1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.360 2006/06/27 15:07:18 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 SC/01 Use a glob alias rather than an array ref in eximstats generated
10 parser. This improves both readability and performance.
12 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
13 Don't display local sender or destination tables in eximstats unless
14 there is data to show.
15 Added average volumes into the eximstats top table text output.
17 SC/03 Collect data on the number of addresses (recipients) as well
18 as the number of messages in eximstats.
20 TF/01 Correct an error in the documentation for the redirect router. Exim
21 does (usually) call initgroups() when daemonizing.
23 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs 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.
38 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
39 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
40 without this. I don't know how relevant this is to other LDAP libraries.
42 PH/02 Add the verb name to the "unknown ACL verb" error.
44 PH/03 Magnus Holmgren's patch for filter_prepend_home.
46 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
52 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
53 other effects) broke the use of negated acl sub-conditions.
55 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
58 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
59 "Deny" causes Exim to reject the incoming connection with a 554 error.
60 Unfortunately, if there is a major crisis, such as a disk failure,
61 tcp-wrappers gives "deny", whereas what one would like would be some
62 kind of temporary error. A kludge has been added to help with this.
63 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
64 554 error is used if errno is still zero or contains ENOENT (which occurs
65 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
68 PH/03 Add -lutil to the default FreeBSD LIBS setting.
70 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
71 errors. Otherwise a message that provokes a temporary error (when other
72 messages do not) can cause a whole host to time out.
74 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
75 addresses were routed directly to files or pipes from a redirect router.
76 File deliveries just didn't batch; pipe deliveries might have suffered
79 PH/06 A failure to get a lock for a hints database would erroneously always say
80 "Failed to get write lock", even when it was really a read lock.
82 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
83 of 0600. This has been changed to use the value of the lockfile_mode
84 option (which defaults to 0600).
86 PH/08 Applied small patch from the Sieve maintainer.
88 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
89 folder from quota calculations, a direct delivery into this folder messed
90 up the contents of the maildirsize file. This was because the regex was
91 used only to exclude .Trash (or whatever) when the size of the mailbox
92 was calculated. There was no check that a delivery was happening into an
93 excluded directory. This bug has been fixed by ignoring all quota
94 processing for deliveries into excluded directories.
96 PH/10 Added the maildirfolder_create_regex option to appendfile.
102 PH/01 The code for finding all the local interface addresses on a FreeBSD
103 system running IPv6 was broken. This may well have applied to all BSD
104 systems, as well as to others that have similar system calls. The broken
105 code found IPv4 interfaces correctly, but gave incorrect values for the
106 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
107 that it would not match correctly against @[] and not recognize the IPv6
110 PH/02 The ipliteral router was not recognizing addresses of the form user@
111 [ipv6:....] because it didn't know about the "ipv6:" prefix.
113 PH/03 Added disable_ipv6.
115 PH/04 Changed $reply_address to use the raw form of the headers instead of the
116 decoded form, because it is most often used to construct To: headers
117 lines in autoreplies, and the decoded form may well be syntactically
118 invalid. However, $reply_address has leading white space removed, and all
119 newlines turned into spaces so that the autoreply transport does not
122 PH/05 If group was specified without a user on a router, and no group or user
123 was specified on a transport, the group from the router was ignored.
125 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
126 for visible compile-time settings that can be used to change these
127 numbers, for those that want even more. Backwards compatibility with old
128 spool files has been maintained. However, going back to a previous Exim
129 release will lost any variables that are in spool files.
131 PH/07 Two small changes when running in the test harness: increase delay when
132 passing a TCP/IP connection to a new process, in case the original
133 process has to generate a bounce, and remove special handling of
134 127.0.0.2 (sic), which is no longer necessary.
136 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
137 be the same on different OS.
139 PH/09 Moved a debug statement in filter processing to avoid a race problem when
142 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
143 whether --show-vars was specified or not
145 JJ/02 exipick: Added support for new ACL variable spool format introduced
148 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
149 syntactically invalid From: or Reply-to: line, and a filter used this to
150 generate an autoreply, and therefore failed to obtain an address for the
151 autoreply, Exim could try to deliver to a non-existent relative file
152 name, causing unrelated and misleading errors. What now happens is that
153 it logs this as a hard delivery error, but does not attempt to create a
156 PH/11 The exinext utility has a -C option for testing purposes, but although
157 the given file was scanned by exinext itself; it wasn't being passed on
158 when Exim was called.
160 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
161 an end-of-file indication when reading a command response.
163 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
164 compiled. In many other places in Exim, IPv6 addresses are always
165 recognized, so I have changed this. It also means that IPv4 domain
166 literals of the form [IPV4:n.n.n.n] are now always recognized.
168 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
169 used if the router is not running as root, for example, when verifying at
170 ACL time, or when using -bh. The debugging output from this situation was
171 non-existent - all you got was a failure to exec. I have made two
174 (a) Failures to set uid/gid, the current directory, or a process leader
175 in a subprocess such as that created by queryprogram now generate
176 suitable debugging ouput when -d is set.
178 (b) The queryprogram router detects when it is not running as root,
179 outputs suitable debugging information if -d is set, and then runs
180 the subprocess without attempting to change uid/gid.
182 PH/15 Minor change to Makefile for building test_host (undocumented testing
185 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
186 additional section of a DNS packet that returns MX or SRV records.
187 Instead, it always explicitly searches for A/AAAA records. This avoids
188 major problems that occur when a DNS server includes only records of one
189 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
190 fixed another bug: if SRV records were looked up and the corresponding
191 address records were *not* found in the additional section, the port
192 values from the SRV records were lost.
194 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
195 using the correct key (the original address) when searching the retry
196 rules in order to find which one to use for generating the retry hint.
198 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
199 adding the default one. Similarly, if it contains a Reply-To: header, the
200 errors_reply_to option, if set, is not used.
202 PH/19 When calculating a retry time, Exim used to measure the "time since
203 failure" by looking at the "first failed" field in the retry record. Now
204 it does not use this if it is later than than the arrival time of the
205 message. Instead it uses the arrival time. This makes for better
206 behaviour in cases where some deliveries succeed, thus re-setting the
207 "first failed" field. An example is a quota failure for a huge message
208 when small messages continue to be delivered. Without this change, the
209 "time since failure" will always be short, possible causing more frequent
210 delivery attempts for the huge message than are intended.
211 [Note: This change was subsequently modified - see PH/04 for 4.62.]
213 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
214 $1, $2, $3) because the numerical variables can be reset during some
215 expansion items (e.g. "match"), thereby losing the authentication data.
217 PH/21 Make -bV show the size of off_t variables so that the test suite can
218 decide whether to run tests for quotas > 2G.
220 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
221 mailbox_size, and mailbox_filecount in the appendfile transport. If a
222 filecount value is greater than 2G or if a quota value is greater than 2G
223 on a system where the size of off_t is not greater than 4, a panic error
226 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
227 never match. The debug and -bh output now contains an explicit error
228 message indicating a malformed IPv4 address or mask.
230 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
231 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
234 PH/25 Do not write to syslog when running in the test harness. The only
235 occasion when this arises is a failure to open the main or panic logs
236 (for which there is an explicit test).
238 PH/26 Added the /no_tell option to "control=freeze".
240 PH/27 If a host name lookup failed very early in a connection, for example, if
241 the IP address matched host_lookup and the reverse lookup yielded a name
242 that did not have a forward lookup, an error message of the form "no IP
243 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
244 could be logged. Now it outputs the IP address instead of "NULL".
246 PH/28 An enabling patch from MH: add new function child_open_exim2() which
247 allows the sender and the authenticated sender to be set when
248 submitting a message from within Exim. Since child_open_exim() is
249 documented for local_scan(), the new function should be too.
251 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
252 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
253 results in an empty string is now treated as unset.
255 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
257 PH/31 Added sender_verify_fail logging option.
259 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
260 needed by RFC 822 but not by RFC 2822 was commented out. I have now
261 tidied the source and removed it altogether.
263 PH/33 When a queue run was abandoned because the load average was too high, a
264 log line was always written; now it is written only if the queue_run log
265 selector is set. In addition, the log line for abandonment now contains
266 information about the queue run such as the pid. This is always present
267 in "start" and "stop" lines but was omitted from the "abandon" line.
269 PH/34 Omit spaces between a header name and the colon in the error message that
270 is given when verify = headers_syntax fails (if there are lots of them,
271 the message gets confusing).
273 PH/35 Change the default for dns_check_names_pattern to allow slashes within
274 names, as there are now some PTR records that contain slashes. This check
275 is only to protect against broken name servers that fall over on strange
276 characters, so the fact that it applies to all lookups doesn't matter.
278 PH/36 Now that the new test suite is complete, we can remove some of the
279 special code in Exim that was needed for the old test suite. For example,
280 sorting DNS records because real resolvers return them in an arbitrary
281 order. The new test suite's fake resolver always returns records in the
284 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
285 bounces) except when queue_only is set, to avoid logging races between
286 the different processes.
288 PH/38 Panic-die if .include specifies a non-absolute path.
290 PH/39 A tweak to the "H" retry rule from its user.
292 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
293 a label. They prevented compilation on older perls.
295 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
296 a warning to be raised on newish perls.
298 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
299 on queue. Changes to match documented behaviour of showing count of
300 messages matching specified criteria.
302 PH/40 Changed the default ident timeout from 30s to 5s.
304 PH/41 Added support for the use of login_cap features, on those BSD systems
305 that have them, for controlling the resources used by pipe deliveries.
307 PH/42 The content-scanning code uses fopen() to create files in which to put
308 message data. Previously it was not paying any attention to the mode of
309 the files. Exim runs with umask(0) because the rest of the code creates
310 files with open(), and sets the required mode explicitly. Thus, these
311 files were ending up world-writeable. This was not a big issue, because,
312 being within the spool directory, they were not world-accessible. I have
313 created a function called modefopen, which takes an additional mode
314 argument. It sets umask(777), creates the file, chmods it to the required
315 mode, then resets the umask. All the relevant calls to fopen() in the
316 content scanning code have been changed to use this function.
318 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
319 to 24 hours. This avoids potential overflow problems when processing G
320 and H retry rules. I suspect nobody ever tinkers with this value.
322 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
324 PH/45 When the plaintext authenticator is running as a client, the server's
325 challenges are checked to ensure they are valid base64 strings. By
326 default, the authentication attempt is cancelled if an invalid string is
327 received. Setting client_ignore_invalid_base64 true ignores these errors.
328 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
329 they are received. Thus, the responses can be made to depend on the
330 challenges. If an invalid string is ignored, an empty string is placed in
333 PH/46 Messages that are created by the autoreply transport now contains a
334 References: header, in accordance with RFCs 2822 and 3834.
336 PH/47 Added authenticated_sender_force to the smtp transport.
338 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
340 PH/49 Installed latest patch from the Sieve maintainer.
342 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
343 was also set, the appendfile transport was unnecessarily scanning a
344 directory of message files (e.g. for maildir delivery) to find the count
345 of files (along with the size), even though it did not need this
346 information. It now does the scan only if it needs to find either the
347 size of the count of files.
349 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
351 PH/52 Two bugs concerned with error handling when the smtp transport is
354 (i) Exim was not creating retry information for temporary errors given
355 for individual recipients after the DATA command when the smtp transport
356 was used in LMTP mode. This meant that they could be retried too
357 frequently, and not timed out correctly.
359 (ii) Exim was setting the flag that allows error details to be returned
360 for LMTP errors on RCPT commands, but not for LMTP errors for individual
361 recipients that were returned after the DATA command.
363 PH/53 This is related to PH/52, but is more general: for any failing address,
364 when detailed error information was permitted to be returned to the
365 sender, but the error was temporary, then after the final timeout, only
366 "retry timeout exceeded" was returned. Now it returns the full error as
367 well as "retry timeout exceeded".
369 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
370 do this, and (what is worse) MTAs that accept it.
372 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
373 will now be deprecated.
375 PH/56 New os.c-cygwin from the Cygwin maintainer.
377 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
378 formats (previously only available in exim formats via -bpr, -bpru,
379 and -bpra. Now also available in native and exiqgrep formats)
381 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
382 with very large, slow to parse queues
384 JJ/08 exipick: added ! as generic prefix to negate any criteria format
386 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
388 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
389 responses to authentication challenges, though it was showing the
390 challenges; (ii) I've removed the CR characters from the debug output for
393 PH/58 Allow for the insertion of a newline as well as a space when a string
394 is turned into more than one encoded-word during RFC 2047 encoding. The
395 Sieve code now uses this.
397 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
398 data_4xx, lost_connection, tls_required.
400 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
401 message was being sent as an SMTP response.
403 PH/61 Add -l and -k options to exicyclog.
405 PH/62 When verifying, if an address was redirected to one new address, so that
406 verification continued, and the new address failed or deferred after
407 having set something in $address_data, the value of $address_data was not
408 passed back to the ACL. This was different to the case when no
409 redirection occurred. The value is now passed back in both cases.
411 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
412 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
413 use login_cap.h, so on its own it isn't the distinguishing feature. The
414 new name refers directly to the setclassresources() function.
416 PH/65 Added configuration files for NetBSD3.
418 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
420 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
421 is preferred over IPv4.
423 PH/68 The bounce_return_message and bounce_return_body options were not being
424 honoured for bounces generated during the reception of non-SMTP messages.
425 In particular, this applied to messages rejected by the ACL. This bug has
426 been fixed. However, if bounce_return_message is true and bounce_return_
427 body is false, the headers that are returned for a non-SMTP message
428 include only those that have been read before the error was detected.
429 (In the case of an ACL rejection, they have all been read.)
431 PH/69 The HTML version of the specification is now built in a directory called
432 spec_html instead of spec.html, because the latter looks like a path with
433 a MIME-type, and this confuses some software.
435 PH/70 Catch two compiler warnings in sieve.c.
437 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
438 function verify_get_ident() calls ip_connect() to connect a socket, but
439 if the "connect()" function timed out, ip_connect() used to close the
440 socket. However, verify_get_ident() also closes the socket later, and in
441 between Exim writes to the log, which may get opened at this point. When
442 the socket was closed in ip_connect(), the log could get the same file
443 descriptor number as the socket. This naturally causes chaos. The fix is
444 not to close the socket in ip_connect(); the socket should be closed by
445 the function that creates it. There was only one place in the code where
446 this was missing, in the iplookup router, which I don't think anybody now
447 uses, but I've fixed it anyway.
449 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
450 well as to direct DNS lookups. Otherwise the handling of names in host
451 lists is inconsistent and therefore confusing.
457 PH/01 Two changes to the default runtime configuration:
459 (1) Move the checks for relay_from_hosts and authenticated clients from
460 after to before the (commented out) DNS black list checks.
462 (2) Add control=submission to the relay_from_hosts and authenticated
463 clients checks, on the grounds that messages accepted by these
464 statements are most likely to be submissions.
466 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
468 (1) Generate an error if the third argument for the ${prvs expansion is
471 (2) Treat a missing third argument of ${prvscheck as if it were an empty
474 (3) Reset the variables that are obtained from the first argument of
475 ${prvscheck and used in the second argument before leaving the code,
476 because their memory is reclaimed, so using them afterwards may do
479 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
480 one (it's much easier than Tom thought :-).
482 (5) Because of (4), we can now allow for the use of $prvscheck_result
483 inside the third argument.
485 PH/03 For some reason, the default setting of PATH when running a command from
486 a pipe transport was just "/usr/bin". I have changed it to
489 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
490 anything to be listed in the output from -bV.
492 PH/05 When a filter generated an autoreply, the entire To: header line was
493 quoted in the delivery log line, like this:
495 => >A.N.Other <ano@some.domain> <original@ddress> ...
497 This has been changed so that it extracts the operative address. There
498 may be more than one such address. If so, they are comma-separated, like
501 => >ano@some.domain,ona@other.domain <original@ddress> ...
503 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
504 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
505 not being looked up in the rDNS to get a host name, Exim was showing the
506 IP address twice in Received: lines, even though the IP addresses were
507 identical. For example:
509 Received: from [1.2.3.4] (helo=[1.2.3.4])
511 However, if the real host name was known, it was omitting the HELO data
512 if it matched the actual IP address. This has been tidied up so that it
513 doesn't show the same IP address twice.
515 PH/07 When both +timestamp and +memory debugging was on, the value given by
516 $tod_xxx expansions could be wrong, because the tod_stamp() function was
517 called by the debug printing, thereby overwriting the timestamp buffer.
518 Debugging no longer uses the tod_stamp() function when +timestamp is set.
520 PH/08 When the original message was included in an autoreply transport, it
521 always said "this is a copy of the message, including all the headers",
522 even if body_only or headers_only was set. It now gives an appropriate
525 PH/09 Applied a patch from the Sieve maintainer which:
527 o fixes some comments
528 o adds the (disabled) notify extension core
529 o adds some debug output for the result of if/elsif tests
530 o points to the current vacation draft in the documentation
531 and documents the missing references header update
535 o fixes a bug in processing the envelope test (when testing
536 multiple envelope elements, the last element determinted the
539 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
540 Electronic Mail") by including:
542 Auto-submitted: auto-generated
544 in the messages that it generates (bounce messages and others, such as
545 warnings). In the case of bounce messages for non-SMTP mesages, there was
546 also a typo: it was using "Auto_submitted" (underscore instead of
547 hyphen). Since every message generated by Exim is necessarily in response
548 to another message, thes have all been changed to:
550 Auto-Submitted: auto-replied
552 in accordance with these statements in the RFC:
554 The auto-replied keyword:
556 - SHOULD be used on messages sent in direct response to another
557 message by an automatic process,
559 - MUST NOT be used on manually-generated messages,
561 - MAY be used on Delivery Status Notifications (DSNs) and Message
562 Disposition Notifications (MDNs),
564 - MUST NOT be used on messages generated by automatic or periodic
565 processes, except for messages which are automatic responses to
568 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
569 to the default Received: header definition.
571 PH/12 Added log selector acl_warn_skipped (default on).
573 PH/13 After a successful wildlsearch lookup, discard the values of numeric
574 variables because (a) they are in the wrong storage pool and (b) even if
575 they were copied, it wouldn't work properly because of the caching.
577 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
578 checking when decoding. Apparently there are clients that generate
579 overlong encoded strings. Why am I not surprised?
581 PH/15 If the first argument of "${if match_address" was not empty, but did not
582 contain an "@" character, Exim crashed. Now it writes a panic log message
583 and treats the condition as false.
585 PH/16 In autoreply, treat an empty string for "once" the same as unset.
587 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
588 extension "envelope-auth". The code is finished and in agreement with
589 other implementations, but there is no documentation so far and in fact,
590 nobody wrote the draft yet. This extension is currently #undef'ed, thus
591 not changing the active code.
593 Print executed "if" and "elsif" statements when debugging is used. This
594 helps a great deal to understand what a filter does.
596 Document more things not specified clearly in RFC3028. I had all this
597 sorted out, when out of a sudden new issues came to my mind. Oops."
599 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
602 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
603 canonical form (as documented). However, after a host name lookup from
604 the IP address, check_host() was doing a simple string comparison with
605 addresses acquired from the DNS when checking that the found name did
606 have the original IP as one of its addresses. Since any found IPv6
607 addresses are likely to be in abbreviated form, the comparison could
608 fail. Luckily, there already exists a function for doing the comparison
609 by converting both addresses to binary, so now that is used instead of
612 PH/20 There was another similar case to PH/19, when a complete host name was
613 given in a host list; looking up its IP address could give an abbreviated
614 form, whereas the current host's name might or might not be abbreviated.
615 The same fix has been applied.
621 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
622 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
625 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
626 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
628 PH/03 Typo: missing ".o" in src/pcre/Makefile.
630 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
631 header line, restrict the check to what is listed in RFCs 2369 and 2929.
632 Also, for "Auto-Submitted", treat anything other than "no" as
633 non-personal, in accordance with RFC 3834. (Previously it treated
634 anything starting "auto-" as non-personal.)
636 TF/01 The control=submission/name=... option had a problem with syntax
637 errors if the name included a slash character. The /name= option
638 now slurps the rest of the string, so it can include any characters
639 but it must come last in the list of options (after /sender_retain
642 PH/05 Some modifications to the interface to the fake nameserver for the new
650 TK/01 Added the "success_on_redirect" address verification option. See
651 NewStuff for rationale and an example.
653 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
655 PH/02 Patch to exigrep to allow it to work on syslog lines.
657 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
658 fread() to skip over the body file's header line, because in Cygwin the
659 header line is locked and is inaccessible.
661 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
662 co-exist for some time) to make it clear that it is the Exim ID that is
663 referenced, not the Message-ID: header line.
665 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
666 string_format() function, because snprintf() does not exist on all
669 PH/06 The use of forbid_filter_existstest now also locks out the use of the
670 ${stat: expansion item.
672 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
673 protocol synchronization error", to keep the pedants happy.
675 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
676 well as for IRIX systems, when gcc is being used. See the host.c source
679 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
681 PH/10 Named domain lists were not working if used in a queue_smtp_domains
684 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
685 transport and to the smtp transport in LMTP mode.
687 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
689 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
690 run a filter in a subprocess. This could lead to confusion in subsequent
691 lookups in the parent process. There should also be a search_tidyup() at
692 the end of the subprocess.
694 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
695 only if the host matched helo_try_verify_hosts, which caused the
696 verification to occur when the EHLO/HELO command was issued. The ACL just
697 tested the remembered result. Now, if a previous verification attempt has
698 not happened, "verify = helo" does it there and then.
700 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
702 TK/03 Fix log output including CR from clamd.
704 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
705 exist provoked a memory error which could cause a segfault.
707 PH/15 Installed PCRE 6.2
709 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
711 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
712 of the problem. Specifically, suggested +O2 rather than +O1 for the
715 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
717 PH/20 If a delivery was routed to a non-standard port by means of an SRV
718 record, the port was not correctly logged when the outgoing_port log
719 selector was set (it logged the transort's default port).
721 PH/21 Added support for host-specific ports to manualroute, queryprogram,
722 fallback_hosts, and "hosts" in the smtp transport.
724 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
725 host errors such as "Connection refused".
727 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
728 authentication with radiusclient 0.4.9:
730 - Error returned from rc_read_config was caught wrongly
731 - Username/password not passed on to radius server due to wrong length.
733 The presumption is that some radiusclient API changes for 4.51/PH/17
734 were not taken care of correctly. The code is still untested by me (my
735 Linux distribution still has 0.3.2 of radiusclient), but it was
736 contributed by a Radius user.
738 PH/24 When doing a callout, the value of $domain wasn't set correctly when
739 expanding the "port" option of the smtp transport.
741 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
742 while reading a MIME header. Thanks to Tom Hughes for a patch.
744 PH/24 Include config.h inside local_scan.h so that configuration settings are
747 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
748 that in an ACL for RCPT (for example), you can examine exactly what was
751 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
752 commands, but it was not correctly comparing the address with the actual
753 client host address. Thus, it would show the EHLO address in Received:
754 header lines when this was not necessary.
756 PH/27 Added the % operator to ${eval:}.
758 PH/28 Exim tries to create and chdir to its spool directory when it starts;
759 it should be ignoring failures (because with -C, for example, it has lost
760 privilege). It wasn't ignoring creation failures other than "already
763 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
766 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
767 because an input line was too long, either on its own, or by virtue of
768 too many continuations, the temporary file was not being removed, and the
769 return code was incorrect.
771 PH/31 Missing "BOOL" in function definition in filtertest.c.
773 PH/32 Applied Sieve patches from the maintainer.
775 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
777 PH/33 Added "verify = not_blind".
779 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
780 Local/Makefile (with some defaults set). These are used in built scripts
781 such as exicyclog, but they have never been used in the exim_install
782 script (though there are many overriding facilities there). I have
783 arranged that the exim_install script now takes note of these two
786 PH/35 Installed configuration files for Dragonfly.
788 PH/36 When a locally submitted message by a trusted user did not contain a
789 From: header, and the sender address was obtained from -f or from an SMTP
790 MAIL command, and the trusted user did not use -F to supply a sender
791 name, $originator_name was incorrectly used when constructing a From:
792 header. Furthermore, $originator_name was used for submission mode
793 messages from external hosts without From: headers in a similar way,
794 which is clearly wrong.
796 PH/37 Added control=suppress_local_fixups.
798 PH/38 When log_selector = +received_sender was set, and the addition of the
799 sender made the log line's construction buffer exactly full, or one byte
800 less than full, an overflow happened when the terminating "\n" was
803 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
804 when the result of a list match is failure because a DNS lookup failed.
806 PH/40 RM_COMMAND is now used in the building process.
808 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
809 the "build-* directories that it finds.
811 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
812 address in a domain literal was a prefix of an interface address.
814 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
815 when verifying a sender address, unless rewrite_headers is false.
817 PH/44 Wrote a long comment about why errors_to addresses are verified as
818 recipients, not senders.
820 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
821 the ratelimit ACL was added.
823 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
825 PH/46 Added extra information about PostgreSQL errors to the error string.
827 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
828 avoiding the need to install special zones in a real server. This is
829 backwards compatible; if it can't find the fake resolver, it drops back.
830 Thus, both old and new test suites can be run.
832 TF/02 Added util/ratelimit.pl
834 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
835 clock is set back in time.
837 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
838 Candler <B.Candler@pobox.com>.
840 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
841 OK for addresses that are the result of redirections.
843 PH/48 A number of further additions for the benefit of the new test suite,
844 including a fake gethostbyname() that interfaces to the fake DNS resolver
847 TF/06 The fix for widen_domains has also been applied to qualify_single and
848 search_parents which are the other dnslookup options that can cause
851 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
854 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
856 TF/07 Exim produced the error message "an SRV record indicated no SMTP
857 service" if it encountered an MX record with an empty target hostname.
858 The message is now "an MX or SRV record indicated no SMTP service".
860 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
861 if the DNS of the sending site is misconfigured. This is quite a
862 common situation. This change restores the behaviour of treating a
863 helo verification defer as a failure.
865 PH/51 If self=fail was set on a router, the bounce message did not include the
866 actual error message.
872 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
874 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
875 command itself ended in error, the underlying message about the transport
876 filter timeout was being overwritten with the pipe command error. Now the
877 underlying error message should be appended to the second error message.
879 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
881 PH/02 Reduce the amount of output that "make" produces by default. Full output
882 can still be requested.
884 PH/03 The warning log line about a condition test deferring for a "warn" verb
885 was being output only once per connection, rather than after each
886 occurrence (because it was using the same function as for successful
887 "warn" verbs). This seems wrong, so I have changed it.
889 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
890 it should not have, which might have caused a crash in the right
891 circumstances, but probably never did.
893 PH/04 Installed a modified version of Tony Finch's patch to make submission
894 mode fix the return path as well as the Sender: header line, and to
895 add a /name= option so that you can make the user's friendly name appear
898 TF/03 Added the control = fakedefer ACL modifier.
900 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
901 Mark Lowes for thorough testing.
903 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
905 TK/03 Merged latest SRS patch from Miles Wilton.
907 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
908 with the definition in sysexits.h (which is #included earlier).
909 Fortunately, Exim does not actually use EX_OK. The code used to try to
910 preserve the sysexits.h value, by assumimg that macro definitions were
911 scanned for macro replacements. I have been disabused of this notion,
912 so now the code just undefines EX_OK before #including unistd.h.
914 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
915 in the smtp transport. When a block could not be written in a single
916 write() function, the timeout was being re-applied to each part-write.
917 This seems wrong - if the receiver was accepting one byte at a time it
918 would take for ever. The timeout is now adjusted when this happens. It
919 doesn't have to be particularly precise.
921 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
922 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
924 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
925 without a domain if the check to <postmaster@domain> fails.
927 SC/01 Eximstats: added -xls and the ability to specify output files
928 (patch written by Frank Heydlauf).
930 SC/02 Eximstats: use FileHandles for outputing results.
932 SC/03 Eximstats: allow any combination of xls, txt, and html output.
934 SC/04 Eximstats: fixed display of large numbers with -nvr option
936 SC/05 Eximstats: fixed merging of reports with empty tables.
938 SC/06 Eximstats: added the -include_original_destination flag
940 SC/07 Eximstats: removed tabs and trailing whitespace.
942 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
944 TK/06 MBOX spool code: Add real "From " MBOX separator line
945 so the .eml file is really in mbox format (even though
946 most programs do not really care). Patch from Alex Miller.
948 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
949 The latter is generated from $received_to and is only set if the
950 message has one envelope recipient. SA can use these headers,
951 obviously out-of-the-box. Patch from Alex Miller.
953 PH/08 The ${def test on a variable was returning false if the variable's
954 value was "0", contrary to what the specification has always said!
955 The result should be true unless the variable is empty.
957 PH/09 The syntax error of a character other than { following "${if
958 def:variable_name" (after optional whitespace) was not being diagnosed.
959 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
960 accidental colon was present, for example, could give incorrect results.
962 PH/10 Tidied the code in a number of places where the st_size field of a stat()
963 result is used (not including appendfile, where other changes are about
966 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
967 This involved changing a lot of size variables from int to off_t. It
968 should work with maildirs and everything.
970 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
971 spamd dying while we are connected to it.
973 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
976 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
977 function for BDB 4.3.
979 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
981 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
982 the sources for PCRE have been re-arranged and more files are now
985 PH/15 The code I had for printing potentially long long variables in PH/11
986 above was not the best (it lost precision). The length of off_t variables
987 is now inspected at build time, and an appropriate printing format (%ld
988 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
989 to be "long long int" or "long int". This is needed for the internal
990 formatting function string_vformat().
992 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
993 the configuration file to be ":syslog", then the script "guesses" where
994 the logs files are, rather than using the compiled in default. In our
995 case the guess is not the same as the compiled default, so the script
996 suddenly stopped working when I started to use syslog. The patch checks
997 to see if log_file_path is "". If so, it attempts to read it from exim
998 with no configuration file to get the compiled in version, before it
999 falls back to the previous guessing code."
1001 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1002 implementing BATV in an Exim configuration. See NewStuff for the gory
1005 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1006 Makefile that are specific to HP-UX.
1008 PH/18 If the "use_postmaster" option was set for a recipient callout together
1009 with the "random" option, the postmaster address was used as the MAIL
1010 FROM address for the random test, but not for the subsequent recipient
1011 test. It is now used for both.
1013 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1014 patch removes a few documentation additions to RFC 3028, because the
1015 latest draft now contains them. It adds the new en;ascii-case comparator
1016 and a new error check for 8bit text in MIME parts. Comparator and
1017 require names are now matched exactly. I enabled the subaddress
1018 extension, but it is not well tested yet (read: it works for me)."
1020 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1021 rework some of the code of TK/09 above to avoid the hardwired use of
1022 "%lld" and "long long". Replaced the call to snprintf() with a call to
1025 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1026 records point to non-existent hosts", "retry timeout exceeded", and
1027 "retry time not reached for any host after a long failure period".
1029 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1030 experimental DomainKeys support:
1032 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1033 (2) On an error such as an illegally used "control", the wrong name for
1034 the control was given.
1036 These problems did NOT occur unless DomainKeys support was compiled.
1038 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1040 PH/24 Added ${if match_ip condition.
1042 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1043 databases so that it will be absolutely obvious if a crash occurs in the
1044 DB library. This is a regular occurrence (often caused by mis-matched
1047 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1048 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1049 picked up on a user's system that detects such things. There doesn't seem
1050 to be a gcc warning option for this - only an attribute that has to be
1051 put on the function's prototype. It seems that in Fedora Core 4 they have
1052 set this on a number of new functions. No doubt there will be more in due
1055 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1056 specify a transport. However, if an address that was verified by such a
1057 router was the subject of a callout, Exim crashed because it tried to
1058 read the rcpt_include_affixes from the non-existent transport. Now it
1059 just assumes that the setting of that option is false. This bug was
1060 introduced by 4.51/PH/31.
1062 PH/28 Changed -d+all to exclude +memory, because that information is very
1063 rarely of interest, but it makes the output a lot bigger. People tend to
1064 do -d+all out of habit.
1066 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1067 code in os-type was giving problems when libc.so lives in lib64, like on
1070 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1071 aren't the modern standard, and it seems that some systems' include files
1072 don't always have them. Exim was already checking for some of the newer
1073 ones like T_AAAA, and defining it itself. I've added checks for all the
1074 record types that Exim uses.
1076 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1077 not automatically generating a new one, as it is supposed to. This
1078 prevented TLS from working. If the file did exist, but contained invalid
1079 data, a new version was generated, as expected. It was only the case of a
1080 non-existent file that was broken.
1082 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1083 with a change in libdomainkeys > 0.64.
1085 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1086 from DNS. If the selector record carries the flag, it now has
1087 precedence over the domain-wide flag.
1089 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1091 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1092 the use of a port name that isn't defined in /etc/services) occurred, the
1093 message was deferred as in a normal delivery, and thus remained on the
1094 spool, instead of being failed because of the mua_wrapper setting. This
1095 is now fixed, and I tidied up some of the mua_wrapper messages at the
1098 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1099 the messages in a hash of arrays rather than using individual hashes.
1100 This is a bit cleaner and results in dramatic memory savings, albeit
1101 at a slight CPU cost.
1103 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1104 as requested by Marc Sherman.
1106 SC/10 Eximstats: added histograms for user specified patterns as requested
1109 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1111 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1112 fopen() in the content-scanning modules that did not already have it.
1118 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1119 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1121 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1123 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1125 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1126 to test Sieve filters that use "vacation".
1128 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1129 that changes the way the GnuTLS parameters are stored in the cache file.
1130 The new format can be generated externally. For backward compatibility,
1131 if the data in the cache doesn't make sense, Exim assumes it has read an
1132 old-format file, and it generates new data and writes a new file. This
1133 means that you can't go back to an older release without removing the
1136 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1137 work if there are any delivery delays because "one_time" forces the
1138 parent to be marked "delivered", so its unseen clone is never tried
1139 again. For this reason, Exim now forbids the simultaneous setting of
1142 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1143 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1144 entry if you want to know the details. The fix, however, neglected to
1145 consider the case when local delivery batching is involved. The test for
1146 "previously delivered" was not happening when checking to see if an
1147 address could be batched with a previous (undelivered) one; under
1148 certain circumstances this could lead to multiple deliveries to the same
1151 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1152 in its include files, and this causes problems building Exim.
1154 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1155 header_syntax) but Exim was just ignoring anything given after a slash.
1156 In particular, this caused confusion with an attempt to use "verify =
1157 reverse_host_lookup/defer_ok". An error is now given when options are
1158 supplied for verify items that do not have them. (Maybe reverse_host_
1159 lookup should have a defer_ok option, but that's a different point.)
1161 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1162 defined by RFC 821) to 2048, because there were problems with some AUTH
1163 commands, and RFC 1869 says the size should be increased for extended
1164 SMTP commands that take arguments.
1166 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1169 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1170 "unknown" error; now it says that the functionality isn't in the binary.
1172 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1173 an address' error message when a string expansion fails (syntax or
1174 whatever). Otherwise the password may appear in the log. Following change
1175 PH/42 below, there is no longer a chance of it appearing in a bounce
1178 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1180 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1181 of its addresses was ever tried. (Bugzilla bug #2.)
1183 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1184 the result incorrectly in the debug output. (It correctly added a newline
1185 to what was transported.)
1187 TF/01 Added $received_time.
1189 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1190 commented out examples of how to interface to a virus scanner and to
1191 SpamAssassin. Also added commented examples of av_scanner and
1192 spamd_address settings.
1194 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1195 and controls are allowed in which ACLs. There were a couple of minor
1196 errors. Some of the entries in the conditions table (which is a table of
1197 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1198 negation of where the condition IS allowed.
1200 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1202 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1203 header file does not have a version number, so I've had to invent a new
1204 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1205 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1206 radiusclient), but it was contributed by a Radius user.
1208 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1209 files or directories, for OpenSSL.
1211 PH/19 When an Exim process that is running as root has to create an Exim log
1212 file, it does so in a subprocess that runs as exim:exim so as to get the
1213 ownership right at creation (otherwise, other Exim processes might see
1214 the file with the wrong ownership). There was no test for failure of this
1215 fork() call, which would lead to the process getting stuck as it waited
1216 for a non-existent subprocess. Forks do occasionally fail when resources
1217 run out. I reviewed all the other calls to fork(); they all seem to check
1220 PH/20 When checking for unexpected SMTP input at connect time (before writing
1221 the banner), Exim was not dealing correctly with a non-positive return
1222 from the read() function. If the client had disconnected by this time,
1223 the result was a log entry for a synchronization error with an empty
1224 string after "input=" when read() returned zero. If read() returned -1
1225 (an event I could not check), uninitialized data bytes were printed.
1226 There were reports of junk text (parts of files, etc) appearing after
1229 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1231 PH/22 Added support for macro redefinition, and (re)definition in between
1232 driver and ACL definitions.
1234 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1235 forgetting to use the resulting value; it was using the unexpanded value.
1237 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1238 hadn't been configured. The fix is from Juergen Kreileder, who
1239 understands it better than I do:
1241 "Here's what I see happening with three configured cyrus_sasl
1242 authenticators configured (plain, login, cram-md5):
1244 On startup auth_cyrus_sasl_init() gets called for each of these.
1245 This means three calls to sasl_listmech() without a specified mech_list.
1246 => SASL tests which mechs of all available mechs actually work
1247 => three warnings about OTP not working
1248 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1250 With the patch, sasl_listmech() also gets called three times. But now
1251 SASL's mech_list option is set to the server_mech specified in the the
1252 authenticator. Or in other words, the answer from sasl_listmech()
1253 gets limited to just the mech you're testing for (which is different
1255 => the return list contains just 'plain' or 'login', 'cram-md5' or
1256 nothing depending on the value of ob->server_mech.
1258 I've just tested the patch: Authentication still works fine,
1259 unavailable mechs specified in the exim configuration are still
1260 caught, and the auth.log warnings about OTP are gone."
1262 PH/25 When debugging is enabled, the contents of the command line are added
1263 to the debugging output, even when log_selector=+arguments is not
1266 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1267 answer is "GNU", and only if the return is "GNU/something" is the answer
1270 PH/27 $acl_verify_message is now set immediately after the failure of a
1271 verification in an ACL, and so is available in subsequent modifiers. In
1272 particular, the message can be preserved by coding like this:
1274 warn !verify = sender
1275 set acl_m0 = $acl_verify_message
1277 Previously, $acl_verify_message was set only while expanding "message"
1278 and "log_message" when a very denied access.
1280 PH/28 Modified OS/os.c-Linux with
1282 -#ifndef OS_LOAD_AVERAGE
1283 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1285 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1286 nomenclature these days.)
1288 PH/29 Installed patch from the Sieve maintainer that adds the options
1289 sieve_useraddress and sieve_subaddress to the redirect router.
1291 PH/30 In these circumstances:
1292 . Two addresses routed to the same list of hosts;
1293 . First host does not offer TLS;
1294 . First host accepts first address;
1295 . First host gives temporary error to second address;
1296 . Second host offers TLS and a TLS session is established;
1297 . Second host accepts second address.
1298 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1299 and peerdn, if requested) that were in fact used only for the second
1302 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1303 attention to any local part prefix or suffix that was matched by the
1304 router that accepted the address. It now behaves in the same way as it
1305 does for delivery: the affixes are removed from the local part unless
1306 rcpt_include_affixes is set on the transport.
1308 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1309 timeout during the DATA phase of an incoming message.
1311 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1312 applied a patch sanctioned by the Sieve maintainer.
1314 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1315 the uid or gid is negative. A case of a negative gid caused this to be
1316 noticed. The fix allows for either to be negative.
1318 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1319 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1322 PH/36 The change PH/12 above was broken. Fixed it.
1324 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1325 the grounds that routing the same address twice would always produce the
1326 same answer. This might have been true once, but it is certainly no
1327 longer true now. Routing a child address may depend on the previous
1328 routing that produced that child. Some complicated redirection strategies
1329 went wrong when messages had multiple recipients, and made Exim's
1330 behaviour dependent on the order in which the addresses were given.
1332 I have moved the duplicate checking until after the routing is complete.
1333 Exim scans the addresses that are assigned to local and remote
1334 transports, and removes any duplicates. This means that more work will be
1335 done, as duplicates will always all be routed, but duplicates are
1336 presumably rare, so I don't expect this is of any significance.
1338 For deliveries to pipes, files, and autoreplies, the duplicate checking
1339 still happens during the routing process, since they are not going to be
1342 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1343 It corrects a timeout issue with spamd. This is Ian's comment: "The
1344 background is that sometimes spamd either never reads data from a
1345 connection it has accepted, or it never writes response data. The exiscan
1346 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1347 blindly assumes that writes won't block so it may never time out."
1349 PH/39 Allow G after quota size as well as K and M.
1351 PH/40 The value set for $authenticated_id in an authenticator may not contain
1352 binary zeroes or newlines because the value is written to log lines and
1353 to spool files. There was no check on this. Now the value is run through
1354 the string_printing() function so that such characters are converted to
1355 printable escape sequences.
1357 PH/41 $message_linecount is a new variable that contains the total number of
1358 lines in the message. Compare $body_linecount, which is the count for the
1361 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1362 bounce and delay warning messages, except in certain special cases, which
1365 (a) An SMTP error message from a remote host;
1366 (b) A message specified in a :fail: redirection;
1367 (c) A message specified in a "fail" command in a system filter;
1368 (d) A message specified in a FAIL return from the queryprogram router;
1369 (e) A message specified by the cannot_route_message router option.
1371 In these cases only, Exim does include the error details in bounce and
1372 warning messages. There are also a few cases where bland messages such
1373 as "unrouteable address" or "local delivery error" are given.
1375 PH/43 $value is now also set for the "else" part of a ${run expansion.
1377 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1378 being worked on, but at least Exim now implements the latest version to
1381 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1382 process to complete was treated as a delivery failure, a timeout while
1383 writing the message to the pipe was logged, but erroneously treated as a
1384 successful delivery. Such timeouts include transport filter timeouts. For
1385 consistency with the overall process timeout, these timeouts are now
1386 treated as errors, giving rise to delivery failures by default. However,
1387 there is now a new Boolean option for the pipe transport called
1388 timeout_defer, which, if set TRUE, converts the failures into defers for
1389 both kinds of timeout. A transport filter timeout is now identified in
1392 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1393 systems where "make" and "gmake" are different, calling "gmake" at top
1394 level broke things. I've arranged for the value of $(MAKE) to be passed
1395 from the Makefile to this script so that it can call the same version of
1399 A note about Exim versions 4.44 and 4.50
1400 ----------------------------------------
1402 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1403 changes of various kinds. As a consequence, a big documentation update was
1404 needed. This delayed the release for rather longer than seemed good, especially
1405 in the light of a couple of (minor) security issues. Therefore, the changes
1406 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1407 release. So 4.44 and 4.50 are in effect two different branches that both start
1410 I have left the 4.50 change log unchanged; it contains all the changes since
1411 4.43. The change log for 4.44 is below; many of its items are identical to
1412 those for 4.50. This seems to be the most sensible way to preserve the
1413 historical information.
1419 1. Minor wording change to the doc/README.SIEVE file.
1421 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1422 computation of the current number of files was incorrect.
1424 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1425 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1428 4. Give more explanation in the error message when the command for a transport
1429 filter fails to execute.
1431 5. There are several places where Exim runs a non-Exim command in a
1432 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1433 was being done only for the command run by the queryprogram router. It is
1434 now done for all such subprocesses. The other cases are: ${run, transport
1435 filters, and the commands run by the lmtp and pipe transports.
1437 6. Added CONFIGURE_GROUP build-time option.
1439 7. Some older OS have a limit of 256 on the maximum number of file
1440 descriptors. Exim was using setrlimit() to set 1000 as a large value
1441 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1442 systems. I've change it so that if it can't get 1000, it tries for 256.
1444 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1445 was an oversight, and furthermore, ever since the addition of extra
1446 controls (e.g. 4.43/32), the checks on when to allow different forms of
1447 "control" were broken. There should now be diagnostics for all cases when a
1448 control that does not make sense is encountered.
1450 9. Added the /retain_sender option to "control=submission".
1452 10. $recipients is now available in the predata ACL (oversight).
1454 11. Tidy the search cache before the fork to do a delivery from a message
1455 received from the command line. Otherwise the child will trigger a lookup
1456 failure and thereby defer the delivery if it tries to use (for example) a
1457 cached ldap connection that the parent has called unbind on.
1459 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1460 of $address_data from the recipient verification was clobbered by the
1461 sender verification.
1463 13. The value of address_data from a sender verification is now available in
1464 $sender_address_data in subsequent conditions in the ACL statement.
1466 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1468 15. Added a new option "connect=<time>" to callout options, to set a different
1471 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1472 was its contents. (It was OK if the option was not defined at all.)
1474 17. A "Completed" log line is now written for messages that are removed from
1475 the spool by the -Mrm option.
1477 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1478 information about exactly what failed.
1480 19. Added -dd to debug only the daemon process.
1482 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1483 handles timeouts, both on the server side and network timeouts. Renamed the
1484 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1486 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1487 It is now set to "smtps".
1489 22. $host_address is now set to the target address during the checking of
1490 ignore_target_hosts.
1492 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1493 being passed; this would have caused $sender_host_name to have been used if
1494 matching the list had actually called for a host name (not very likely,
1495 since this list is usually IP addresses). A host name is now passed as
1498 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1499 code that specifies a non-restarting handler (typically sigaction() in
1500 modern systems) in an attempt to fix a rare and obscure crash bug.
1502 25. Narrowed the window for a race in the daemon that could cause it to ignore
1503 SIGCHLD signals. This is not a major problem, because they are used only to
1504 wake it up if nothing else does.
1506 26. A malformed maildirsize file could cause Exim to calculate negative values
1507 for the mailbox size or file count. Odd effects could occur as a result.
1508 The maildirsize information is now recalculated if the size or filecount
1511 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1512 support for a long time. Removed HAVE_SYS_VFS_H.
1514 28. Installed the latest version of exipick from John Jetmore.
1516 29. In an address list, if the pattern was not a regular expression, an empty
1517 subject address (from a bounce message) matched only if the pattern was an
1518 empty string. Non-empty patterns were not even tested. This was the wrong
1519 because it is perfectly reasonable to use an empty address as part of a
1520 database query. An empty address is now tested by patterns that are
1521 lookups. However, all the other forms of pattern expect the subject to
1522 contain a local part and a domain, and therefore, for them, an empty
1523 address still always fails if the pattern is not itself empty.
1525 30. Exim went into a mad DNS loop when attempting to do a callout where the
1526 host was specified on an smtp transport, and looking it up yielded more
1527 than one IP address.
1529 31. Re-factored the code for checking spool and log partition space into a
1530 function that finds that data and another that does the check. The former
1531 is then used to implement four new variables: $spool_space, $log_space,
1532 $spool_inodes, and $log_inodes.
1534 32. The RFC2047 encoding function was originally intended for short strings
1535 such as real names; it was not keeping to the 75-character limit for
1536 encoded words that the RFC imposes. It now respects the limit, and
1537 generates multiple encoded words if necessary. To be on the safe side, I
1538 have increased the buffer size for the ${rfc2047: expansion operator from
1541 33. It is now permitted to omit both strings after an "if" condition; if the
1542 condition is true, the result is "true". As before, when the second string
1543 is omitted, a false condition yields an empty string. This makes it less
1544 cumbersome to write custom ACL and router conditions.
1546 34. Failure to deliver a bounce message always caused it to be frozen, even if
1547 there was an errors_to setting on the router. The errors_to setting is now
1550 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1551 canonical form (fully expanded) before being placed in
1552 $sender_host_address.
1554 36. The table in the code that translates DNS record types into text (T_A to
1555 "A" for instance) was missing entries for NS and CNAME. It is just possible
1556 that this could have caused confusion if both these types were looked up
1557 for the same domain, because the text type is used as part of Exim's
1558 per-process caching. But the chance of anyone hitting this buglet seems
1561 37. The dnsdb lookup has been extended in a number of ways.
1563 (1) There is a new type, "zns", which walks up the domain tree until it
1564 finds some nameserver records. It should be used with care.
1566 (2) There is a new type, "mxh", which is like "mx" except that it returns
1567 just the host names, not the priorities.
1569 (3) It is now possible to give a list of domains (or IP addresses) to be
1570 looked up. The behaviour when one of the lookups defers can be
1571 controlled by a keyword.
1573 (4) It is now possible to specify the separator character for use when
1574 multiple records are returned.
1576 38. The dnslists ACL condition has been extended: it is now possible to supply
1577 a list of IP addresses and/or domains to be looked up in a particular DNS
1580 39. Added log_selector=+queue_time_overall.
1582 40. When running the queue in the test harness, wait just a tad after forking a
1583 delivery process, to get repeatability of debugging output.
1585 41. Include certificate and key file names in error message when GnuTLS fails
1586 to set them up, because the GnuTLS error message doesn't include the name
1587 of the failing file when there is a problem reading it.
1589 42. Allow both -bf and -bF in the same test run.
1591 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1593 44. The "Exiscan patch" is now merged into the mainline Exim source.
1595 45. Sometimes the final signoff response after QUIT could fail to get
1596 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1597 < 0 before doing a fflush(). This bug looks as though it goes back to the
1598 introduction of TLS in release 3.20, but "sometimes" must have been rare
1599 because the tests only now provoked it.
1601 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1602 (this can affect the format of dates).
1604 47. exim_tidydb, when checking for the continued existence of a message for
1605 which it has found a message-specific retry record, was not finding
1606 messages that were in split spool directories. Consequently, it was
1607 deleting retry records that should have stayed in existence.
1609 48. Steve fixed some bugs in eximstats.
1611 49. The SPA authentication driver was not abandoning authentication and moving
1612 on to the next authenticator when an expansion was forced to fail,
1613 contradicting the general specification for all authenticators. Instead it
1614 was generating a temporary error. It now behaves as specified.
1616 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1617 (the order specifies the preference for clients). The order is now AES256,
1618 AES128, 3DES, ARCFOUR128.
1620 51. Small patch to Sieve code - explicitly set From: when generating an
1623 52. Exim crashed if a remote delivery caused a very long error message to be
1624 recorded - for instance if somebody sent an entire SpamAssassin report back
1625 as a large number of 550 error lines. This bug was coincidentally fixed by
1626 increasing the size of one of Exim's internal buffers (big_buffer) that
1627 happened as part of the Exiscan merge. However, to be on the safe side, I
1628 have made the code more robust (and fixed the comments that describe what
1631 53. Now that there can be additional text after "Completed" in log lines (if
1632 the queue_time_overall log selector is set), a one-byte patch to exigrep
1633 was needed to allow it to recognize "Completed" as not the last thing in
1636 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1637 patch that reportedly fixes this has been added. I am not expert enough to
1638 create a test for it. This is what the patch creator wrote:
1640 "I found a little strange behaviour of ldap code when working with
1641 Windows 2003 AD Domain, where users was placed in more than one
1642 Organization Units. When I tried to give exim partial DN, the exit code
1643 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1644 But simultaneously result of request was absolutely normal ldap result,
1645 so I produce this patch..."
1647 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1648 REFERENCE, so I have modified the code to exclude the patch when that macro
1651 55. Some experimental protocols are using DNS PTR records for new purposes. The
1652 keys for these records are domain names, not reversed IP addresses. The
1653 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1654 leaves it alone. Component reversal etc. now happens only for IP addresses.
1657 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1659 57. Double the size of the debug message buffer (to 2048) so that more of very
1660 long debug lines gets shown.
1662 58. The exicyclog utility now does better if the number of log files to keep
1663 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1665 59. Two changes related to the smtp_active_hostname option:
1667 (1) $smtp_active_hostname is now available as a variable.
1668 (2) The default for smtp_banner uses $smtp_active_hostname instead
1669 of $primary_hostname.
1671 60. The host_aton() function is supposed to be passed a string that is known
1672 to be a valid IP address. However, in the case of IPv6 addresses, it was
1673 not checking this. This is a hostage to fortune. Exim now panics and dies
1674 if the condition is not met. A case was found where this could be provoked
1675 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1676 components; fortuitously, this particular loophole had already been fixed
1677 by change 4.50/55 above.
1679 If there are any other similar loopholes, the new check in host_aton()
1680 itself should stop them being exploited. The report I received stated that
1681 data on the command line could provoke the exploit when Exim was running as
1682 exim, but did not say which command line option was involved. All I could
1683 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1684 running as the user.
1687 61. There was a buffer overflow vulnerability in the SPA authentication code
1688 (which came originally from the Samba project). I have added a test to the
1689 spa_base64_to_bits() function which I hope fixes it.
1692 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1693 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1695 63. The daemon start-up calls getloadavg() while still root for those OS that
1696 need the first call to be done as root, but it missed one case: when
1697 deliver_queue_load_max is set with deliver_drop_privilege. This is
1698 necessary for the benefit of the queue runner, because there is no re-exec
1699 when deliver_drop_privilege is set.
1701 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1702 This has been fixed.
1704 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1705 was in use, was not putting the data itself into the right store pool;
1706 consequently, it could be overwritten for a subsequent message in the same
1707 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1710 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1712 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1713 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1714 the function were treating the return as a boolean value, which happened to
1715 work because 0=false and not-0=true, but is not correct code.
1717 68. The host_aton() function was not handling scoped IPv6 addresses (those
1718 with, for example, "%eth0" on the end) correctly.
1720 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1721 negated items (that is, ~something) in unsigned ints. Some compilers
1722 apparently mutter when there is no cast.
1724 70. If an address verification called from an ACL failed, and did not produce a
1725 user-specific message (i.e. there was only a "system" message), nothing was
1726 put in $acl_verify_message. In this situation, it now puts the system
1729 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1730 session; change 4.31/43 added the unwanted input to the log line - except
1731 that it did not do this in the start of session case. It now does.
1733 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1734 This is wrong and can cause the other end to generate a synchronization
1735 error if it is another Exim or anything else that does the synchronization
1736 check. A QUIT command is no longer sent after a timeout.
1738 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1739 during host lookups.
1741 74. The defer_ok option of callout verification was not working if it was used
1742 when verifying addresses in header lines, that is, for this case:
1744 verify = header_sender/callout=defer_ok
1746 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1747 those file descriptors could be used for SMTP connections. If anything
1748 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1749 could be sent to the SMTP client, causing chaos. The daemon now opens
1750 stdin, stdout, and stderr to /dev/null when it puts itself into the
1753 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1754 log by default. The user can override this with suitable Perl magic.
1756 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1757 the log message when discard triggers, was not working for the DATA ACL or
1758 for the non-SMTP ACL.
1760 78. Error message wording change in sieve.c.
1762 79. If smtp_accept_max_per_host was set, the number of connections could be
1763 restricted to fewer than expected, because the daemon was trying to set up
1764 a new connection before checking whether the processes handling previous
1765 connections had finished. The check for completed processes is now done
1766 earlier. On busy systems, this bug wouldn't be noticed because something
1767 else would have woken the daemon, and it would have reaped the completed
1770 80. If a message was submitted locally by a user whose login name contained one
1771 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1772 It caused a spool format error. I have fixed the spool reading code. A
1773 related problem was that the "from" clause in the Received: line became
1774 illegal because of the space(s). It is now covered by ${quote_local_part.
1776 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1779 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1781 83. Patch from Sieve maintainer to fix unterminated string problem in
1782 "vacation" handling.
1784 84. Some minor changes to the Linux configuration files to help with other
1785 OS variants using glibc.
1787 85. One more patch for Sieve to update vacation handling to latest spec.
1790 ----------------------------------------------------
1791 See the note above about the 4.44 and 4.50 releases.
1792 ----------------------------------------------------
1798 1. Change 4.43/35 introduced a bug that caused file counts to be
1799 incorrectly computed when quota_filecount was set in an appendfile
1802 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1803 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1806 3. Give more explanation in the error message when the command for a transport
1807 filter fails to execute.
1809 4. There are several places where Exim runs a non-Exim command in a
1810 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1811 was being done only for the command run by the queryprogram router. It is
1812 now done for all such subprocesses. The other cases are: ${run, transport
1813 filters, and the commands run by the lmtp and pipe transports.
1815 5. Some older OS have a limit of 256 on the maximum number of file
1816 descriptors. Exim was using setrlimit() to set 1000 as a large value
1817 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1818 systems. I've change it so that if it can't get 1000, it tries for 256.
1820 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1821 was an oversight, and furthermore, ever since the addition of extra
1822 controls (e.g. 4.43/32), the checks on when to allow different forms of
1823 "control" were broken. There should now be diagnostics for all cases when a
1824 control that does not make sense is encountered.
1826 7. $recipients is now available in the predata ACL (oversight).
1828 8. Tidy the search cache before the fork to do a delivery from a message
1829 received from the command line. Otherwise the child will trigger a lookup
1830 failure and thereby defer the delivery if it tries to use (for example) a
1831 cached ldap connection that the parent has called unbind on.
1833 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1834 of $address_data from the recipient verification was clobbered by the
1835 sender verification.
1837 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1838 was its contents. (It was OK if the option was not defined at all.)
1840 11. A "Completed" log line is now written for messages that are removed from
1841 the spool by the -Mrm option.
1843 12. $host_address is now set to the target address during the checking of
1844 ignore_target_hosts.
1846 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1847 being passed; this would have caused $sender_host_name to have been used if
1848 matching the list had actually called for a host name (not very likely,
1849 since this list is usually IP addresses). A host name is now passed as
1852 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1853 code that specifies a non-restarting handler (typically sigaction() in
1854 modern systems) in an attempt to fix a rare and obscure crash bug.
1856 15. Narrowed the window for a race in the daemon that could cause it to ignore
1857 SIGCHLD signals. This is not a major problem, because they are used only to
1858 wake it up if nothing else does.
1860 16. A malformed maildirsize file could cause Exim to calculate negative values
1861 for the mailbox size or file count. Odd effects could occur as a result.
1862 The maildirsize information is now recalculated if the size or filecount
1865 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1866 support for a long time. Removed HAVE_SYS_VFS_H.
1868 18. Updated exipick to current release from John Jetmore.
1870 19. Allow an empty sender to be matched against a lookup in an address list.
1871 Previously the only cases considered were a regular expression, or an
1874 20. Exim went into a mad DNS lookup loop when doing a callout where the
1875 host was specified on the transport, if the DNS lookup yielded more than
1878 21. The RFC2047 encoding function was originally intended for short strings
1879 such as real names; it was not keeping to the 75-character limit for
1880 encoded words that the RFC imposes. It now respects the limit, and
1881 generates multiple encoded words if necessary. To be on the safe side, I
1882 have increased the buffer size for the ${rfc2047: expansion operator from
1885 22. Failure to deliver a bounce message always caused it to be frozen, even if
1886 there was an errors_to setting on the router. The errors_to setting is now
1889 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1890 canonical form (fully expanded) before being placed in
1891 $sender_host_address.
1893 24. Updated eximstats to version 1.33
1895 25. Include certificate and key file names in error message when GnuTLS fails
1896 to set them up, because the GnuTLS error message doesn't include the name
1897 of the failing file when there is a problem reading it.
1899 26. Expand error message when OpenSSL has problems setting up cert/key files.
1902 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1903 (this can affect the format of dates).
1905 28. exim_tidydb, when checking for the continued existence of a message for
1906 which it has found a message-specific retry record, was not finding
1907 messages that were in split spool directories. Consequently, it was
1908 deleting retry records that should have stayed in existence.
1910 29. eximstats updated to version 1.35
1911 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1912 1.35 - bugfix such that pie charts by volume are generated correctly
1914 30. The SPA authentication driver was not abandoning authentication and moving
1915 on to the next authenticator when an expansion was forced to fail,
1916 contradicting the general specification for all authenticators. Instead it
1917 was generating a temporary error. It now behaves as specified.
1919 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1920 (the order specifies the preference for clients). The order is now AES256,
1921 AES128, 3DES, ARCFOUR128.
1923 31. Small patch to Sieve code - explicitly set From: when generating an
1926 32. Exim crashed if a remote delivery caused a very long error message to be
1927 recorded - for instance if somebody sent an entire SpamAssassin report back
1928 as a large number of 550 error lines. This bug was coincidentally fixed by
1929 increasing the size of one of Exim's internal buffers (big_buffer) that
1930 happened as part of the Exiscan merge. However, to be on the safe side, I
1931 have made the code more robust (and fixed the comments that describe what
1934 33. Some experimental protocols are using DNS PTR records for new purposes. The
1935 keys for these records are domain names, not reversed IP addresses. The
1936 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1937 leaves it alone. Component reversal etc. now happens only for IP addresses.
1940 34. The host_aton() function is supposed to be passed a string that is known
1941 to be a valid IP address. However, in the case of IPv6 addresses, it was
1942 not checking this. This is a hostage to fortune. Exim now panics and dies
1943 if the condition is not met. A case was found where this could be provoked
1944 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1945 components; fortuitously, this particular loophole had already been fixed
1946 by change 4.50/55 or 4.44/33 above.
1948 If there are any other similar loopholes, the new check in host_aton()
1949 itself should stop them being exploited. The report I received stated that
1950 data on the command line could provoke the exploit when Exim was running as
1951 exim, but did not say which command line option was involved. All I could
1952 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1953 running as the user.
1956 35. There was a buffer overflow vulnerability in the SPA authentication code
1957 (which came originally from the Samba project). I have added a test to the
1958 spa_base64_to_bits() function which I hope fixes it.
1961 36. The daemon start-up calls getloadavg() while still root for those OS that
1962 need the first call to be done as root, but it missed one case: when
1963 deliver_queue_load_max is set with deliver_drop_privilege. This is
1964 necessary for the benefit of the queue runner, because there is no re-exec
1965 when deliver_drop_privilege is set.
1967 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1968 was in use, was not putting the data itself into the right store pool;
1969 consequently, it could be overwritten for a subsequent message in the same
1970 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1973 38. Sometimes the final signoff response after QUIT could fail to get
1974 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1975 < 0 before doing a fflush(). This bug looks as though it goes back to the
1976 introduction of TLS in release 3.20, but "sometimes" must have been rare
1977 because the tests only now provoked it.
1983 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1984 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1985 Now it returns an integer. A number of calls were still expecting a T/F
1986 return. Fortuitously, in all cases, the tests worked in OK situations,
1987 which is the norm. However, things would have gone wrong on any write
1988 failures on the smtp file descriptor. This function is used when sending
1989 messages over SMTP and also when doing verify callouts.
1991 2. When Exim is called to do synchronous delivery of a locally submitted
1992 message (the -odf or -odi options), it no longer closes stderr before doing
1995 3. Implemented the mua_wrapper option.
1997 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1999 5. Implemented the functions header_remove(), header_testname(),
2000 header_add_at_position(), and receive_remove_recipient(), and exported them
2003 6. If an ACL "warn" statement specified the addition of headers, Exim already
2004 inserted X-ACL-Warn: at the start if there was no header name. However, it
2005 was not making this test for the second and subsequent header lines if
2006 there were newlines in the string. This meant that an invalid header could
2007 be inserted if Exim was badly configured.
2009 7. Allow an ACL "warn" statement to add header lines at the start or after all
2010 the Received: headers, as well as at the end.
2012 8. Added the rcpt_4xx retry error code.
2014 9. Added postmaster_mailfrom=xxx to callout verification option.
2016 10. Added mailfrom=xxxx to the callout verification option, for verify=
2019 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2020 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2022 12. Inserted some casts to stop certain compilers warning when using pointer
2023 differences as field lengths or precisions in printf-type calls (mostly
2024 affecting debugging statements).
2026 13. Added optional readline() support for -be (dynamically loaded).
2028 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2029 same clock tick as a message's arrival, so that its received time was the
2030 same as the "first fail" time on the retry record, and that message
2031 remained on the queue past the ultimate address timeout, every queue runner
2032 would try a delivery (because it was past the ultimate address timeout) but
2033 after another failure, the ultimate address timeout, which should have then
2034 bounced the address, did not kick in. This was a "< instead of <=" error;
2035 in most cases the first failure would have been in the next clock tick
2036 after the received time, and all would be well.
2038 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2039 being recognized when the domain list was tested by the match_domain
2040 condition in an expansion string.
2042 16. Added the ${str2b64: operator.
2044 17. Exim was always calling setrlimit() to set a large limit for the number of
2045 processes, without checking whether the existing limit was already
2046 adequate. (It did check for the limit on file descriptors.) Furthermore,
2047 errors from getrlimit() and setrlimit() were being ignored. Now they are
2048 logged to the main and panic logs, but Exim does carry on, to try to do its
2049 job under whatever limits there are.
2051 18. Imported PCRE 5.0.
2053 19. Trivial typo in log message " temporarily refused connection" (the leading
2056 20. If the log selector return_path_on_delivery was set and an address was
2057 redirected to /dev/null, the delivery process crashed because it assumed
2058 that a return path would always be set for a "successful" delivery. In this
2059 case, the whole delivery is bypassed as an optimization, and therefore no
2062 21. Internal re-arrangement: the function for sending a challenge and reading
2063 a response while authentication was assuming a zero-terminated challenge
2064 string. It's now changed to take a pointer and a length, to allow for
2065 binary data in such strings.
2067 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2069 23. Exim was not respecting finduser_retries when seeking the login of the
2070 uid under which it was called; it was always trying 10 times. (The default
2071 setting of finduser_retries is zero.) Also, it was sleeping after the final
2072 failure, which is pointless.
2074 24. Implemented tls_on_connect_ports.
2076 25. Implemented acl_smtp_predata.
2078 26. If the domain in control=submission is set empty, Exim assumes that the
2079 authenticated id is a complete email address when it generates From: or
2080 Sender: header lines.
2082 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2083 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2084 chown and chgrp in /bin and hostname in /usr/bin.
2086 28. Exim was keeping the "process log" file open after each use, just as it
2087 does for the main log. This opens the possibility of it remaining open for
2088 long periods when the USR1 signal hits a daemon. Occasional processlog
2089 errors were reported, that could have been caused by this. Anyway, it seems
2090 much more sensible not to leave this file open at all, so that is what now
2093 29. The long-running daemon process does not normally write to the log once it
2094 has entered its main loop, and it closes the log before doing so. This is
2095 so that log files can straightforwardly be renamed and moved. However,
2096 there are a couple of unusual error situations where the daemon does write
2097 log entries, and I had neglected to close the log afterwards.
2099 30. The text of an SMTP error response that was received during a remote
2100 delivery was being truncated at 512 bytes. This is too short for some of
2101 the long messages that one sometimes sees. I've increased the limit to
2104 31. It is now possible to make retry rules that apply only when a message has a
2105 specific sender, in particular, an empty sender.
2107 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2108 it possible to be selective about when SMTP synchronization is enforced.
2110 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2112 32. Implemented hosts_connection_nolog.
2114 33. Added an ACL for QUIT.
2116 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2119 35. Added mailbox_size and mailbox_filecount to appendfile.
2121 36. Added control = no_multiline_responses to ACLs.
2123 37. There was a bug in the logic of the code that waits for the clock to tick
2124 in the case where the clock went backwards by a substantial amount such
2125 that the microsecond fraction of "now" was more than the microsecond
2126 fraction of "then" (but the whole seconds number was less).
2128 38. Added support for the libradius Radius client library this is found on
2129 FreeBSD (previously only the radiusclient library was supported).
2135 1. When certain lookups returned multiple values in the form name=value, the
2136 quoting of the values was not always being done properly. Specifically:
2137 (a) If the value started with a double quote, but contained no whitespace,
2139 (b) If the value contained whitespace other than a space character (i.e.
2140 tabs or newlines or carriage returns) it was not quoted.
2141 This fix has been applied to the mysql and pgsql lookups by writing a
2142 separate quoting function and calling it from the lookup code. The fix
2143 should probably also be applied to nisplus, ibase and oracle lookups, but
2144 since I cannot test any of those, I have not disturbed their existing code.
2146 2. A hit in the callout cache for a specific address caused a log line with no
2147 reason for rejecting RCPT. Now it says "Previous (cached) callout
2148 verification failure".
2150 3. There was an off-by-one bug in the queryprogram router. An over-long
2151 return line was truncated at 256 instead of 255 characters, thereby
2152 overflowing its buffer with the terminating zero. As well as fixing this, I
2153 have increased the buffer size to 1024 (and made a note to document this).
2155 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2156 when Exim is waiting for an SMTP response from a remote server, Exim
2157 restarts its select() call on the socket, thereby resetting its timeout.
2158 This is not a problem when such interrupts are rare. Somebody set up a cron
2159 job to run exiwhat every 2 minutes, which is less than the normal select()
2160 timeout (5 or 10 minutes). This meant that the select() timeout never
2161 kicked in because it was always reset. I have fixed this by comparing the
2162 time when an interrupt arrives with the time at the start of the first call
2163 to select(). If more time than the timeout has elapsed, the interrupt is
2164 treated as a timeout.
2166 5. Some internal re-factoring in preparation for the addition of Sieve
2167 extensions (by MH). In particular, the "personal" test is moved to a
2168 separate function, and given an option for scanning Cc: and Bcc: (which is
2169 not set for Exim filters).
2171 6. When Exim created an email address using the login of the caller as the
2172 local part (e.g. when creating a From: or Sender: header line), it was not
2173 quoting the local part when it contained special characters such as @.
2175 7. Installed new OpenBSD configuration files.
2177 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2178 try to make them clearer.
2180 9. Callout options, other than the timeout value, were being ignored when
2181 verifying sender addresses in header lines. For example, when using
2183 verify = header_sender/callout=no_cache
2185 the cache was (incorrectly) being used.
2187 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2188 only the Cygwin environment.
2190 11. When return_path_on_delivery was set as a log selector, if different remote
2191 addresses in the same message used different return paths and parallel
2192 remote delivery occurred, the wrong values would sometimes be logged.
2193 (Whenever a remote delivery process finished, the return path value from
2194 the most recently started remote delivery process was logged.)
2196 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2197 lines when AUTH and/or TLS are in use. This is the "received protocol"
2198 field. Exim used to use "asmtp" for authenticated SMTP, without any
2199 indication (in the protocol name) for TLS use. Now it follows the RFC and
2200 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2201 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2202 names appear in log lines as well as in Received: header lines.
2204 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2205 extensions, and comparison tests, and to fix some bugs.
2207 14. Changes to the "personal" filter test:
2209 (1) The test was buggy in that it was just doing the equivalent of
2210 "contains" tests on header lines. For example, if a user's address was
2211 anne@some.where, the "personal" test would incorrectly be true for
2213 To: susanne@some.where
2215 This test is now done by extracting each address from the header in turn,
2216 and checking the entire address. Other tests that are part of "personal"
2217 are now done using regular expressions (for example, to check local parts
2218 of addresses in From: header lines).
2220 (2) The list of non-personal local parts in From: addresses has been
2221 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2222 taken from the Sieve specification recommendations.
2224 (3) If the message contains any header line starting with "List-" it is
2225 treated as non-personal.
2227 (4) The test for "circular" in the Subject: header line has been removed
2228 because it now seems ill-conceived.
2230 15. Minor typos in src/EDITME comments corrected.
2232 16. Installed latest exipick from John Jetmore.
2234 17. If headers_add on a router specified a text string that was too long for
2235 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2236 of string_sprintf() is now avoided.
2238 18. $message_body_size was not set (it was always zero) when running the DATA
2239 ACL and the local_scan() function.
2241 19. For the "mail" command in an Exim filter, no default was being set for
2242 the once_repeat time, causing a random time value to be used if "once" was
2243 specified. (If the value happened to be <= 0, no repeat happened.) The
2244 default is now 0s, meaning "never repeat". The "vacation" command was OK
2245 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2246 (I found it when inspecting the code).
2248 20. There is now an overall timeout for performing a callout verification. It
2249 defaults to 4 times the callout timeout, which applies to individual SMTP
2250 commands during the callout. The overall timeout applies when there is more
2251 than one host that can be tried. The timeout is checked before trying the
2252 next host. This prevents very long delays if there are a large number of
2253 hosts and all are timing out (e.g. when the network connections are timing
2254 out). The value of the overall timeout can be changed by specifying an
2255 additional sub-option for "callout", called "maxwait". For example:
2257 verify = sender/callout=5s,maxwait=20s
2259 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2260 to the end before writing, but this should make it even safer).
2262 22. Exim was forgetting that it had advertised PIPELINING for the second and
2263 subsequent messages on an SMTP connection. It was also not resetting its
2264 memory on STARTTLS and an internal HELO.
2266 23. When Exim logs an SMTP synchronization error within a session, it now
2267 records whether PIPELINING has been advertised or not.
2269 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2270 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2271 rather than long int.
2273 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2275 26. Added the never_mail option to autoreply.
2281 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2282 crash if the getsockname() call failed; this can happen if a connection is
2283 closed very soon after it is established. The problem was simply in the
2284 order in which certain operations were done, causing Exim to try to write
2285 to the SMTP stream before it had set up the file descriptor. The bug has
2286 been fixed by making things happen in the correct order.
2292 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2293 before the connection was closed, thus losing the rejection response.
2295 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2296 some early Solaris releases, but causes trouble in current releases where
2297 socklen_t is defined.
2299 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2302 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2305 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2306 that releasing the top of it at the end releases what was used for sub-
2307 expansions (unless the block got too big). However, discard this block if
2308 the first thing is a variable or header, so that we can use its block when
2309 it is dynamic (useful for very large $message_headers, for example).
2311 6. Lookups now cache *every* query, not just the most recent. A new, separate
2312 store pool is used for this. It can be recovered when all lookup caches are
2313 flushed. Lookups now release memory at the end of their result strings.
2314 This has involved some general refactoring of the lookup sources.
2316 7. Some code has been added to the store_xxx() functions to reduce the amount
2317 of flapping under certain conditions.
2319 8. log_incoming_interface used to affect only the <= reception log lines. Now
2320 it causes the local interface and port to be added to several more SMTP log
2321 lines, for example "SMTP connection from", and rejection lines.
2323 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2325 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2327 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2328 could be overwritten at the end of the current message (or the start of a
2329 new message if it was set in a HELO ACL). The value is now preserved for
2330 the duration of the SMTP connection.
2332 12. If a transport had a headers_rewrite setting, and a matching header line
2333 contained an unqualified address, that address was qualified, even if it
2334 did not match any rewriting rules. The underlying bug was that the values
2335 of the flags that permit the existence of unqualified sender and recipient
2336 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2337 non-local messages, and by -bnq for local messages) were not being
2338 preserved with the message after it was received.
2340 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2341 "next input=" as part of the text comprising the host identity instead of
2342 the correct text. The code was using the same buffer for two different
2343 strings. However, depending on which order the printing function evaluated
2344 its arguments, the bug did not always show up. Under Linux, for example, my
2345 test suite worked just fine.
2347 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2348 doesn't work with some older versions of Perl. It has been changed to "my",
2349 which in any case is probably the better facility to use.
2351 15. A really picky compiler found some instances of statements for creating
2352 error messages that either had too many or two few arguments for the format
2355 16. The size of the buffer for calls to the DNS resolver has been increased
2356 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2357 for addresses that have a lot of PTR records. This alleviates a problem; it
2358 does not fully solve it.
2360 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2361 buffer now truncates the list and logs the incident, which is the same
2362 action as happens when Exim is looking up a host name and its aliases.
2363 Previously in this situation something unpredictable would happen;
2364 sometimes it was "internal error: store_reset failed".
2366 18. If a server dropped the connection unexpectedly when an Exim client was
2367 using GnuTLS and trying to read a response, the client delivery process
2368 crashed while trying to generate an error log message.
2370 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2371 string, for example:
2373 warn message = H1: something\nH2: something
2375 the text was added as a single header line from Exim's point of view
2376 though it ended up OK in the delivered message. However, searching for the
2377 second and subsequent header lines using $h_h2: did not work. This has been
2378 fixed. Similarly, if a system filter added multiple headers in this way,
2379 the routers could not see them.
2381 20. Expanded the error message when iplsearch is called with an invalid key to
2382 suggest using net-iplsearch in a host list.
2384 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2385 ACLs are no longer actually imposed (and a message to that effect is
2388 22. If a "gecos" field in a passwd entry contained escaped characters, in
2389 particular, if it contained a \" sequence, Exim got it wrong when building
2390 a From: or a Sender: header from that name. A second bug also caused
2391 incorrect handling when an unquoted " was present following a character
2392 that needed quoting.
2394 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2395 was not being matched caselessly.
2397 24. Arranged for all hyphens in the exim.8 source to be escaped with
2400 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2401 properly. Recipient callout cache records were still being keyed to include
2402 the sender, even when use_sender was set false. This led to far more
2403 callouts that were necessary. The sender is no longer included in the key
2404 when use_sender is false.
2406 26. Added "control = submission" modifier to ACLs.
2408 27. Added the ${base62d: operator to decode base 62 numbers.
2410 28. dnsdb lookups can now access SRV records.
2412 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2413 the configuration file.
2415 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2416 (-v) mode. This makes the output for a verbose queue run more intelligible.
2418 31. Added a use_postmaster feature to recipient callouts.
2420 32. Added the $body_zerocount variable, containing the number of binary zero
2421 bytes in the message body.
2423 33. The time of last modification of the "new" subdirectory is now used as the
2424 "mailbox time last read" when there is a quota error for a maildir
2427 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2429 35. Added +ignore_unknown as a special item in host lists.
2431 36. Code for decoding IPv6 addresses in host lists is now included, even if
2432 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2433 address was recognized as an IP address, but was then not correctly decoded
2434 into binary, causing unexpected and incorrect effects when compared with
2441 1. Very minor rewording of debugging text in manualroute to say "list of
2442 hosts" instead of "hostlist".
2444 2. If verify=header_syntax was set, and a header line with an unqualified
2445 address (no domain) and a large number of spaces between the end of the
2446 name and the colon was received, the reception process suffered a buffer
2447 overflow, and (when I tested it) crashed. This was caused by some obsolete
2448 code that should have been removed. The fix is to remove it!
2450 3. When running in the test harness, delay a bit after writing a bounce
2451 message to get a bit more predictability in the log output.
2453 4. Added a call to search_tidyup() just before forking a reception process. In
2454 theory, someone could use a lookup in the expansion of smtp_accept_max_
2455 per_host which, without the tidyup, could leave open a database connection.
2457 5. Added the variables $recipient_data and $sender_data which get set from a
2458 lookup success in an ACL "recipients" or "senders" condition, or a router
2459 "senders" option, similar to $domain_data and $local_part_data.
2461 6. Moved the writing of debug_print from before to after the "senders" test
2464 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2465 problems for message scanning, either using a data ACL, or using
2466 local_scan() because the Received: header was not generated till after they
2467 were called (in order to set the time as the time of reception completion).
2468 I have revised the way this works. The header is now generated after the
2469 body is received, but before the ACL or local_scan() are called. After they
2470 are run, the timestamp in the header is updated.
2476 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2477 before starting a queue runner without re-exec. This happened only when
2478 deliver_drop_privilege was set or when the Exim user was set to root. The
2479 effect of the bug was that timeouts during subsequent deliveries caused
2480 crashes instead of being properly handled. The handler is now left at its
2481 default (and expected) setting.
2483 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2484 message, again when deliver_drop_privilege is set or Exim is run as root.
2485 The bug described in (1) was not present in this case, but the tidying up
2486 of the other signals was missing. I have made the two cases consistent.
2488 3. The ignore_target_hosts setting on a manualroute router was being ignored
2489 for hosts that were looked up using the /MX notation.
2491 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2494 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2495 operated on the sender address. After changing the $sender_address to <>
2496 for the sender address verify, Exim was re-instated it as the original
2497 (before rewriting) address, but remembering that it had rewritten it, so it
2498 wasn't rewriting it again. This bug also had the effect of breaking the
2499 sender address verification caching when the sender address was rewritten.
2501 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2502 This has been changed so that if the ip literal address matches
2503 ignore_target_hosts, the router declines.
2505 7. Added expansion conditions match_domain, match_address, and match_local_
2506 part (NOT match_host).
2508 8. The placeholder for the Received: header didn't have a length field set.
2510 9. Added code to Exim itself and to exim_lock to test for a specific race
2511 condition that could lead to file corruption when using MBX delivery. The
2512 issue is with the lockfile that is created in /tmp. If this file is removed
2513 after a process has opened it but before that process has acquired a lock,
2514 there is the potential for a second process to recreate the file and also
2515 acquire a lock. This could lead to two Exim processes writing to the file
2516 at the same time. The added code performs the same test as UW imapd; it
2517 checks after acquiring the lock that its file descriptor still refers to
2518 the same named file.
2520 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2521 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2524 11. Added the smtp_active_hostname option. If used, this will typically be made
2525 to depend on the incoming interface address. Because $interface_address is
2526 not set up until the daemon has forked a reception process, error responses
2527 that can happen earlier (such as "too many connections") no longer contain
2530 12. If an expansion in a condition on a "warn" statement fails because a lookup
2531 defers, the "warn" statement is abandoned, and the next ACL statement is
2532 processed. Previously this caused the whole ACL to be aborted.
2534 13. Added the iplsearch lookup type.
2536 14. Added ident_timeout as a log selector.
2538 15. Added tls_certificate_verified as a log selector.
2540 16. Added a global option tls_require_ciphers (compare the smtp transport
2541 option of the same name). This controls incoming TLS connections.
2543 17. I finally figured out how to make tls_require_ciphers do a similar thing
2544 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2545 before starting the TLS session.
2547 18. Tabs are now shown as \t in -bP output.
2549 19. If the log selector return_path_on_delivery was set, Exim crashed when
2550 bouncing a message because it had too many Received: header lines.
2552 20. If two routers both had headers_remove settings, and the first one included
2553 a superfluous trailing colon, the final name in the first list and the
2554 first name in the second list were incorrectly joined into one item (with a
2555 colon in the middle).
2561 1. Added -C and -D options to the exinext utility, mainly to make it easier
2562 to include in the automated testing, but these could be helpful when
2563 multiple configurations are in use.
2565 2. The exinext utility was not formatting the output nicely when there was
2566 an alternate port involved in the retry record key, nor when there was a
2567 message id as well (for retries that were specific to a specific message
2568 and a specific host). It was also confused by IPv6 addresses, because of
2569 the additional colons they contain. I have fixed the IPv4 problem, and
2570 patched it up to do a reasonable job for IPv6.
2572 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2573 delivery, the log line now contains "pipelined" if PIPELINING was used.
2575 4. An SMTP transport process used to panic and die if the bind() call to set
2576 an explicit outgoing interface failed. This has been changed; it is now
2577 treated in the same way as a connect() failure.
2579 5. A reference to $sender_host_name in the part of a conditional expansion
2580 that was being skipped was still causing a DNS lookup. This no longer
2583 6. The def: expansion condition was not recognizing references to header lines
2584 that used bh_ and bheader_.
2586 7. Added the _cache feature to named lists.
2588 8. The code for checking quota_filecount in the appendfile transport was
2589 allowing one more file than it should have been.
2591 9. For compatibility with Sendmail, the command line option
2599 and sets the incoming protocol and host name (for trusted callers). The
2600 host name and its colon can be omitted when only the protocol is to be set.
2601 Note the Exim already has two private options, -pd and -ps, that refer to
2602 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2603 "s", but I don't think that's a major issue.
2605 10. A number of refactoring changes to the code, none of which should affect
2608 (a) The number of logging options was getting close to filling up the
2609 32-bit word that was used as a bit map. I have split them into two classes:
2610 those that are passed in the argument to log_write(), and those that are
2611 only ever tested independently outside of that function. These are now in
2612 separate 32-bit words, so there is plenty of room for expansion again.
2613 There is no change in the user interface or the logging behaviour.
2615 (b) When building, for example, log lines, the code previously used a
2616 macro that called string_cat() twice, in order to add two strings. This is
2617 not really sufficiently general. Furthermore, there was one instance where
2618 it was actually wrong because one of the argument was used twice, and in
2619 one call a function was used. (As it happened, calling the function twice
2620 did not affect the overall behaviour.) The macro has been replaced by a
2621 function that can join an arbitrary number of extra strings onto a growing
2624 (c) The code for expansion conditions now uses a table and a binary chop
2625 instead of a serial search (which was left over from when there were very
2626 few conditions). Also, it now recognizes conditions like "pam" even when
2627 the relevant support is not compiled in: a suitably worded error message is
2628 given if an attempt is made to use such a condition.
2630 11. Added ${time_interval:xxxxx}.
2632 12. A bug was causing one of the ddress fields not to be passed back correctly
2633 from remote delivery subprocesses. The field in question was not being
2634 subsequently used, so this caused to problems in practice.
2636 13. Added new log selectors queue_time and deliver_time.
2638 14. Might have fixed a bug in maildirsizefile handling that threw up
2639 "unexpected character" debug warnings, and recalculated the data
2640 unnecessarily. In any case, I expanded the warning message to give more
2643 15. Added the message "Restricted characters in address" to the statements in
2644 the default ACL that block characters like @ and % in local parts.
2646 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2647 Three changes have been made:
2649 (a) There was a serious bug; a negative response to MAIL caused the whole
2650 recipient domain to be cached as invalid, thereby blocking all messages
2651 to all local parts at the same domain, from all senders. This bug has
2652 been fixed. The domain is no longer cached after a negative response to
2653 MAIL if the sender used is not empty.
2655 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2658 (c) A new callout option, "use_sender" has been added for people who want
2659 the modified behaviour.
2665 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2668 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2669 indeed breaks things for older releases.
2671 3. Added additional logging to the case where there is a problem reading data
2672 from a filter that is running in a subprocess using a pipe, in order to
2673 try to track down a specific problem.
2675 4. Testing facility fudge: when running in the test harness and attempting
2676 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2677 getting "No route to host". Convert this to a timeout.
2679 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2682 6. Some OS don't have socklen_t but use size_t instead. This affects the
2683 fifth argument of getsockopt() amongst other things. This is now
2684 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2685 can be set for individual OS. I have set it for SunOS5, OSF1, and
2686 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2687 some earlier ones do not.
2689 7. Change 4.30/15 was not doing the test caselessly.
2691 8. The standard form for an IPv6 address literal was being rejected by address
2692 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2693 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2694 this, as well as the form without the "IPv6" on the front (but only when
2695 address literals are enabled, of course).
2697 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2699 10. Exim crashed if a message with an empty sender address specified by -f
2700 encountered a router with an errors_to setting. This could be provoked only
2701 by a command such as
2705 where an empty string was supplied; "<>" did not hit this bug.
2707 11. Installed PCRE release 4.5.
2709 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2710 remained set. It is now erased.
2712 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2713 times from message ids (which are base 36 rather than the normal 62).
2715 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2716 were being counted as actual protocol errors, and logged if the log
2717 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2718 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2719 and DATA following a set of rejected RCPTs do not count as protocol errors.
2720 In other words, Exim assumes they were pipelined, though this may not
2721 actually be the case. Of course, in all cases the client gets an
2722 appropriate error code.
2724 15. If a lookup fails in an ACL condition, a message about the failure may
2725 be available; it is used if testing the ACL cannot continue, because most
2726 such messages specify what the cause of the deferral is. However, some
2727 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2728 that caused an old message to be retained and used if a later statement
2729 caused a defer, replacing the real cause of the deferral.
2731 16. If an IP address had so many PTR records that the DNS lookup buffer
2732 was not large enough to hold them, Exim could crash while trying to process
2733 the truncated data. It now detects and logs this case.
2735 17. Further to 4.21/58, another change has been made: if (and only if) the
2736 first line of a message (the first header line) ends with CRLF, a bare LF
2737 in a subsequent header line has a space inserted after it, so as not to
2738 terminate the header.
2740 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2741 unnecessarily, used atoi() instead of strtol(), and didn't check the
2742 termination when getting file sizes from file names by regex.
2744 19. Completely re-implemented the support for maildirsize files, in the light
2745 of a number of problems with the previous contributed implementation
2746 (4.30/29). In particular:
2748 . If the quota is zero, the maildirsize file is maintained, but no quota is
2751 . If the maildir directory does not exist, it is created before any attempt
2752 to write a maildirsize file.
2754 . The quota value in the file is just a cache; if the quota is changed in
2755 the transport, the new value overrides.
2757 . A regular expression is available for excluding directories from the
2760 20. The autoreply transport checks the characters in options that define the
2761 message's headers; it allows continued headers, but it was checking with
2762 isspace() after an embedded newline instead of explicitly looking for a
2765 21. If all the "regular" hosts to which an address was routed had passed their
2766 expiry times, and had not reached their retry times, the address was
2767 bounced, even if fallback hosts were defined. Now Exim should go on to try
2770 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2771 equivalent code in the SMTP transport. Some hosts send humungous responses
2772 to HELO/EHLO, more than 1024 it seems.
2774 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2775 gives compiler warnings in some environments. I've now done it "properly",
2778 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2779 (because of problems with the built-in one) was declared to return uschar *
2780 instead of char *, causing compiler failure.
2782 25. Fixed a file descriptor leak when processing alias/forward files.
2784 26. Fixed a minor format string issue in dbfn.c.
2786 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2788 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2789 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2792 29. When a sender address is verified, it is cached, to save repeating the test
2793 when there is more than one recipient in a message. However, when the
2794 verification involves a callout, it is possible for different callout
2795 options to be set for different recipients. It is too complicated to keep
2796 track of this in the cache, so now Exim always runs a verification when a
2797 callout is required, relying on the callout cache for the optimization.
2798 The overhead is duplication of the address routing, but this should not be
2801 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2802 to be verified with callout=postmaster, and the main callout worked but the
2803 postmaster check failed, the verification correctly failed. However, if a
2804 subsequent RCPT command asked for sender verification *without* the
2805 postmaster check, incorrect caching caused this verification also to fail,
2808 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2809 it was not caching the DNS options (qualify_single, search_parents) that
2810 were used when the lookup failed. A subsequent lookup with different
2811 options therefore always gave the same answer, though there were cases
2812 where it should not have. (Example: a "domains = !$mx_any" option on a
2813 dnslookup router: the "domains" option is always processed without any
2814 widening, but the router might have qualify_single set.) Now Exim uses the
2815 cached value only when the same options are set.
2817 32. Added John Jetmore's "exipick" utility to the distribution.
2819 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2820 than a timeout (e.g. a certificate is required, and is not provided), an
2821 Exim server now closes the connection immediately. Previously it waited for
2822 the client to close - but if the client is SSL, it seems that they each
2823 wait for each other, leading to a delay before one of them times out.
2825 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2826 maintained 0.8.x compatibility because I don't think many are using it, and
2827 it is clearly obsolete.
2829 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2832 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2833 client certificate was expired. A simple patch fixes this, though I don't
2834 understand the full logic of why the verify callback is called multiple
2837 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2838 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2839 which causes problems with some clients (such as the Certicom SSL Plus
2840 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2841 disables the coutermeasure allowing Eudora to connect."
2843 38. Exim was not checking that a write() to a log file succeeded. This could
2844 lead to Bad Things if a log got too big, in particular if it hit a file
2845 size limit. Exim now panics and dies if it cannot write to a log file, just
2846 as it does if it cannot open a log file.
2848 39. Modified OS/Makefile-Linux so that it now contains
2850 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2852 The two -D definitions ensure that Exim is compiled with large file
2853 support, which makes it possible to handle log files that are bigger than
2856 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2857 instance) a domain was checked against a named list that involved a lookup,
2858 causing $domain_data to be set, then another domain was checked against the
2859 same list, then the first domain was re-checked, the value of $domain_data
2860 after the final check could be wrong. In particular, if the second check
2861 failed, it could be set empty. This bug probably also applied to
2864 41. The strip_trailing_dot option was not being applied to the address given
2865 with the -f command-line option.
2867 42. The code for reading a message's header from the spool was incrementing
2868 $received_count, but never initializing it. This meant that the value was
2869 incorrect (doubled) while delivering a message in the same process in which
2870 it was received. In the most common configuration of Exim, this never
2871 happens - a fresh exec is done - but it can happen when
2872 deliver_drop_privilege is set.
2874 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2875 it now includes up to 150 characters of the unexpected data in the log
2878 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2879 and building data strings. The size of both of these buffers was 10 000
2880 bytes - far larger than anybody would *ever* want, thought I. Needless to
2881 say, somebody hit the limit. I have increased the maximum line length to
2882 20 000 and the maximum data length of concatenated lines to 100 000. I have
2883 also fixed two bugs, because there was no checking on these buffers. Tsk,
2884 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2885 buffer is too small.
2887 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2888 lsearch lookups. Now it does.
2890 46. When parsing a route_list item in a manualroute router, a fixed-length
2891 buffer was used for the list of hosts. I made this 1024 bytes long,
2892 thinking that nobody would ever have a list of hosts that long. Wrong.
2893 Somebody had a whole pile of complicated expansion conditions, and the
2894 string was silently truncated, leading to an expansion error. It turns out
2895 that it is easier to change to an unlimited length (owing to other changes
2896 that have happened since this code was originally written) than to build
2897 structure for giving a limitation error. The length of the item that
2898 expands into the list of hosts is now unlimited.
2900 47. The lsearch lookup could not handle data where the length of text line was
2901 more than 4095 characters. Such lines were truncated, leading to shortened
2902 data being returned. It should now handle lines of any length.
2904 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2905 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2908 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2909 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2910 no longer happens in comment lines. A list of replacements is now placed
2911 at the head of all of the source files, except those whose only change is
2912 to replace PERL_COMMAND in the very first #! line.
2914 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2915 messages on the queue, with a bottom-up merge sort, using code contributed
2916 by Michael Haardt. This should make operations like -bp somewhat faster on
2917 large queues. It won't affect queue runners, except when queue_run_in_order
2920 51. Installed eximstats 1.31 in the distribution.
2922 52. Added support for SRV lookups to the dnslookup router.
2924 53. If an ACL referred to $message_body or $message_body_end, the value was not
2925 reset for any messages that followed in the same SMTP session.
2927 54. The store-handling optimization for building very long strings was not
2928 differentiating between the different store pools. I don't think this
2929 actually made any difference in practice, but I've tidied it.
2931 55. While running the routers to verify a sender address, $sender_address
2932 was still set to the sender address. This is wrong, because when routing to
2933 send a bounce to the sender, it would be empty. Therefore, I have changed
2934 it so that, while verifying a sender address, $sender_address is set to <>.
2935 (There is no change to what happens when verifying a recipient address.)
2937 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2938 target A or AAAA records (if not already returned) without resetting the
2939 qualify_single or search_parents options of the DNS resolver. These are
2940 inappropriate in this case because the targets of MX and SRV records must
2941 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2942 target that, when qualified, matched something in the local domain. These
2943 two options are now turned off when doing these lookups.
2945 57. It seems that at least some releases of Reiserfs (which does not have the
2946 concept of a fixed number of inodes) returns zero and not -1 for the
2947 number of available inodes. This interacted badly with check_spool_inodes,
2948 which assumed that -1 was the "no such thing" setting. What I have done is
2949 to check that the total number of inodes is greater than zero before doing
2950 the test of how many are available.
2952 58. When a "warn" ACL statement has a log_message modifier, the message is
2953 remembered, and not repeated. This is to avoid a lot of repetition when a
2954 message has many recipients that cause the same warning to be written.
2955 Howewer, Exim was preserving the list of already written lines for an
2956 entire SMTP session, which doesn't seem right. The memory is now reset if a
2957 new message is started.
2959 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2960 debugging output unless log_rewrite was also set.
2962 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2963 of (int)(handle) when we know that handle contains (void *)(-1).
2965 61. The Exim daemon panic-logs an error return when it closes the incoming
2966 connection. However "connection reset by peer" seems to be common, and
2967 isn't really an error worthy of noting specially, so that particular error
2970 62. When Exim is trying to find all the local interfaces, it used to panic and
2971 die if the ioctl to get the interface flags failed. However, it seems that
2972 on at least one OS (Solaris 9) it is possible to have an interface that is
2973 included in the list of interfaces, but for which you get a failure error
2974 for this call. This happens when the interface is not "plumbed" into a
2975 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2976 failure of the "get flags" call assumes that the interface is down.
2978 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2979 makes life easier for people who are doing arithmetic on fields extracted
2980 from dates, where you often get leading zeros that should not be
2981 interpreted as octal.
2983 64. Added qualify_domain to the redirect router, to override the global
2986 65. If a pathologically long header line contained very many addresses (the
2987 report of this problem mentioned 10 000) and each of them was rewritten,
2988 Exim could use up a very large amount of memory. (It kept on making new
2989 copies of the header line as it rewrote, and never released the old ones.)
2990 At the expense of a bit more processing, the header rewriting function has
2991 been changed so that it no longer eats memory in this way.
2993 66. The generation of the Received: header has been moved from the time that a
2994 message starts to be received, to the time that it finishes. The timestamp
2995 in the Received: header should now be very close to that of the <= log
2996 line. There are two side-effects of this change:
2998 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2999 logged header lines no longer include the local Received: line, because
3000 it has not yet been created. The same applies to a copy of the message
3001 that is returned to a non-SMTP sender when a message is rejected.
3003 (b) When a filter file is tested using -bf, no additional Received: header
3004 is added to the test message. After some thought, I decided that this
3007 This change does not affect the value of $received_for. It is still set
3008 after address rewriting, but before local_scan() is called.
3010 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3012 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3013 gave an unhelpful panic error message, and a defer error. I have managed to
3014 change this behaviour so that it now rejects any supplied certificate,
3015 which seems right, as the list of acceptable certificates is empty.
3017 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3018 gave an unhelpful defer error. I have not managed to make this reject any
3019 supplied certificates, but the error message it gives is "no certificate
3020 supplied", which is not helpful.
3022 70. exigrep's output now also includes lines that are not associated with any
3023 message, but which match the given pattern. Implemented by a patch from
3024 Martin Sluka, which also tidied up the Perl a bit.
3026 71. Recipient callout verification, like sender verification, was using <> in
3027 the MAIL FROM command. This isn't really the right thing, since the actual
3028 sender may affect whether the remote host accepts the recipient or not. I
3029 have changed it to use the actual sender in the callout; this means that
3030 the cache record is now keyed on a recipient/sender pair, not just the
3031 recipient address. There doesn't seem to be a real danger of callout loops,
3032 since a callout by the remote host to check the sender would use <>.
3033 [SEE ABOVE: changed after hitting problems.]
3035 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3036 temporary errors. However, in the case of such a code being given after
3037 the end of a data transmission (i.e. after ".") Exim was failing to write
3038 a retry record for the message. (Yes, there was some broken host that was
3039 actually sending 8xx at this point.)
3041 73. An unknown lookup type in a host list could cause Exim to panic-die when
3042 the list was checked. (An example that provoked this was putting <; in the
3043 middle of a list instead of at the start.) If this happened during a DATA
3044 ACL check, a -D file could be left lying around. This kind of configuration
3045 error no longer causes Exim to die; instead it causes a defer errror. The
3046 incident is still logged to the main and panic logs.
3048 74. Buglet left over from Exim 3 conversion. The message "too many messages
3049 in one connection" was written to the rejectlog but not the mainlog, except
3050 when address rewriting (yes!) was being logged.
3052 75. Added write_rejectlog option.
3054 76. When a system filter was run not as root (that is, when system_filter_user
3055 was set), the values of the $n variables were not being returned to the
3056 main process; thus, they were not subsequently available in the $sn
3059 77. Added +return_path_on_delivery log selector.
3061 78. A connection timeout was being treated differently from recipients deferred
3062 when testing hosts_max_try with a message that was older than the host's
3063 retry timeout. (The host should not be counted, thus allowing all hosts to
3064 be tried at least once before bouncing.) This may have been the cause of an
3065 occasionally reported bug whereby a message would remain on the queue
3066 longer than the retry timeout, but would be bounced if a delivery was
3067 forced. I say "may" because I never totally pinned down the problem;
3068 setting up timeout/retry tests is difficult. See also the next item.
3070 79. The ultimate address timeout was not being applied to errors that involved
3071 a combination of host plus message (for example, a timeout on a MAIL
3072 command). When an address resolved to a number of possible hosts, and they
3073 were not all tried for each delivery (e.g. because of hosts_max_try), a
3074 message could remain on the queue longer than the retry timeout.
3076 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3079 81. Fixed an obscure SMTP outgoing bug which required at least the following
3080 conditions: (a) there was another message waiting for the same server;
3081 (b) the server returned 5xx to all RCPT commands in the first message so
3082 that the message was not completed; (c) the server dropped the connection
3083 or gave a negative response to the RSET that Exim sends to abort the
3084 transaction. The observed case was a dropped connection after DATA that had
3085 been sent in pipelining mode. That is, the server had advertised PIPELINING
3086 but was not implementing it correctly. The effect of the bug was incorrect
3087 behaviour, such as trying another host, and this could lead to a crash.
3093 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3094 and daemon.c were passed as pointers to ints; they should have been
3095 pointers to socklen_t variables (which are typically unsigned ints).
3097 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3100 3. Fixed a really odd bug that affected only the testing scheme; patching a
3101 certain fixed string in the binary changed the value of another string that
3102 happened to be identical to the end of the original first string.
3104 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3105 name", it returns that address as the IP address. On some operating
3106 systems (e.g. Solaris), it also passes back the IP address string as the
3107 "host name". However, on others (e.g. Linux), it passes back an empty
3108 string. Exim wasn't checking for this, and was changing the host name to an
3109 empty string, assuming it had been canonicized.
3111 5. Although rare, it is permitted to have more than one PTR record for a given
3112 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3113 all the names associated with an address, because they do in Solaris.
3114 However, it seems that they do not in Linux for data that comes from the
3115 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3116 I found this out when I moved to a new Linux workstation and tried to run
3117 the Exim test suite.
3119 To get round this problem I have changed the code so that it now does its
3120 own call to the DNS to look up PTR records when searching for a host name.
3121 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3122 addresses that are only in /etc/hosts are still found.
3124 This behaviour is, however, controlled by an option called host_lookup_
3125 order, which defaults to "bydns:byaddr". If people want to use the other
3126 order, or indeed, just use one or the other means of lookup, they can
3127 specify it in this variable.
3129 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3130 some operating systems, this comes back from gethostbyaddr() as an empty
3131 string, and this is what Exim used to test for. However, it seems that in
3132 other systems, "." is yielded. Exim now tests for this case too.
3134 7. The values of check_spool_space and check_log_space are now held internally
3135 as a number of kilobytes instead of an absolute number of bytes. If a
3136 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3137 kilobyte. This means that much larger values can be stored.
3139 8. Exim monitor: an attempt to get the action menu when not actually pointing
3140 at a message produces an empty menu entitled "No message selected". This
3141 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3142 no entries in it ("Shell widget menu has zero width and/or height"). So I
3143 have added a single, blank menu entry in this case.
3145 9. Added ${quote_local_part.
3147 10. MIME decoding is now applied to the contents of Subject: header lines when
3150 11. Now that a reference to $sender_host_address automatically causes a reverse
3151 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3152 host lookup before query-style lookups in lists that might use this
3153 variable. This has therefore been abolished, and the "net-" prefix is no
3154 longer necessary for query-style lookups.
3156 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3157 have been a typo for LFLAGS, so it has been changed.
3159 13. The install script calls Exim with "-C /dev/null" in order to find the
3160 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3161 to be output. Howeve, since Exim outputs its version number before the
3162 error, it didn't break the script. It just looked ugly. I fixed this by
3163 always allowing "-C /dev/null" if the caller is root.
3165 14. Ignore overlarge ACL variable number when reading spool file - insurance
3166 against a later release with more variables having written the file.
3168 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3169 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3170 this, as well as the form without the "IPv6" on the front.
3172 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3173 OS/Makefile-Darwin file.
3175 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3176 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3178 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3179 variable. However, some OS use ints for inodes. I've added cast to long int
3180 to get rid of the compiler warning.
3182 19. I had forgotten to lock out "/../" in configuration file names when
3183 ALT_CONFIG_PREFIX was set.
3185 20. Routers used for verification do not need to specify transports. However,
3186 if such a router generated a host list, and callout was configured, Exim
3187 crashed, because it could not find a port number from the (non-existent)
3188 transport. It now assumes port 25 in this circumstance.
3190 21. Added the -t option to exigrep.
3192 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3193 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3194 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3195 them is set, LOOKUP_LSEARCH is forced.
3197 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3199 24. Added sender and host information to the "rejected by local_scan()" log
3200 line; previously there was no indication of these.
3202 25. Added .include_if_exists.
3204 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3205 Linux. It turns out that not all file systems support this. Apparently some
3206 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3207 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3208 not supported on the file descriptor, is now ignored when Exim is trying to
3209 sync a directory. This applies only to Linux.
3211 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3213 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3214 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3217 29. Added support for maildirsize files from supplied patch (modified a bit).
3219 30. The use of :fail: followed by an empty string could lead Exim to respond to
3220 sender verification failures with (e.g.):
3222 550 Verification failed for <xxx>
3223 550 Sender verify failed
3225 where the first response line was missing the '-' that indicates it is not
3226 the final line of the response.
3228 31. The loop for finding the name of the user that called Exim had a hardwired
3229 limit of 10; it now uses the value of finduser_retries, which is used for
3230 all other user lookups.
3232 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3235 33. Exim was neglecting to zero errno before one call of strtol() when
3236 expanding a string and expecting an integer value. On some systems this
3237 resulted in spurious "integer overflow" errors. Also, it was casting the
3238 result into an int without checking.
3240 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3241 did not work. The code looks as if it has *never* worked, though it appears
3242 to have been documented since at least releast 1.62. I have made it work.
3244 35. The "timeout_DNS" error in retry rules, also documented since at least
3245 1.62, also never worked. As it isn't clear exactly what this means, and
3246 clearly it isn't a major issue, I have abolished the feature by treating it
3247 as "timeout", and writing a warning to the main and panic logs.
3249 36. The display of retry rules for -brt wasn't always showing the error code
3252 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3253 timeout_connect_A, timeout_connect_MX.
3255 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3256 to the empty sender.
3258 39. The daemon was not analysing the content of -oX till after it had closed
3259 stderr and disconnected from the controlling terminal. This meant that any
3260 syntax errors were only noted on the panic log, and the return code from
3261 the command was 0. By re-arranging the code a little, I've made the
3262 decoding happen first, so such errors now appear on stderr, and the return
3263 code is 1. However, the actual setting up of the sockets still happens in
3264 the disconnected process, so errors there are still only recorded on the
3267 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3268 connections (as happens on some IP stacks) was logged at start up time as
3269 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3270 it from "IPv6 and IPv4", which means that two separate sockets are being
3273 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3274 says whether AF_INET or AF_INET6 was passed as an argument.
3276 42. Exiwhat output was messed up when time zones were included in log
3279 43. Exiwhat now gives more information about the daemon's listening ports,
3280 and whether -tls-on-connect was used.
3282 44. The "port" option of the smtp transport is now expanded.
3284 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3285 silently ignored. Now an error message is written to the main and panic
3288 46. There's a new ACL modifier called "logwrite" which writes to a log file
3289 as soon as it is encountered.
3291 47. Added $local_user_uid and $local_user_gid at routing time.
3293 48. Exim crashed when trying to verify a sender address that was being
3296 49. Exim was recognizing only a space character after ".include". It now also
3297 recognizes a tab character.
3299 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3300 extracting the relevant information from the specification. The man page no
3301 longer contains scrambled data for the -d option, and I've added a section
3302 at the front about calling Exim under different names.
3304 51. Added "extra_headers" argument to the "mail" command in filter files.
3306 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3309 53. Installed eximstats 1.29.
3311 54. Added transport_filter_timeout as a generic transport option.
3313 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3314 Cc: header lines. This was required by RFC 822, but it not required by RFC
3317 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3318 incoming messages that did not have them. Now it does so only if the
3319 message originates locally, that is, if there is no associated remote host
3320 address. When Resent- header lines are present, this applies to the Resent-
3321 lines rather than the non-Resent- lines.
3323 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3324 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3326 58. Messages for configuration errors now include the name of the main
3327 configuration file - useful now that there may be more than one file in a
3328 list (.included file names were always shown).
3330 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3331 for those rare installations that do not start the daemon as root or run it
3332 setuid root. I've cut out the call to initgroups() if the daemon is not
3335 60. The Exim user and group can now be bound into the binary as text strings
3336 that are looked up at the start of Exim's processing.
3338 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3340 62. Added $mailstore_basename variable.
3342 63. Installed patch to sieve.c from Michael Haardt.
3344 64. When Exim failed to open the panic log after failing to open the main log,
3345 the original message it was trying to log was written to stderr and debug
3346 output, but if they were not available (the usual case in production), it
3347 was lost. Now it is written to syslog before the two lines that record the
3348 failures to open the logs.
3350 65. Users' Exim filters run in subprocesses under the user's uid. It is
3351 possible for a "deliver" command or an alias in a "personal" command to
3352 provoke an address rewrite. If logging of address rewriting is configured,
3353 this fails because the process is not running as root or exim. There may be
3354 a better way of dealing with this, but for the moment (because 4.30 needs
3355 to be released), I have disabled address rewrite logging when running a
3356 filter in a non-root, non-exim process.
3362 1. The buildconfig auxiliary program wasn't quoting the value set for
3363 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3364 not defined. This bug was masked in 4.22 by the effect that was fixed in
3367 2. Some messages that were rejected after a message id was allocated were
3368 shown as "incomplete" by exigrep. It no longer does this for messages that
3369 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3371 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3372 have allow_domain_literals set, the ID did not get logged in the <= line.
3373 Domain literals are now always recognized in Message-ID: header lines.
3375 4. The first argument for a ${extract expansion item is the key name or field
3376 number. Leading and trailing spaces in this item were not being ignored,
3377 causing some misleading effects.
3379 5. When deliver_drop_privilege was set, single queue runner processes started
3380 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3381 same command in the process it spins off) were not dropping privilege.
3383 6. When the daemon running as "exim" started a queue runner, it always
3384 re-executed Exim in the spun-off process. This is a waste of effort when
3385 deliver_drop_privilege is set. The new process now just calls the
3386 queue-runner function directly.
3392 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3395 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3396 ignored. Though the use of -oP was forcing the writing of a pid file, it
3397 was always written to the default place.
3399 3. If the message "no IP address found for host xxxx" is generated during
3400 incoming verification, it is now followed by identification of the incoming
3401 connection (so you can more easily find what provoked it).
3403 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3405 5. Added some features to "harden" Exim a bit more against certain attacks:
3407 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3408 be put in Local/Makefile. This is like the never_users runtime option,
3409 but it cannot be overridden. The default setting is "root".
3411 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3412 prefix string with which any file named in a -C command line option
3415 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3416 is retained for -C and -D only if the caller of Exim is root. Without
3417 it, the exim user may also use -C and -D and retain privilege.
3419 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3420 command line option is disabled.
3422 6. Macro names set by the -D option must start with an upper case letter, just
3423 like macro names defined in the configuration file.
3425 7. Added "dereference=" facility to LDAP.
3427 8. Two instances of the typo "uknown" in the source files are fixed.
3429 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3430 the Configure-Makefile script screwed up while processing it.
3432 10. Incorporated PCRE 4.4.
3434 11. The SMTP synchronization check was not operating right at the start of an
3435 SMTP session. For example, it could not catch a HELO sent before the client
3436 waited for the greeting. There is now a check for outstanding input at the
3437 point when the greeting is written. Because of the duplex, asynchronous
3438 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3439 way, but not yet received, when the check is performed.
3441 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3442 on TCP/IP sockets, because this apparently causes some broken clients to
3445 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3446 unchanged) from the Cygwin maintainer.
3448 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3449 was supported instead of testing for mbx. Effectively a typo.
3451 15. The spa authenticator server code was not checking that the input it
3452 received was valid base64.
3454 16. The debug output line for the "set" modifier in ACLs was not showing the
3455 name of the variable that was being set.
3457 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3459 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3460 lookup on its own. Something else was needed to trigger the lookup. For
3461 example, a match in host_lookup or the need for a host name in a host list.
3462 Now, if $sender_host_name is referenced and the host name has not yet been
3463 looked up, a lookup is performed. If the lookup fails, the variable remains
3464 empty, and $host_lookup_failed is set to "1".
3466 19. Added "eqi" as a case-independent comparison operator.
3468 20. The saslauthd authentication condition could segfault if neither service
3469 nor realm was specified.
3471 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3472 error message that was logged was misleading, and incoming SMTP
3473 connections were dropped. The message is now more accurate, and temporary
3474 errors are given to SMTP connections.
3476 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3477 (except RSET) with a 421 error, until QUIT is received. However, it was
3478 failing to send a response to QUIT.
3480 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3481 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3482 ACL is now run after the helo_try_verify_hosts code.
3484 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3485 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3486 case-independent, so other case variants are also recognized). Apparently
3487 some systems use these upper case variants.
3489 25. If more than two messages were waiting for the same host, and a transport
3490 filter was specified for the transport, Exim sent two messages over the
3491 same TCP/IP connection, and then failed with "socket operation on non-
3492 socket" when it tried to send the third.
3494 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3496 27. The extern definition of crypt16() in expand.c was not being excluded when
3497 the OS had its own crypt16() function.
3499 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3500 as a preferred synonym for return_size_limit, both as an option and as an
3503 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3505 30. Changed the default configuration ACL to relax the local part checking rule
3506 for addresses that are not in any local domains. For these addresses,
3507 slashes and pipe symbols are allowed within local parts, but the sequence
3508 /../ is explicitly forbidden.
3510 31. SPA server authentication was not clearing the challenge buffer before
3513 32. log_message in a "warn" ACL statement was writing to the reject log as
3514 well as to the main log, which contradicts the documentation and doesn't
3515 seem right (because no rejection is happening). So I have stopped it.
3517 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3518 However, I am unable to do any testing of this.
3520 34. Fixed an infelicity in the appendfile transport. When checking directories
3521 for a mailbox, to see if any needed to be created, it was accidentally
3522 using path names with one or more superfluous leading slashes; tracing
3523 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3525 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3526 contents are added to the log line that is written for every discarded
3527 recipient. (Previously a log_message setting was ignored.)
3529 36. The ${quote: operator now quotes the string if it is empty.
3531 37. The install script runs exim in order to find its version number. If for
3532 some reason other than non-existence or emptiness, which it checks, it
3533 could not run './exim', it was installing it with an empty version number,
3534 i.e. as "exim-". This error state is now caught, and the installation is
3537 38. An argument was missing from the function that creates an error message
3538 when Exim fails to connect to the socket for saslauthd authentication.
3539 This could cause Exim to crash, or give a corrupted message.
3541 39. Added isip, isip4, and isip6 to ${if conditions.
3543 40. The ACL variables $acl_xx are now saved with the message, and can be
3544 accessed later in routers, transports, and filters.
3546 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3547 strings in the file are not string-expanded.
3549 42. If a MAIL command specified a SIZE value that was too large to fit into an
3550 int variable, the check against message_size_limit failed. Such values are
3551 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3552 day this will have to be increased, but I don't think I want to be around
3553 when emails are that large.
3560 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3561 iconv() is not standard in FreeBSD.
3563 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3564 IPv6 enabled. The observed symptom was a segmentation fault on return from
3565 the function os_common_find_running_interfaces() in src/os.c.
3567 3. In the check_special_case() function in daemon.c I had used "errno" as an
3568 argument name, which causes warnings on some systems. This was basically a
3569 typo, since it was named "eno" in the comments!
3571 4. The code that waits for the clock to tick (at a resolution of some fraction
3572 of a second) so as to ensure message-id uniqueness was always waiting for
3573 at least one whole tick, when it could have waited for less. [This is
3574 almost certainly not relevant at current processor speeds, where it is
3575 unlikely to ever wait at all. But we try to future-proof.]
3577 5. The function that sleeps for a time interval that includes fractions of a
3578 second contained a race. It did not block SIGALRM between setting the
3579 timer, and suspending (a couple of lines later). If the interval was short
3580 and the sigsuspend() was delayed until after it had expired, the suspension
3581 never ended. On busy systems this could lead to processes getting stuck for
3584 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3585 process, before it forks any delivery processes. The open lookup caching
3586 mechanism meant that the open file or database connection was passed into
3587 the delivery process. The problem was that delivery processes always tidy
3588 up cached lookup data. This could cause a problem for the next delivery
3589 process started by the queue runner, because the external queue runner
3590 process does not know about the closure. So the next delivery process
3591 still has data in the lookup cache. In the case of a file lookup, there was
3592 no problem because closing a file descriptor in a subprocess doesn't affect
3593 the parent. However, if the lookup was caching a connection to a database,
3594 the connection was closed, and the second delivery process was likely to
3595 see errors such as "PGSQL: query failed: server closed the connection
3596 unexpectedly". The problem has been fixed by closing all cached lookups
3597 in a queue runner before running a delivery process.
3599 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3600 seem to have the "const" qualifier which it has on other OS. I've
3603 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3604 *of the same type* (client or server) with the same public name that an
3605 error should be diagnosed.
3607 9. When Exim looked up a host name for an IP address, but failed to find the
3608 original IP address when looking up the host name (a safety check), it
3609 output the message "<ip address> does not match any IP for NULL", which was
3610 confusing, to say the least. The bug was that the host name should have
3611 appeared instead of "NULL".
3613 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3614 user that is built into the binary, and the -C or -D options is used, root
3615 privilege is dropped before the configuration file is read. In addition,
3616 logging is switched to stderr instead of the normal log files. If the
3617 configuration then re-defines the Exim user, the unprivileged environment
3618 is probably not what is expected, so Exim logs a panic warning message (but
3621 However, if deliver_drop_privilege is set, the unprivileged state may well
3622 be exactly what is intended, so the warning has been cut out in that case,
3623 and Exim is allowed to try to write to its normal log files.
3629 1. smtp_return_error_details was not giving details for temporary sender
3630 or receiver verification errors.
3632 2. Diagnose a configuration error if two authenticators have the same public
3635 3. Exim used not to create the message log file for a message until the first
3636 delivery attempt. This could be confusing when incoming messages were held
3637 for policy or load reasons. The message log file is now created at the time
3638 the message is received, and an initial "Received" line is written to it.
3640 4. The automatically generated man page for command line options had a minor
3641 bug that caused no ill effects; however, a more serious problem was that
3642 the procedure for building the man page automatically didn't always
3643 operate. Consequently, release 4.20 contains an out-of-date version. This
3644 shouldn't happen again.
3646 5. When building Exim with embedded Perl support, the script that builds the
3647 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3648 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3650 6. The freeze_tell option was not being used for messages that were frozen on
3651 arrival, either by an ACL or by local_scan().
3653 7. Added the smtp_incomplete_transaction log selector.
3655 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3656 was accepting AUTH without a new EHLO.
3658 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3659 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3662 10. Logging of TCP/IP connections (when configured) now happens in the main
3663 daemon process instead of the child process, so that the TCP/IP connection
3664 count is more accurate (but it can never be perfect).
3666 11. The use of "drop" in a nested ACL was not being handled correctly in the
3667 outer ACL. Now, if condition failure induced by the nested "drop" causes
3668 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3669 or "require"), the connection is dropped.
3671 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3672 that yield "discard" can now be used with an "accept" or a "discard" verb,
3673 but an error is generated for any others (because I can't see a useful way
3674 to define what should happen).
3676 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3677 are now processed in the same way as lines in the Exim configuration file.
3678 In particular, continuation lines are supported.
3680 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3682 15. Added -ti meaning -t -i.
3684 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3685 domains, and warn by logging if others are found.
3687 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3688 returned from GIFCONF when Exim is trying to find the list of interfaces on
3689 a host. The code in os.c has been modified to copy each ifreq to an aligned
3690 structure in all cases.
3692 Also, in some cases, the returned ifreq's were being copied to a 'struct
3693 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3694 means the last couple of bytes of an IPv6 address could be chopped if the
3695 ifreq contained only a normal sockaddr (14 bytes storage).
3697 18. Named domain lists were not supported in the hosts_treat_as_local option.
3698 An entry such as +xxxx was not recognized, and was treated as a literal
3701 19. Ensure that header lines added by a DATA ACL are included in the reject log
3702 if the ACL subsequently rejects the message.
3704 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3705 MD5 (which is deprecated).
3707 21. When testing a filter file using -bf, Exim was writing a message when it
3708 took the sender from a "From " line in the message, but it was not doing so
3709 when it took $return_path from a Return-Path: header line. It now does.
3711 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3712 with a valid header line field name (a series of printing characters
3713 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3715 23. Changed "disc" in the source to "disk" to conform to the documentation and
3716 the book and for uniformity.
3718 24. Ignore Sendmail's -Ooption=value command line item.
3720 25. When execve() failed while trying to run a command in a pipe transport,
3721 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3722 could be confused with a return value of 69 from the command itself. This
3723 has been changed to 127, the value the shell returns if it is asked to run
3724 a non-existent command. The wording for the related log line suggests a
3725 non-existent command as the problem.
3727 26. If received_header_text expands to an empty string, do not add a Received:
3728 header line to the message. (Well, it adds a token one on the spool, but
3729 marks it "old" so that it doesn't get used or transmitted.)
3731 27. Installed eximstats 1.28 (addition of -nt option).
3733 28. There was no check for failure on the call to getsockname() in the daemon
3734 code. This can fail if there is a shortage of resources on the system, with
3735 ENOMEM, for example. A temporary error is now given on failure.
3737 29. Contrary to the C standard, it seems that in some environments, the
3738 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3739 program. Exim now does this explicitly; it affects the formatting of
3740 timestamps using strftime().
3742 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3743 complaints. I've now initialized all the variables, to avoid this.
3745 32. Header lines added by a system filter were not being "seen" during
3746 transport-time rewrites.
3748 33. The info_callback() function passed to OpenSSL is set up with type void
3749 (*)(SSL *, int, int), as described somewhere. However, when calling the
3750 function (actually a macro) that sets it up, the type void(*)() is
3751 expected. I've put in a cast to prevent warnings from picky compilers.
3753 34. If a DNS black list lookup found a CNAME record, but there were no A
3754 records associated with the domain it pointed at, Exim crashed.
3756 35. If a DNS black list lookup returned more than one A record, Exim ignored
3757 all but the first. It now scans all returned addresses if a particular IP
3758 value is being sought. In this situation, the contents of the
3759 $dnslist_value variable are a list of all the addresses, separated by a
3762 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3763 to accept a host name and all its aliases if the forward lookup for any of
3764 them yielded the IP address of the incoming connection. Now it accepts only
3765 those names whose forward lookup yields the correct IP address. Any other
3766 names are discarded. This closes a loophole whereby a rogue DNS
3767 administrator could create reverse DNS records to break through a
3768 wildcarded host restriction in an ACL.
3770 37. If a user filter or a system filter that ran in a subprocess used any of
3771 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3772 the wrong values were passed to the pipe command ($thisaddress had the
3773 value of $0, $0 had the value of $1, etc). This bug was introduced by
3774 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3776 38. Improved the line breaking for long SMTP error messages from ACLs.
3777 Previously, if there was no break point between 40 and 75 characters, Exim
3778 left the rest of the message alone. Two changes have been made: (a) I've
3779 reduced the minimum length to 35 characters; (b) if it can't find a break
3780 point between 35 and 75 characters, it looks ahead and uses the first one
3781 that it finds. This may give the occasional overlong line, but at least the
3782 remaining text gets split now.
3784 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3785 file descriptors might be low, and that setting 1000 would always raise it.
3786 It turns out that in some environments, the limit is already over 1000 and
3787 that lowering it causes trouble. So now Exim takes care not to decrease it.
3789 40. When delivering a message, the value of $return_path is set to $sender_
3790 address at the start of routing (routers may change the value). By an
3791 oversight, this default was not being set up when an address was tested by
3792 -bt or -bv, which affected the outcome if any router or filter referred to
3795 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3796 log only when "message" or "log_message" are set. However, if one of the
3797 conditions was an address verification, or a call to a nested ACL, the
3798 messages generated by the underlying test were being passed through. This
3799 no longer happens. The underlying message is available in $acl_verify_
3800 message for both "message" and "log_message" expansions, so it can be
3801 passed through if needed.
3803 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3804 new expansion $bh_ to give the encoded byte string without charset
3805 translation. Translation happens only if iconv() is available; HAVE_ICONV
3806 indicates this at build time. HEADERS_CHARSET gives the charset to
3807 translate to; headers_charset can change it in the configuration, and
3808 "headers charset" can change it in an individual filter file.
3810 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3811 that creates RFC 2047 encoded "words" labels them as that charset instead
3812 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3813 expansion operator; (ii) when Exim creates a From: line for a local
3814 message; (iii) when a header line is rewritten to include a "phrase" part.
3816 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3817 buggy, causing it to skip the first lines of messages whose message ID
3818 ended in 'D'. This would not have bitten before Exim release 4.14, because
3819 message IDs were unlikely to end in 'D' before then. The effect was to have
3820 incorrect size information for certain domains.
3822 45. #include "config.h" was missing at the start of the crypt16.c module. This
3823 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3826 46. If there was a timeout during a "random" callout check, Exim treated it as
3827 a failure of the random address, and carried on sending RSET and the real
3828 address. If the delay was just some slowness somewhere, the response to the
3829 original RCPT would be taken as a response to RSET and so on, causing
3830 mayhem of various kinds.
3832 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3833 when I implemented it. It didn't allow for the fact that some option values
3834 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3835 the right test for positive values.
3837 48. Domain names in DNS records are case-independent. Exim always looks them up
3838 in lower case. Some resolvers return domain names in exactly the case they
3839 appear in the zone file, that is, they may contain uppercase letters. Not
3840 all resolvers do this - some return always lower case. Exim was treating a
3841 change of case by a resolver as a change of domain, similar to a widening
3842 of a domain abbreviation. This triggered its re-routing code and so it was
3843 trying to route what was effectively the same domain again. This normally
3844 caused routing to fail (because the router wouldn't handle the domain
3845 twice). Now Exim checks for this case specially, and just changes the
3846 casing of the domain that it ultimately uses when it transmits the message
3849 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3852 50. If a filter generated a file delivery with a non-absolute name (possible if
3853 no home directory exists for the router), the forbid_file option was not
3856 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3857 the existing equality matching.
3859 52. Exim was using ints instead of ino_t variables in some places where it was
3860 dealing with inode numbers.
3862 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3863 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3864 called TMPDIR, and if it finds it is different, it changes its value.
3866 54. The smtp_printf() function is now made available to local_scan() so
3867 additional output lines can be written before returning. There is also an
3868 smtp_fflush() function to enable the detection of a dropped connection.
3869 The variables smtp_input and smtp_batched_input are exported to
3872 55. Changed the default runtime configuration: the message "Unknown user"
3873 has been removed from the ACL, and instead placed on the localuser router,
3874 using the cannot_route_message feature. This means that any verification
3875 failures that generate their own messages won't get overridden. Similarly,
3876 the "Unrouteable address" message that was in the ACL for unverifiable
3877 relay addresses has also been removed.
3879 56. Added hosts_avoid_esmtp to the smtp transport.
3881 57. The exicyclog script was not checking for the esoteric option
3882 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3883 will work only if exicyclog is run under the appropriate euid.
3885 58. Following a discussion on the list, the rules by which Exim recognises line
3886 endings on incoming messages have been changed. The -dropcr and drop_cr
3887 options are now no-ops, retained only for backwards compatibility. The
3888 following line terminators are recognized: LF CRLF CR. However, special
3889 processing applies to CR:
3891 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3892 nor a local message in the state where . is a terminator.
3894 (ii) If a bare CR is encountered in a header line, an extra space is added
3895 after the line terminator so as not to end the header. The reasoning
3896 behind this is that bare CRs in header lines are most likely either
3897 to be mistakes, or people trying to play silly games.
3899 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3900 was being incorrectly given as 18 bytes larger than it should have been.
3901 This is a VOB (very old bug).
3903 60. This may never have affected anything current, but just in case it has:
3904 When the local host is found other than at the start of a list of hosts,
3905 the local host, those with the same MX, and any that follow, are discarded.
3906 When the list in question was part of a longer list of hosts, the following
3907 hosts (not currently being processed) were also being discarded. This no
3908 longer happens. I'm not sure if this situation could ever has previously
3911 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3914 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3915 header. This is something that is recommended in a new Internet Draft, and
3916 is something that is documented as being done by Sendmail. There are two
3917 possible values. For messages generated by the autoreply transport, Exim
3920 Auto-Submitted: auto-replied
3922 whereas for all other generated messages (e.g. bounces) it adds
3924 Auto-Submitted: auto-generated
3926 63. The "personal" condition in filters now includes a test for the
3927 Auto-Submitted: header. If it contains the string "auto-" the message it
3928 not considered personal.
3930 64. Added rcpt_include_affixes as a generic transport option.
3932 65. Added queue_only_override (default true).
3934 66. Added the syslog_duplication option.
3936 67. If what should have been the first header line of a message consisted of
3937 a space followed by a colon, Exim was mis-interpreting it as a header line.
3938 It isn't of course - it is syntactically invalid and should therefore be
3939 treated as the start of the message body. The misbehaviour could have
3940 caused a number of strange effects, including loss of data in subsequent
3941 header lines, and spool format errors.
3943 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3944 client host had authenticated. This control can now be exercised by an ACL
3945 for more flexibility.
3947 69. By default, callouts do not happen when testing with -bh. There is now a
3948 variant, -bhc, which does actually run the callout code, including
3949 consulting and updating the callout cache.
3951 70. Added support for saslauthd authentication, courtesy of Alexander
3954 71. If statvfs() failed on the spool or log directories while checking their
3955 size for availability, Exim confusingly gave the error "space shortage".
3956 Furthermore, in debugging mode it crashed with a floating point exception.
3957 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3958 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3959 serious problem, Exim now writes to the main and panic logs when this
3960 happens, with details of the failure. It then refuses to accept the
3961 incoming message, giving the message "spool directory problem" or "log
3962 directory problem" with a 421 code for SMTP messages.
3964 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3965 0, 1, and 2 exist, because some OS complain for execs without them (see
3966 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3967 descriptors. However, the code omitted to check that the open succeeded,
3968 causing mysterious errors if for some reason the permissions on /dev/null
3969 got screwed. Now Exim writes a message to the main and panic logs, and
3970 bombs out if it can't open /dev/null.
3972 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3973 interact so that it is all more flexible. It is supposed to remain
3974 backwards compatible. Also added extra_local_interfaces.
3976 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3977 to bomb out with an assertion failure - to the client this appears as a
3978 connection drop. This problem occurs in the part of the code that was taken
3979 from the Samba project. Fortunately, the assertion is in a very simple
3980 function, so I have fixed this by reproducing the function inline in the
3981 one place where it is called, and arranging for authentication to fail
3982 instead of killing the process with assert().
3984 75. The SPA client code was not working when the server requested OEM rather
3985 than Unicode encoding.
3987 76. Added code to make require_files with a specific uid setting more usable in
3988 the case where statting the file as root fails - usually a non-root-mounted
3989 NFS file system. When this happens and the failure is EACCES, Exim now
3990 forks a subprocess and does the per-uid checking as the relevant uid.
3992 77. Added process_log_path.
3994 78. If log_file_path was not explicitly set, a setting of check_log_space or
3995 check_log_inodes was ignored.
3997 79. If a space check for the spool or log partitions fails, the incident is now
3998 logged. Of course, in the latter case the data may get lost...
4000 80. Added the %p formatting code to string_format() so that it can be used to
4001 print addresses in debug_print(). Adjusted all the address printing in the
4002 debugging in store.c to use %p rather than %d.
4004 81. There was a concern that a line of code in smtp_in.c could overflow a
4005 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4006 initially expressed, the concern was not well-founded, because trailing
4007 spaces are removed early. However, if the trailing spaces were followed by
4008 a NULL, they did not get removed, so the overflow was possible. Two fixes
4011 (a) I re-wrote the offending code in a cleaner fashion.
4012 (b) If an incoming SMTP command contains a NULL character, it is rejected
4015 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4016 runs initgroups(), so that if the Exim user is in any additional groups,
4017 they will be used during message reception.
4023 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob