that ran in a subprocess used any of
3761 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3762 the wrong values were passed to the pipe command ($thisaddress had the
3763 value of $0, $0 had the value of $1, etc). This bug was introduced by
3764 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3766 38. Improved the line breaking for long SMTP error messages from ACLs.
3767 Previously, if there was no break point between 40 and 75 characters, Exim
3768 left the rest of the message alone. Two changes have been made: (a) I've
3769 reduced the minimum length to 35 characters; (b) if it can't find a break
3770 point between 35 and 75 characters, it looks ahead and uses the first one
3771 that it finds. This may give the occasional overlong line, but at least the
3772 remaining text gets split now.
3774 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3775 file descriptors might be low, and that setting 1000 would always raise it.
3776 It turns out that in some environments, the limit is already over 1000 and
3777 that lowering it causes trouble. So now Exim takes care not to decrease it.
3779 40. When delivering a message, the value of $return_path is set to $sender_
3780 address at the start of routing (routers may change the value). By an
3781 oversight, this default was not being set up when an address was tested by
3782 -bt or -bv, which affected the outcome if any router or filter referred to
3785 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3786 log only when "message" or "log_message" are set. However, if one of the
3787 conditions was an address verification, or a call to a nested ACL, the
3788 messages generated by the underlying test were being passed through. This
3789 no longer happens. The underlying message is available in $acl_verify_
3790 message for both "message" and "log_message" expansions, so it can be
3791 passed through if needed.
3793 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3794 new expansion $bh_ to give the encoded byte string without charset
3795 translation. Translation happens only if iconv() is available; HAVE_ICONV
3796 indicates this at build time. HEADERS_CHARSET gives the charset to
3797 translate to; headers_charset can change it in the configuration, and
3798 "headers charset" can change it in an individual filter file.
3800 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3801 that creates RFC 2047 encoded "words" labels them as that charset instead
3802 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3803 expansion operator; (ii) when Exim creates a From: line for a local
3804 message; (iii) when a header line is rewritten to include a "phrase" part.
3806 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3807 buggy, causing it to skip the first lines of messages whose message ID
3808 ended in 'D'. This would not have bitten before Exim release 4.14, because
3809 message IDs were unlikely to end in 'D' before then. The effect was to have
3810 incorrect size information for certain domains.
3812 45. #include "config.h" was missing at the start of the crypt16.c module. This
3813 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3816 46. If there was a timeout during a "random" callout check, Exim treated it as
3817 a failure of the random address, and carried on sending RSET and the real
3818 address. If the delay was just some slowness somewhere, the response to the
3819 original RCPT would be taken as a response to RSET and so on, causing
3820 mayhem of various kinds.
3822 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3823 when I implemented it. It didn't allow for the fact that some option values
3824 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3825 the right test for positive values.
3827 48. Domain names in DNS records are case-independent. Exim always looks them up
3828 in lower case. Some resolvers return domain names in exactly the case they
3829 appear in the zone file, that is, they may contain uppercase letters. Not
3830 all resolvers do this - some return always lower case. Exim was treating a
3831 change of case by a resolver as a change of domain, similar to a widening
3832 of a domain abbreviation. This triggered its re-routing code and so it was
3833 trying to route what was effectively the same domain again. This normally
3834 caused routing to fail (because the router wouldn't handle the domain
3835 twice). Now Exim checks for this case specially, and just changes the
3836 casing of the domain that it ultimately uses when it transmits the message
3839 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3842 50. If a filter generated a file delivery with a non-absolute name (possible if
3843 no home directory exists for the router), the forbid_file option was not
3846 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3847 the existing equality matching.
3849 52. Exim was using ints instead of ino_t variables in some places where it was
3850 dealing with inode numbers.
3852 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3853 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3854 called TMPDIR, and if it finds it is different, it changes its value.
3856 54. The smtp_printf() function is now made available to local_scan() so
3857 additional output lines can be written before returning. There is also an
3858 smtp_fflush() function to enable the detection of a dropped connection.
3859 The variables smtp_input and smtp_batched_input are exported to
3862 55. Changed the default runtime configuration: the message "Unknown user"
3863 has been removed from the ACL, and instead placed on the localuser router,
3864 using the cannot_route_message feature. This means that any verification
3865 failures that generate their own messages won't get overridden. Similarly,
3866 the "Unrouteable address" message that was in the ACL for unverifiable
3867 relay addresses has also been removed.
3869 56. Added hosts_avoid_esmtp to the smtp transport.
3871 57. The exicyclog script was not checking for the esoteric option
3872 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3873 will work only if exicyclog is run under the appropriate euid.
3875 58. Following a discussion on the list, the rules by which Exim recognises line
3876 endings on incoming messages have been changed. The -dropcr and drop_cr
3877 options are now no-ops, retained only for backwards compatibility. The
3878 following line terminators are recognized: LF CRLF CR. However, special
3879 processing applies to CR:
3881 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3882 nor a local message in the state where . is a terminator.
3884 (ii) If a bare CR is encountered in a header line, an extra space is added
3885 after the line terminator so as not to end the header. The reasoning
3886 behind this is that bare CRs in header lines are most likely either
3887 to be mistakes, or people trying to play silly games.
3889 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3890 was being incorrectly given as 18 bytes larger than it should have been.
3891 This is a VOB (very old bug).
3893 60. This may never have affected anything current, but just in case it has:
3894 When the local host is found other than at the start of a list of hosts,
3895 the local host, those with the same MX, and any that follow, are discarded.
3896 When the list in question was part of a longer list of hosts, the following
3897 hosts (not currently being processed) were also being discarded. This no
3898 longer happens. I'm not sure if this situation could ever has previously
3901 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3904 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3905 header. This is something that is recommended in a new Internet Draft, and
3906 is something that is documented as being done by Sendmail. There are two
3907 possible values. For messages generated by the autoreply transport, Exim
3910 Auto-Submitted: auto-replied
3912 whereas for all other generated messages (e.g. bounces) it adds
3914 Auto-Submitted: auto-generated
3916 63. The "personal" condition in filters now includes a test for the
3917 Auto-Submitted: header. If it contains the string "auto-" the message it
3918 not considered personal.
3920 64. Added rcpt_include_affixes as a generic transport option.
3922 65. Added queue_only_override (default true).
3924 66. Added the syslog_duplication option.
3926 67. If what should have been the first header line of a message consisted of
3927 a space followed by a colon, Exim was mis-interpreting it as a header line.
3928 It isn't of course - it is syntactically invalid and should therefore be
3929 treated as the start of the message body. The misbehaviour could have
3930 caused a number of strange effects, including loss of data in subsequent
3931 header lines, and spool format errors.
3933 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3934 client host had authenticated. This control can now be exercised by an ACL
3935 for more flexibility.
3937 69. By default, callouts do not happen when testing with -bh. There is now a
3938 variant, -bhc, which does actually run the callout code, including
3939 consulting and updating the callout cache.
3941 70. Added support for saslauthd authentication, courtesy of Alexander
3944 71. If statvfs() failed on the spool or log directories while checking their
3945 size for availability, Exim confusingly gave the error "space shortage".
3946 Furthermore, in debugging mode it crashed with a floating point exception.
3947 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3948 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3949 serious problem, Exim now writes to the main and panic logs when this
3950 happens, with details of the failure. It then refuses to accept the
3951 incoming message, giving the message "spool directory problem" or "log
3952 directory problem" with a 421 code for SMTP messages.
3954 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3955 0, 1, and 2 exist, because some OS complain for execs without them (see
3956 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3957 descriptors. However, the code omitted to check that the open succeeded,
3958 causing mysterious errors if for some reason the permissions on /dev/null
3959 got screwed. Now Exim writes a message to the main and panic logs, and
3960 bombs out if it can't open /dev/null.
3962 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3963 interact so that it is all more flexible. It is supposed to remain
3964 backwards compatible. Also added extra_local_interfaces.
3966 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3967 to bomb out with an assertion failure - to the client this appears as a
3968 connection drop. This problem occurs in the part of the code that was taken
3969 from the Samba project. Fortunately, the assertion is in a very simple
3970 function, so I have fixed this by reproducing the function inline in the
3971 one place where it is called, and arranging for authentication to fail
3972 instead of killing the process with assert().
3974 75. The SPA client code was not working when the server requested OEM rather
3975 than Unicode encoding.
3977 76. Added code to make require_files with a specific uid setting more usable in
3978 the case where statting the file as root fails - usually a non-root-mounted
3979 NFS file system. When this happens and the failure is EACCES, Exim now
3980 forks a subprocess and does the per-uid checking as the relevant uid.
3982 77. Added process_log_path.
3984 78. If log_file_path was not explicitly set, a setting of check_log_space or
3985 check_log_inodes was ignored.
3987 79. If a space check for the spool or log partitions fails, the incident is now
3988 logged. Of course, in the latter case the data may get lost...
3990 80. Added the %p formatting code to string_format() so that it can be used to
3991 print addresses in debug_print(). Adjusted all the address printing in the
3992 debugging in store.c to use %p rather than %d.
3994 81. There was a concern that a line of code in smtp_in.c could overflow a
3995 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3996 initially expressed, the concern was not well-founded, because trailing
3997 spaces are removed early. However, if the trailing spaces were followed by
3998 a NULL, they did not get removed, so the overflow was possible. Two fixes
4001 (a) I re-wrote the offending code in a cleaner fashion.
4002 (b) If an incoming SMTP command contains a NULL character, it is rejected
4005 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4006 runs initgroups(), so that if the Exim user is in any additional groups,
4007 they will be used during message reception.
4013 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob