1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.353 2006/05/22 18:44:24 fanf2 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 SC/01 Use a glob alias rather than an array ref in eximstats generated
10 parser. This improves both readability and performance.
12 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
13 Don't display local sender or destination tables in eximstats unless
14 there is data to show.
15 Added average volumes into the eximstats top table text output.
17 SC/03 Collect data on the number of addresses (recipients) as well
18 as the number of messages in eximstats.
20 TF/01 Correct an error in the documentation for the redirect router. Exim
21 does (usually) call initgroups() when daemonizing.
23 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs with
24 consistent privilege compared to when running as a daemon.
29 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
30 other effects) broke the use of negated acl sub-conditions.
32 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
35 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
36 "Deny" causes Exim to reject the incoming connection with a 554 error.
37 Unfortunately, if there is a major crisis, such as a disk failure,
38 tcp-wrappers gives "deny", whereas what one would like would be some
39 kind of temporary error. A kludge has been added to help with this.
40 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
41 554 error is used if errno is still zero or contains ENOENT (which occurs
42 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
45 PH/03 Add -lutil to the default FreeBSD LIBS setting.
47 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
48 errors. Otherwise a message that provokes a temporary error (when other
49 messages do not) can cause a whole host to time out.
51 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
52 addresses were routed directly to files or pipes from a redirect router.
53 File deliveries just didn't batch; pipe deliveries might have suffered
56 PH/06 A failure to get a lock for a hints database would erroneously always say
57 "Failed to get write lock", even when it was really a read lock.
59 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
60 of 0600. This has been changed to use the value of the lockfile_mode
61 option (which defaults to 0600).
63 PH/08 Applied small patch from the Sieve maintainer.
65 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
66 folder from quota calculations, a direct delivery into this folder messed
67 up the contents of the maildirsize file. This was because the regex was
68 used only to exclude .Trash (or whatever) when the size of the mailbox
69 was calculated. There was no check that a delivery was happening into an
70 excluded directory. This bug has been fixed by ignoring all quota
71 processing for deliveries into excluded directories.
73 PH/10 Added the maildirfolder_create_regex option to appendfile.
79 PH/01 The code for finding all the local interface addresses on a FreeBSD
80 system running IPv6 was broken. This may well have applied to all BSD
81 systems, as well as to others that have similar system calls. The broken
82 code found IPv4 interfaces correctly, but gave incorrect values for the
83 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
84 that it would not match correctly against @[] and not recognize the IPv6
87 PH/02 The ipliteral router was not recognizing addresses of the form user@
88 [ipv6:....] because it didn't know about the "ipv6:" prefix.
90 PH/03 Added disable_ipv6.
92 PH/04 Changed $reply_address to use the raw form of the headers instead of the
93 decoded form, because it is most often used to construct To: headers
94 lines in autoreplies, and the decoded form may well be syntactically
95 invalid. However, $reply_address has leading white space removed, and all
96 newlines turned into spaces so that the autoreply transport does not
99 PH/05 If group was specified without a user on a router, and no group or user
100 was specified on a transport, the group from the router was ignored.
102 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
103 for visible compile-time settings that can be used to change these
104 numbers, for those that want even more. Backwards compatibility with old
105 spool files has been maintained. However, going back to a previous Exim
106 release will lost any variables that are in spool files.
108 PH/07 Two small changes when running in the test harness: increase delay when
109 passing a TCP/IP connection to a new process, in case the original
110 process has to generate a bounce, and remove special handling of
111 127.0.0.2 (sic), which is no longer necessary.
113 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
114 be the same on different OS.
116 PH/09 Moved a debug statement in filter processing to avoid a race problem when
119 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
120 whether --show-vars was specified or not
122 JJ/02 exipick: Added support for new ACL variable spool format introduced
125 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
126 syntactically invalid From: or Reply-to: line, and a filter used this to
127 generate an autoreply, and therefore failed to obtain an address for the
128 autoreply, Exim could try to deliver to a non-existent relative file
129 name, causing unrelated and misleading errors. What now happens is that
130 it logs this as a hard delivery error, but does not attempt to create a
133 PH/11 The exinext utility has a -C option for testing purposes, but although
134 the given file was scanned by exinext itself; it wasn't being passed on
135 when Exim was called.
137 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
138 an end-of-file indication when reading a command response.
140 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
141 compiled. In many other places in Exim, IPv6 addresses are always
142 recognized, so I have changed this. It also means that IPv4 domain
143 literals of the form [IPV4:n.n.n.n] are now always recognized.
145 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
146 used if the router is not running as root, for example, when verifying at
147 ACL time, or when using -bh. The debugging output from this situation was
148 non-existent - all you got was a failure to exec. I have made two
151 (a) Failures to set uid/gid, the current directory, or a process leader
152 in a subprocess such as that created by queryprogram now generate
153 suitable debugging ouput when -d is set.
155 (b) The queryprogram router detects when it is not running as root,
156 outputs suitable debugging information if -d is set, and then runs
157 the subprocess without attempting to change uid/gid.
159 PH/15 Minor change to Makefile for building test_host (undocumented testing
162 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
163 additional section of a DNS packet that returns MX or SRV records.
164 Instead, it always explicitly searches for A/AAAA records. This avoids
165 major problems that occur when a DNS server includes only records of one
166 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
167 fixed another bug: if SRV records were looked up and the corresponding
168 address records were *not* found in the additional section, the port
169 values from the SRV records were lost.
171 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
172 using the correct key (the original address) when searching the retry
173 rules in order to find which one to use for generating the retry hint.
175 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
176 adding the default one. Similarly, if it contains a Reply-To: header, the
177 errors_reply_to option, if set, is not used.
179 PH/19 When calculating a retry time, Exim used to measure the "time since
180 failure" by looking at the "first failed" field in the retry record. Now
181 it does not use this if it is later than than the arrival time of the
182 message. Instead it uses the arrival time. This makes for better
183 behaviour in cases where some deliveries succeed, thus re-setting the
184 "first failed" field. An example is a quota failure for a huge message
185 when small messages continue to be delivered. Without this change, the
186 "time since failure" will always be short, possible causing more frequent
187 delivery attempts for the huge message than are intended.
188 [Note: This change was subsequently modified - see PH/04 for 4.62.]
190 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
191 $1, $2, $3) because the numerical variables can be reset during some
192 expansion items (e.g. "match"), thereby losing the authentication data.
194 PH/21 Make -bV show the size of off_t variables so that the test suite can
195 decide whether to run tests for quotas > 2G.
197 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
198 mailbox_size, and mailbox_filecount in the appendfile transport. If a
199 filecount value is greater than 2G or if a quota value is greater than 2G
200 on a system where the size of off_t is not greater than 4, a panic error
203 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
204 never match. The debug and -bh output now contains an explicit error
205 message indicating a malformed IPv4 address or mask.
207 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
208 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
211 PH/25 Do not write to syslog when running in the test harness. The only
212 occasion when this arises is a failure to open the main or panic logs
213 (for which there is an explicit test).
215 PH/26 Added the /no_tell option to "control=freeze".
217 PH/27 If a host name lookup failed very early in a connection, for example, if
218 the IP address matched host_lookup and the reverse lookup yielded a name
219 that did not have a forward lookup, an error message of the form "no IP
220 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
221 could be logged. Now it outputs the IP address instead of "NULL".
223 PH/28 An enabling patch from MH: add new function child_open_exim2() which
224 allows the sender and the authenticated sender to be set when
225 submitting a message from within Exim. Since child_open_exim() is
226 documented for local_scan(), the new function should be too.
228 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
229 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
230 results in an empty string is now treated as unset.
232 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
234 PH/31 Added sender_verify_fail logging option.
236 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
237 needed by RFC 822 but not by RFC 2822 was commented out. I have now
238 tidied the source and removed it altogether.
240 PH/33 When a queue run was abandoned because the load average was too high, a
241 log line was always written; now it is written only if the queue_run log
242 selector is set. In addition, the log line for abandonment now contains
243 information about the queue run such as the pid. This is always present
244 in "start" and "stop" lines but was omitted from the "abandon" line.
246 PH/34 Omit spaces between a header name and the colon in the error message that
247 is given when verify = headers_syntax fails (if there are lots of them,
248 the message gets confusing).
250 PH/35 Change the default for dns_check_names_pattern to allow slashes within
251 names, as there are now some PTR records that contain slashes. This check
252 is only to protect against broken name servers that fall over on strange
253 characters, so the fact that it applies to all lookups doesn't matter.
255 PH/36 Now that the new test suite is complete, we can remove some of the
256 special code in Exim that was needed for the old test suite. For example,
257 sorting DNS records because real resolvers return them in an arbitrary
258 order. The new test suite's fake resolver always returns records in the
261 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
262 bounces) except when queue_only is set, to avoid logging races between
263 the different processes.
265 PH/38 Panic-die if .include specifies a non-absolute path.
267 PH/39 A tweak to the "H" retry rule from its user.
269 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
270 a label. They prevented compilation on older perls.
272 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
273 a warning to be raised on newish perls.
275 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
276 on queue. Changes to match documented behaviour of showing count of
277 messages matching specified criteria.
279 PH/40 Changed the default ident timeout from 30s to 5s.
281 PH/41 Added support for the use of login_cap features, on those BSD systems
282 that have them, for controlling the resources used by pipe deliveries.
284 PH/42 The content-scanning code uses fopen() to create files in which to put
285 message data. Previously it was not paying any attention to the mode of
286 the files. Exim runs with umask(0) because the rest of the code creates
287 files with open(), and sets the required mode explicitly. Thus, these
288 files were ending up world-writeable. This was not a big issue, because,
289 being within the spool directory, they were not world-accessible. I have
290 created a function called modefopen, which takes an additional mode
291 argument. It sets umask(777), creates the file, chmods it to the required
292 mode, then resets the umask. All the relevant calls to fopen() in the
293 content scanning code have been changed to use this function.
295 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
296 to 24 hours. This avoids potential overflow problems when processing G
297 and H retry rules. I suspect nobody ever tinkers with this value.
299 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
301 PH/45 When the plaintext authenticator is running as a client, the server's
302 challenges are checked to ensure they are valid base64 strings. By
303 default, the authentication attempt is cancelled if an invalid string is
304 received. Setting client_ignore_invalid_base64 true ignores these errors.
305 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
306 they are received. Thus, the responses can be made to depend on the
307 challenges. If an invalid string is ignored, an empty string is placed in
310 PH/46 Messages that are created by the autoreply transport now contains a
311 References: header, in accordance with RFCs 2822 and 3834.
313 PH/47 Added authenticated_sender_force to the smtp transport.
315 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
317 PH/49 Installed latest patch from the Sieve maintainer.
319 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
320 was also set, the appendfile transport was unnecessarily scanning a
321 directory of message files (e.g. for maildir delivery) to find the count
322 of files (along with the size), even though it did not need this
323 information. It now does the scan only if it needs to find either the
324 size of the count of files.
326 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
328 PH/52 Two bugs concerned with error handling when the smtp transport is
331 (i) Exim was not creating retry information for temporary errors given
332 for individual recipients after the DATA command when the smtp transport
333 was used in LMTP mode. This meant that they could be retried too
334 frequently, and not timed out correctly.
336 (ii) Exim was setting the flag that allows error details to be returned
337 for LMTP errors on RCPT commands, but not for LMTP errors for individual
338 recipients that were returned after the DATA command.
340 PH/53 This is related to PH/52, but is more general: for any failing address,
341 when detailed error information was permitted to be returned to the
342 sender, but the error was temporary, then after the final timeout, only
343 "retry timeout exceeded" was returned. Now it returns the full error as
344 well as "retry timeout exceeded".
346 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
347 do this, and (what is worse) MTAs that accept it.
349 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
350 will now be deprecated.
352 PH/56 New os.c-cygwin from the Cygwin maintainer.
354 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
355 formats (previously only available in exim formats via -bpr, -bpru,
356 and -bpra. Now also available in native and exiqgrep formats)
358 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
359 with very large, slow to parse queues
361 JJ/08 exipick: added ! as generic prefix to negate any criteria format
363 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
365 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
366 responses to authentication challenges, though it was showing the
367 challenges; (ii) I've removed the CR characters from the debug output for
370 PH/58 Allow for the insertion of a newline as well as a space when a string
371 is turned into more than one encoded-word during RFC 2047 encoding. The
372 Sieve code now uses this.
374 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
375 data_4xx, lost_connection, tls_required.
377 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
378 message was being sent as an SMTP response.
380 PH/61 Add -l and -k options to exicyclog.
382 PH/62 When verifying, if an address was redirected to one new address, so that
383 verification continued, and the new address failed or deferred after
384 having set something in $address_data, the value of $address_data was not
385 passed back to the ACL. This was different to the case when no
386 redirection occurred. The value is now passed back in both cases.
388 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
389 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
390 use login_cap.h, so on its own it isn't the distinguishing feature. The
391 new name refers directly to the setclassresources() function.
393 PH/65 Added configuration files for NetBSD3.
395 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
397 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
398 is preferred over IPv4.
400 PH/68 The bounce_return_message and bounce_return_body options were not being
401 honoured for bounces generated during the reception of non-SMTP messages.
402 In particular, this applied to messages rejected by the ACL. This bug has
403 been fixed. However, if bounce_return_message is true and bounce_return_
404 body is false, the headers that are returned for a non-SMTP message
405 include only those that have been read before the error was detected.
406 (In the case of an ACL rejection, they have all been read.)
408 PH/69 The HTML version of the specification is now built in a directory called
409 spec_html instead of spec.html, because the latter looks like a path with
410 a MIME-type, and this confuses some software.
412 PH/70 Catch two compiler warnings in sieve.c.
414 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
415 function verify_get_ident() calls ip_connect() to connect a socket, but
416 if the "connect()" function timed out, ip_connect() used to close the
417 socket. However, verify_get_ident() also closes the socket later, and in
418 between Exim writes to the log, which may get opened at this point. When
419 the socket was closed in ip_connect(), the log could get the same file
420 descriptor number as the socket. This naturally causes chaos. The fix is
421 not to close the socket in ip_connect(); the socket should be closed by
422 the function that creates it. There was only one place in the code where
423 this was missing, in the iplookup router, which I don't think anybody now
424 uses, but I've fixed it anyway.
426 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
427 well as to direct DNS lookups. Otherwise the handling of names in host
428 lists is inconsistent and therefore confusing.
434 PH/01 Two changes to the default runtime configuration:
436 (1) Move the checks for relay_from_hosts and authenticated clients from
437 after to before the (commented out) DNS black list checks.
439 (2) Add control=submission to the relay_from_hosts and authenticated
440 clients checks, on the grounds that messages accepted by these
441 statements are most likely to be submissions.
443 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
445 (1) Generate an error if the third argument for the ${prvs expansion is
448 (2) Treat a missing third argument of ${prvscheck as if it were an empty
451 (3) Reset the variables that are obtained from the first argument of
452 ${prvscheck and used in the second argument before leaving the code,
453 because their memory is reclaimed, so using them afterwards may do
456 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
457 one (it's much easier than Tom thought :-).
459 (5) Because of (4), we can now allow for the use of $prvscheck_result
460 inside the third argument.
462 PH/03 For some reason, the default setting of PATH when running a command from
463 a pipe transport was just "/usr/bin". I have changed it to
466 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
467 anything to be listed in the output from -bV.
469 PH/05 When a filter generated an autoreply, the entire To: header line was
470 quoted in the delivery log line, like this:
472 => >A.N.Other <ano@some.domain> <original@ddress> ...
474 This has been changed so that it extracts the operative address. There
475 may be more than one such address. If so, they are comma-separated, like
478 => >ano@some.domain,ona@other.domain <original@ddress> ...
480 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
481 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
482 not being looked up in the rDNS to get a host name, Exim was showing the
483 IP address twice in Received: lines, even though the IP addresses were
484 identical. For example:
486 Received: from [1.2.3.4] (helo=[1.2.3.4])
488 However, if the real host name was known, it was omitting the HELO data
489 if it matched the actual IP address. This has been tidied up so that it
490 doesn't show the same IP address twice.
492 PH/07 When both +timestamp and +memory debugging was on, the value given by
493 $tod_xxx expansions could be wrong, because the tod_stamp() function was
494 called by the debug printing, thereby overwriting the timestamp buffer.
495 Debugging no longer uses the tod_stamp() function when +timestamp is set.
497 PH/08 When the original message was included in an autoreply transport, it
498 always said "this is a copy of the message, including all the headers",
499 even if body_only or headers_only was set. It now gives an appropriate
502 PH/09 Applied a patch from the Sieve maintainer which:
504 o fixes some comments
505 o adds the (disabled) notify extension core
506 o adds some debug output for the result of if/elsif tests
507 o points to the current vacation draft in the documentation
508 and documents the missing references header update
512 o fixes a bug in processing the envelope test (when testing
513 multiple envelope elements, the last element determinted the
516 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
517 Electronic Mail") by including:
519 Auto-submitted: auto-generated
521 in the messages that it generates (bounce messages and others, such as
522 warnings). In the case of bounce messages for non-SMTP mesages, there was
523 also a typo: it was using "Auto_submitted" (underscore instead of
524 hyphen). Since every message generated by Exim is necessarily in response
525 to another message, thes have all been changed to:
527 Auto-Submitted: auto-replied
529 in accordance with these statements in the RFC:
531 The auto-replied keyword:
533 - SHOULD be used on messages sent in direct response to another
534 message by an automatic process,
536 - MUST NOT be used on manually-generated messages,
538 - MAY be used on Delivery Status Notifications (DSNs) and Message
539 Disposition Notifications (MDNs),
541 - MUST NOT be used on messages generated by automatic or periodic
542 processes, except for messages which are automatic responses to
545 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
546 to the default Received: header definition.
548 PH/12 Added log selector acl_warn_skipped (default on).
550 PH/13 After a successful wildlsearch lookup, discard the values of numeric
551 variables because (a) they are in the wrong storage pool and (b) even if
552 they were copied, it wouldn't work properly because of the caching.
554 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
555 checking when decoding. Apparently there are clients that generate
556 overlong encoded strings. Why am I not surprised?
558 PH/15 If the first argument of "${if match_address" was not empty, but did not
559 contain an "@" character, Exim crashed. Now it writes a panic log message
560 and treats the condition as false.
562 PH/16 In autoreply, treat an empty string for "once" the same as unset.
564 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
565 extension "envelope-auth". The code is finished and in agreement with
566 other implementations, but there is no documentation so far and in fact,
567 nobody wrote the draft yet. This extension is currently #undef'ed, thus
568 not changing the active code.
570 Print executed "if" and "elsif" statements when debugging is used. This
571 helps a great deal to understand what a filter does.
573 Document more things not specified clearly in RFC3028. I had all this
574 sorted out, when out of a sudden new issues came to my mind. Oops."
576 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
579 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
580 canonical form (as documented). However, after a host name lookup from
581 the IP address, check_host() was doing a simple string comparison with
582 addresses acquired from the DNS when checking that the found name did
583 have the original IP as one of its addresses. Since any found IPv6
584 addresses are likely to be in abbreviated form, the comparison could
585 fail. Luckily, there already exists a function for doing the comparison
586 by converting both addresses to binary, so now that is used instead of
589 PH/20 There was another similar case to PH/19, when a complete host name was
590 given in a host list; looking up its IP address could give an abbreviated
591 form, whereas the current host's name might or might not be abbreviated.
592 The same fix has been applied.
598 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
599 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
602 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
603 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
605 PH/03 Typo: missing ".o" in src/pcre/Makefile.
607 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
608 header line, restrict the check to what is listed in RFCs 2369 and 2929.
609 Also, for "Auto-Submitted", treat anything other than "no" as
610 non-personal, in accordance with RFC 3834. (Previously it treated
611 anything starting "auto-" as non-personal.)
613 TF/01 The control=submission/name=... option had a problem with syntax
614 errors if the name included a slash character. The /name= option
615 now slurps the rest of the string, so it can include any characters
616 but it must come last in the list of options (after /sender_retain
619 PH/05 Some modifications to the interface to the fake nameserver for the new
627 TK/01 Added the "success_on_redirect" address verification option. See
628 NewStuff for rationale and an example.
630 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
632 PH/02 Patch to exigrep to allow it to work on syslog lines.
634 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
635 fread() to skip over the body file's header line, because in Cygwin the
636 header line is locked and is inaccessible.
638 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
639 co-exist for some time) to make it clear that it is the Exim ID that is
640 referenced, not the Message-ID: header line.
642 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
643 string_format() function, because snprintf() does not exist on all
646 PH/06 The use of forbid_filter_existstest now also locks out the use of the
647 ${stat: expansion item.
649 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
650 protocol synchronization error", to keep the pedants happy.
652 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
653 well as for IRIX systems, when gcc is being used. See the host.c source
656 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
658 PH/10 Named domain lists were not working if used in a queue_smtp_domains
661 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
662 transport and to the smtp transport in LMTP mode.
664 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
666 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
667 run a filter in a subprocess. This could lead to confusion in subsequent
668 lookups in the parent process. There should also be a search_tidyup() at
669 the end of the subprocess.
671 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
672 only if the host matched helo_try_verify_hosts, which caused the
673 verification to occur when the EHLO/HELO command was issued. The ACL just
674 tested the remembered result. Now, if a previous verification attempt has
675 not happened, "verify = helo" does it there and then.
677 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
679 TK/03 Fix log output including CR from clamd.
681 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
682 exist provoked a memory error which could cause a segfault.
684 PH/15 Installed PCRE 6.2
686 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
688 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
689 of the problem. Specifically, suggested +O2 rather than +O1 for the
692 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
694 PH/20 If a delivery was routed to a non-standard port by means of an SRV
695 record, the port was not correctly logged when the outgoing_port log
696 selector was set (it logged the transort's default port).
698 PH/21 Added support for host-specific ports to manualroute, queryprogram,
699 fallback_hosts, and "hosts" in the smtp transport.
701 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
702 host errors such as "Connection refused".
704 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
705 authentication with radiusclient 0.4.9:
707 - Error returned from rc_read_config was caught wrongly
708 - Username/password not passed on to radius server due to wrong length.
710 The presumption is that some radiusclient API changes for 4.51/PH/17
711 were not taken care of correctly. The code is still untested by me (my
712 Linux distribution still has 0.3.2 of radiusclient), but it was
713 contributed by a Radius user.
715 PH/24 When doing a callout, the value of $domain wasn't set correctly when
716 expanding the "port" option of the smtp transport.
718 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
719 while reading a MIME header. Thanks to Tom Hughes for a patch.
721 PH/24 Include config.h inside local_scan.h so that configuration settings are
724 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
725 that in an ACL for RCPT (for example), you can examine exactly what was
728 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
729 commands, but it was not correctly comparing the address with the actual
730 client host address. Thus, it would show the EHLO address in Received:
731 header lines when this was not necessary.
733 PH/27 Added the % operator to ${eval:}.
735 PH/28 Exim tries to create and chdir to its spool directory when it starts;
736 it should be ignoring failures (because with -C, for example, it has lost
737 privilege). It wasn't ignoring creation failures other than "already
740 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
743 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
744 because an input line was too long, either on its own, or by virtue of
745 too many continuations, the temporary file was not being removed, and the
746 return code was incorrect.
748 PH/31 Missing "BOOL" in function definition in filtertest.c.
750 PH/32 Applied Sieve patches from the maintainer.
752 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
754 PH/33 Added "verify = not_blind".
756 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
757 Local/Makefile (with some defaults set). These are used in built scripts
758 such as exicyclog, but they have never been used in the exim_install
759 script (though there are many overriding facilities there). I have
760 arranged that the exim_install script now takes note of these two
763 PH/35 Installed configuration files for Dragonfly.
765 PH/36 When a locally submitted message by a trusted user did not contain a
766 From: header, and the sender address was obtained from -f or from an SMTP
767 MAIL command, and the trusted user did not use -F to supply a sender
768 name, $originator_name was incorrectly used when constructing a From:
769 header. Furthermore, $originator_name was used for submission mode
770 messages from external hosts without From: headers in a similar way,
771 which is clearly wrong.
773 PH/37 Added control=suppress_local_fixups.
775 PH/38 When log_selector = +received_sender was set, and the addition of the
776 sender made the log line's construction buffer exactly full, or one byte
777 less than full, an overflow happened when the terminating "\n" was
780 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
781 when the result of a list match is failure because a DNS lookup failed.
783 PH/40 RM_COMMAND is now used in the building process.
785 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
786 the "build-* directories that it finds.
788 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
789 address in a domain literal was a prefix of an interface address.
791 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
792 when verifying a sender address, unless rewrite_headers is false.
794 PH/44 Wrote a long comment about why errors_to addresses are verified as
795 recipients, not senders.
797 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
798 the ratelimit ACL was added.
800 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
802 PH/46 Added extra information about PostgreSQL errors to the error string.
804 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
805 avoiding the need to install special zones in a real server. This is
806 backwards compatible; if it can't find the fake resolver, it drops back.
807 Thus, both old and new test suites can be run.
809 TF/02 Added util/ratelimit.pl
811 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
812 clock is set back in time.
814 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
815 Candler <B.Candler@pobox.com>.
817 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
818 OK for addresses that are the result of redirections.
820 PH/48 A number of further additions for the benefit of the new test suite,
821 including a fake gethostbyname() that interfaces to the fake DNS resolver
824 TF/06 The fix for widen_domains has also been applied to qualify_single and
825 search_parents which are the other dnslookup options that can cause
828 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
831 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
833 TF/07 Exim produced the error message "an SRV record indicated no SMTP
834 service" if it encountered an MX record with an empty target hostname.
835 The message is now "an MX or SRV record indicated no SMTP service".
837 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
838 if the DNS of the sending site is misconfigured. This is quite a
839 common situation. This change restores the behaviour of treating a
840 helo verification defer as a failure.
842 PH/51 If self=fail was set on a router, the bounce message did not include the
843 actual error message.
849 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
851 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
852 command itself ended in error, the underlying message about the transport
853 filter timeout was being overwritten with the pipe command error. Now the
854 underlying error message should be appended to the second error message.
856 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
858 PH/02 Reduce the amount of output that "make" produces by default. Full output
859 can still be requested.
861 PH/03 The warning log line about a condition test deferring for a "warn" verb
862 was being output only once per connection, rather than after each
863 occurrence (because it was using the same function as for successful
864 "warn" verbs). This seems wrong, so I have changed it.
866 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
867 it should not have, which might have caused a crash in the right
868 circumstances, but probably never did.
870 PH/04 Installed a modified version of Tony Finch's patch to make submission
871 mode fix the return path as well as the Sender: header line, and to
872 add a /name= option so that you can make the user's friendly name appear
875 TF/03 Added the control = fakedefer ACL modifier.
877 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
878 Mark Lowes for thorough testing.
880 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
882 TK/03 Merged latest SRS patch from Miles Wilton.
884 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
885 with the definition in sysexits.h (which is #included earlier).
886 Fortunately, Exim does not actually use EX_OK. The code used to try to
887 preserve the sysexits.h value, by assumimg that macro definitions were
888 scanned for macro replacements. I have been disabused of this notion,
889 so now the code just undefines EX_OK before #including unistd.h.
891 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
892 in the smtp transport. When a block could not be written in a single
893 write() function, the timeout was being re-applied to each part-write.
894 This seems wrong - if the receiver was accepting one byte at a time it
895 would take for ever. The timeout is now adjusted when this happens. It
896 doesn't have to be particularly precise.
898 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
899 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
901 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
902 without a domain if the check to <postmaster@domain> fails.
904 SC/01 Eximstats: added -xls and the ability to specify output files
905 (patch written by Frank Heydlauf).
907 SC/02 Eximstats: use FileHandles for outputing results.
909 SC/03 Eximstats: allow any combination of xls, txt, and html output.
911 SC/04 Eximstats: fixed display of large numbers with -nvr option
913 SC/05 Eximstats: fixed merging of reports with empty tables.
915 SC/06 Eximstats: added the -include_original_destination flag
917 SC/07 Eximstats: removed tabs and trailing whitespace.
919 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
921 TK/06 MBOX spool code: Add real "From " MBOX separator line
922 so the .eml file is really in mbox format (even though
923 most programs do not really care). Patch from Alex Miller.
925 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
926 The latter is generated from $received_to and is only set if the
927 message has one envelope recipient. SA can use these headers,
928 obviously out-of-the-box. Patch from Alex Miller.
930 PH/08 The ${def test on a variable was returning false if the variable's
931 value was "0", contrary to what the specification has always said!
932 The result should be true unless the variable is empty.
934 PH/09 The syntax error of a character other than { following "${if
935 def:variable_name" (after optional whitespace) was not being diagnosed.
936 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
937 accidental colon was present, for example, could give incorrect results.
939 PH/10 Tidied the code in a number of places where the st_size field of a stat()
940 result is used (not including appendfile, where other changes are about
943 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
944 This involved changing a lot of size variables from int to off_t. It
945 should work with maildirs and everything.
947 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
948 spamd dying while we are connected to it.
950 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
953 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
954 function for BDB 4.3.
956 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
958 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
959 the sources for PCRE have been re-arranged and more files are now
962 PH/15 The code I had for printing potentially long long variables in PH/11
963 above was not the best (it lost precision). The length of off_t variables
964 is now inspected at build time, and an appropriate printing format (%ld
965 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
966 to be "long long int" or "long int". This is needed for the internal
967 formatting function string_vformat().
969 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
970 the configuration file to be ":syslog", then the script "guesses" where
971 the logs files are, rather than using the compiled in default. In our
972 case the guess is not the same as the compiled default, so the script
973 suddenly stopped working when I started to use syslog. The patch checks
974 to see if log_file_path is "". If so, it attempts to read it from exim
975 with no configuration file to get the compiled in version, before it
976 falls back to the previous guessing code."
978 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
979 implementing BATV in an Exim configuration. See NewStuff for the gory
982 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
983 Makefile that are specific to HP-UX.
985 PH/18 If the "use_postmaster" option was set for a recipient callout together
986 with the "random" option, the postmaster address was used as the MAIL
987 FROM address for the random test, but not for the subsequent recipient
988 test. It is now used for both.
990 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
991 patch removes a few documentation additions to RFC 3028, because the
992 latest draft now contains them. It adds the new en;ascii-case comparator
993 and a new error check for 8bit text in MIME parts. Comparator and
994 require names are now matched exactly. I enabled the subaddress
995 extension, but it is not well tested yet (read: it works for me)."
997 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
998 rework some of the code of TK/09 above to avoid the hardwired use of
999 "%lld" and "long long". Replaced the call to snprintf() with a call to
1002 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1003 records point to non-existent hosts", "retry timeout exceeded", and
1004 "retry time not reached for any host after a long failure period".
1006 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1007 experimental DomainKeys support:
1009 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1010 (2) On an error such as an illegally used "control", the wrong name for
1011 the control was given.
1013 These problems did NOT occur unless DomainKeys support was compiled.
1015 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1017 PH/24 Added ${if match_ip condition.
1019 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1020 databases so that it will be absolutely obvious if a crash occurs in the
1021 DB library. This is a regular occurrence (often caused by mis-matched
1024 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1025 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1026 picked up on a user's system that detects such things. There doesn't seem
1027 to be a gcc warning option for this - only an attribute that has to be
1028 put on the function's prototype. It seems that in Fedora Core 4 they have
1029 set this on a number of new functions. No doubt there will be more in due
1032 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1033 specify a transport. However, if an address that was verified by such a
1034 router was the subject of a callout, Exim crashed because it tried to
1035 read the rcpt_include_affixes from the non-existent transport. Now it
1036 just assumes that the setting of that option is false. This bug was
1037 introduced by 4.51/PH/31.
1039 PH/28 Changed -d+all to exclude +memory, because that information is very
1040 rarely of interest, but it makes the output a lot bigger. People tend to
1041 do -d+all out of habit.
1043 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1044 code in os-type was giving problems when libc.so lives in lib64, like on
1047 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1048 aren't the modern standard, and it seems that some systems' include files
1049 don't always have them. Exim was already checking for some of the newer
1050 ones like T_AAAA, and defining it itself. I've added checks for all the
1051 record types that Exim uses.
1053 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1054 not automatically generating a new one, as it is supposed to. This
1055 prevented TLS from working. If the file did exist, but contained invalid
1056 data, a new version was generated, as expected. It was only the case of a
1057 non-existent file that was broken.
1059 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1060 with a change in libdomainkeys > 0.64.
1062 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1063 from DNS. If the selector record carries the flag, it now has
1064 precedence over the domain-wide flag.
1066 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1068 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1069 the use of a port name that isn't defined in /etc/services) occurred, the
1070 message was deferred as in a normal delivery, and thus remained on the
1071 spool, instead of being failed because of the mua_wrapper setting. This
1072 is now fixed, and I tidied up some of the mua_wrapper messages at the
1075 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1076 the messages in a hash of arrays rather than using individual hashes.
1077 This is a bit cleaner and results in dramatic memory savings, albeit
1078 at a slight CPU cost.
1080 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1081 as requested by Marc Sherman.
1083 SC/10 Eximstats: added histograms for user specified patterns as requested
1086 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1088 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1089 fopen() in the content-scanning modules that did not already have it.
1095 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1096 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1098 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1100 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1102 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1103 to test Sieve filters that use "vacation".
1105 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1106 that changes the way the GnuTLS parameters are stored in the cache file.
1107 The new format can be generated externally. For backward compatibility,
1108 if the data in the cache doesn't make sense, Exim assumes it has read an
1109 old-format file, and it generates new data and writes a new file. This
1110 means that you can't go back to an older release without removing the
1113 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1114 work if there are any delivery delays because "one_time" forces the
1115 parent to be marked "delivered", so its unseen clone is never tried
1116 again. For this reason, Exim now forbids the simultaneous setting of
1119 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1120 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1121 entry if you want to know the details. The fix, however, neglected to
1122 consider the case when local delivery batching is involved. The test for
1123 "previously delivered" was not happening when checking to see if an
1124 address could be batched with a previous (undelivered) one; under
1125 certain circumstances this could lead to multiple deliveries to the same
1128 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1129 in its include files, and this causes problems building Exim.
1131 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1132 header_syntax) but Exim was just ignoring anything given after a slash.
1133 In particular, this caused confusion with an attempt to use "verify =
1134 reverse_host_lookup/defer_ok". An error is now given when options are
1135 supplied for verify items that do not have them. (Maybe reverse_host_
1136 lookup should have a defer_ok option, but that's a different point.)
1138 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1139 defined by RFC 821) to 2048, because there were problems with some AUTH
1140 commands, and RFC 1869 says the size should be increased for extended
1141 SMTP commands that take arguments.
1143 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1146 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1147 "unknown" error; now it says that the functionality isn't in the binary.
1149 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1150 an address' error message when a string expansion fails (syntax or
1151 whatever). Otherwise the password may appear in the log. Following change
1152 PH/42 below, there is no longer a chance of it appearing in a bounce
1155 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1157 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1158 of its addresses was ever tried. (Bugzilla bug #2.)
1160 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1161 the result incorrectly in the debug output. (It correctly added a newline
1162 to what was transported.)
1164 TF/01 Added $received_time.
1166 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1167 commented out examples of how to interface to a virus scanner and to
1168 SpamAssassin. Also added commented examples of av_scanner and
1169 spamd_address settings.
1171 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1172 and controls are allowed in which ACLs. There were a couple of minor
1173 errors. Some of the entries in the conditions table (which is a table of
1174 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1175 negation of where the condition IS allowed.
1177 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1179 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1180 header file does not have a version number, so I've had to invent a new
1181 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1182 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1183 radiusclient), but it was contributed by a Radius user.
1185 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1186 files or directories, for OpenSSL.
1188 PH/19 When an Exim process that is running as root has to create an Exim log
1189 file, it does so in a subprocess that runs as exim:exim so as to get the
1190 ownership right at creation (otherwise, other Exim processes might see
1191 the file with the wrong ownership). There was no test for failure of this
1192 fork() call, which would lead to the process getting stuck as it waited
1193 for a non-existent subprocess. Forks do occasionally fail when resources
1194 run out. I reviewed all the other calls to fork(); they all seem to check
1197 PH/20 When checking for unexpected SMTP input at connect time (before writing
1198 the banner), Exim was not dealing correctly with a non-positive return
1199 from the read() function. If the client had disconnected by this time,
1200 the result was a log entry for a synchronization error with an empty
1201 string after "input=" when read() returned zero. If read() returned -1
1202 (an event I could not check), uninitialized data bytes were printed.
1203 There were reports of junk text (parts of files, etc) appearing after
1206 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1208 PH/22 Added support for macro redefinition, and (re)definition in between
1209 driver and ACL definitions.
1211 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1212 forgetting to use the resulting value; it was using the unexpanded value.
1214 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1215 hadn't been configured. The fix is from Juergen Kreileder, who
1216 understands it better than I do:
1218 "Here's what I see happening with three configured cyrus_sasl
1219 authenticators configured (plain, login, cram-md5):
1221 On startup auth_cyrus_sasl_init() gets called for each of these.
1222 This means three calls to sasl_listmech() without a specified mech_list.
1223 => SASL tests which mechs of all available mechs actually work
1224 => three warnings about OTP not working
1225 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1227 With the patch, sasl_listmech() also gets called three times. But now
1228 SASL's mech_list option is set to the server_mech specified in the the
1229 authenticator. Or in other words, the answer from sasl_listmech()
1230 gets limited to just the mech you're testing for (which is different
1232 => the return list contains just 'plain' or 'login', 'cram-md5' or
1233 nothing depending on the value of ob->server_mech.
1235 I've just tested the patch: Authentication still works fine,
1236 unavailable mechs specified in the exim configuration are still
1237 caught, and the auth.log warnings about OTP are gone."
1239 PH/25 When debugging is enabled, the contents of the command line are added
1240 to the debugging output, even when log_selector=+arguments is not
1243 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1244 answer is "GNU", and only if the return is "GNU/something" is the answer
1247 PH/27 $acl_verify_message is now set immediately after the failure of a
1248 verification in an ACL, and so is available in subsequent modifiers. In
1249 particular, the message can be preserved by coding like this:
1251 warn !verify = sender
1252 set acl_m0 = $acl_verify_message
1254 Previously, $acl_verify_message was set only while expanding "message"
1255 and "log_message" when a very denied access.
1257 PH/28 Modified OS/os.c-Linux with
1259 -#ifndef OS_LOAD_AVERAGE
1260 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1262 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1263 nomenclature these days.)
1265 PH/29 Installed patch from the Sieve maintainer that adds the options
1266 sieve_useraddress and sieve_subaddress to the redirect router.
1268 PH/30 In these circumstances:
1269 . Two addresses routed to the same list of hosts;
1270 . First host does not offer TLS;
1271 . First host accepts first address;
1272 . First host gives temporary error to second address;
1273 . Second host offers TLS and a TLS session is established;
1274 . Second host accepts second address.
1275 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1276 and peerdn, if requested) that were in fact used only for the second
1279 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1280 attention to any local part prefix or suffix that was matched by the
1281 router that accepted the address. It now behaves in the same way as it
1282 does for delivery: the affixes are removed from the local part unless
1283 rcpt_include_affixes is set on the transport.
1285 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1286 timeout during the DATA phase of an incoming message.
1288 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1289 applied a patch sanctioned by the Sieve maintainer.
1291 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1292 the uid or gid is negative. A case of a negative gid caused this to be
1293 noticed. The fix allows for either to be negative.
1295 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1296 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1299 PH/36 The change PH/12 above was broken. Fixed it.
1301 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1302 the grounds that routing the same address twice would always produce the
1303 same answer. This might have been true once, but it is certainly no
1304 longer true now. Routing a child address may depend on the previous
1305 routing that produced that child. Some complicated redirection strategies
1306 went wrong when messages had multiple recipients, and made Exim's
1307 behaviour dependent on the order in which the addresses were given.
1309 I have moved the duplicate checking until after the routing is complete.
1310 Exim scans the addresses that are assigned to local and remote
1311 transports, and removes any duplicates. This means that more work will be
1312 done, as duplicates will always all be routed, but duplicates are
1313 presumably rare, so I don't expect this is of any significance.
1315 For deliveries to pipes, files, and autoreplies, the duplicate checking
1316 still happens during the routing process, since they are not going to be
1319 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1320 It corrects a timeout issue with spamd. This is Ian's comment: "The
1321 background is that sometimes spamd either never reads data from a
1322 connection it has accepted, or it never writes response data. The exiscan
1323 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1324 blindly assumes that writes won't block so it may never time out."
1326 PH/39 Allow G after quota size as well as K and M.
1328 PH/40 The value set for $authenticated_id in an authenticator may not contain
1329 binary zeroes or newlines because the value is written to log lines and
1330 to spool files. There was no check on this. Now the value is run through
1331 the string_printing() function so that such characters are converted to
1332 printable escape sequences.
1334 PH/41 $message_linecount is a new variable that contains the total number of
1335 lines in the message. Compare $body_linecount, which is the count for the
1338 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1339 bounce and delay warning messages, except in certain special cases, which
1342 (a) An SMTP error message from a remote host;
1343 (b) A message specified in a :fail: redirection;
1344 (c) A message specified in a "fail" command in a system filter;
1345 (d) A message specified in a FAIL return from the queryprogram router;
1346 (e) A message specified by the cannot_route_message router option.
1348 In these cases only, Exim does include the error details in bounce and
1349 warning messages. There are also a few cases where bland messages such
1350 as "unrouteable address" or "local delivery error" are given.
1352 PH/43 $value is now also set for the "else" part of a ${run expansion.
1354 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1355 being worked on, but at least Exim now implements the latest version to
1358 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1359 process to complete was treated as a delivery failure, a timeout while
1360 writing the message to the pipe was logged, but erroneously treated as a
1361 successful delivery. Such timeouts include transport filter timeouts. For
1362 consistency with the overall process timeout, these timeouts are now
1363 treated as errors, giving rise to delivery failures by default. However,
1364 there is now a new Boolean option for the pipe transport called
1365 timeout_defer, which, if set TRUE, converts the failures into defers for
1366 both kinds of timeout. A transport filter timeout is now identified in
1369 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1370 systems where "make" and "gmake" are different, calling "gmake" at top
1371 level broke things. I've arranged for the value of $(MAKE) to be passed
1372 from the Makefile to this script so that it can call the same version of
1376 A note about Exim versions 4.44 and 4.50
1377 ----------------------------------------
1379 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1380 changes of various kinds. As a consequence, a big documentation update was
1381 needed. This delayed the release for rather longer than seemed good, especially
1382 in the light of a couple of (minor) security issues. Therefore, the changes
1383 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1384 release. So 4.44 and 4.50 are in effect two different branches that both start
1387 I have left the 4.50 change log unchanged; it contains all the changes since
1388 4.43. The change log for 4.44 is below; many of its items are identical to
1389 those for 4.50. This seems to be the most sensible way to preserve the
1390 historical information.
1396 1. Minor wording change to the doc/README.SIEVE file.
1398 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1399 computation of the current number of files was incorrect.
1401 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1402 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1405 4. Give more explanation in the error message when the command for a transport
1406 filter fails to execute.
1408 5. There are several places where Exim runs a non-Exim command in a
1409 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1410 was being done only for the command run by the queryprogram router. It is
1411 now done for all such subprocesses. The other cases are: ${run, transport
1412 filters, and the commands run by the lmtp and pipe transports.
1414 6. Added CONFIGURE_GROUP build-time option.
1416 7. Some older OS have a limit of 256 on the maximum number of file
1417 descriptors. Exim was using setrlimit() to set 1000 as a large value
1418 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1419 systems. I've change it so that if it can't get 1000, it tries for 256.
1421 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1422 was an oversight, and furthermore, ever since the addition of extra
1423 controls (e.g. 4.43/32), the checks on when to allow different forms of
1424 "control" were broken. There should now be diagnostics for all cases when a
1425 control that does not make sense is encountered.
1427 9. Added the /retain_sender option to "control=submission".
1429 10. $recipients is now available in the predata ACL (oversight).
1431 11. Tidy the search cache before the fork to do a delivery from a message
1432 received from the command line. Otherwise the child will trigger a lookup
1433 failure and thereby defer the delivery if it tries to use (for example) a
1434 cached ldap connection that the parent has called unbind on.
1436 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1437 of $address_data from the recipient verification was clobbered by the
1438 sender verification.
1440 13. The value of address_data from a sender verification is now available in
1441 $sender_address_data in subsequent conditions in the ACL statement.
1443 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1445 15. Added a new option "connect=<time>" to callout options, to set a different
1448 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1449 was its contents. (It was OK if the option was not defined at all.)
1451 17. A "Completed" log line is now written for messages that are removed from
1452 the spool by the -Mrm option.
1454 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1455 information about exactly what failed.
1457 19. Added -dd to debug only the daemon process.
1459 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1460 handles timeouts, both on the server side and network timeouts. Renamed the
1461 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1463 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1464 It is now set to "smtps".
1466 22. $host_address is now set to the target address during the checking of
1467 ignore_target_hosts.
1469 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1470 being passed; this would have caused $sender_host_name to have been used if
1471 matching the list had actually called for a host name (not very likely,
1472 since this list is usually IP addresses). A host name is now passed as
1475 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1476 code that specifies a non-restarting handler (typically sigaction() in
1477 modern systems) in an attempt to fix a rare and obscure crash bug.
1479 25. Narrowed the window for a race in the daemon that could cause it to ignore
1480 SIGCHLD signals. This is not a major problem, because they are used only to
1481 wake it up if nothing else does.
1483 26. A malformed maildirsize file could cause Exim to calculate negative values
1484 for the mailbox size or file count. Odd effects could occur as a result.
1485 The maildirsize information is now recalculated if the size or filecount
1488 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1489 support for a long time. Removed HAVE_SYS_VFS_H.
1491 28. Installed the latest version of exipick from John Jetmore.
1493 29. In an address list, if the pattern was not a regular expression, an empty
1494 subject address (from a bounce message) matched only if the pattern was an
1495 empty string. Non-empty patterns were not even tested. This was the wrong
1496 because it is perfectly reasonable to use an empty address as part of a
1497 database query. An empty address is now tested by patterns that are
1498 lookups. However, all the other forms of pattern expect the subject to
1499 contain a local part and a domain, and therefore, for them, an empty
1500 address still always fails if the pattern is not itself empty.
1502 30. Exim went into a mad DNS loop when attempting to do a callout where the
1503 host was specified on an smtp transport, and looking it up yielded more
1504 than one IP address.
1506 31. Re-factored the code for checking spool and log partition space into a
1507 function that finds that data and another that does the check. The former
1508 is then used to implement four new variables: $spool_space, $log_space,
1509 $spool_inodes, and $log_inodes.
1511 32. The RFC2047 encoding function was originally intended for short strings
1512 such as real names; it was not keeping to the 75-character limit for
1513 encoded words that the RFC imposes. It now respects the limit, and
1514 generates multiple encoded words if necessary. To be on the safe side, I
1515 have increased the buffer size for the ${rfc2047: expansion operator from
1518 33. It is now permitted to omit both strings after an "if" condition; if the
1519 condition is true, the result is "true". As before, when the second string
1520 is omitted, a false condition yields an empty string. This makes it less
1521 cumbersome to write custom ACL and router conditions.
1523 34. Failure to deliver a bounce message always caused it to be frozen, even if
1524 there was an errors_to setting on the router. The errors_to setting is now
1527 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1528 canonical form (fully expanded) before being placed in
1529 $sender_host_address.
1531 36. The table in the code that translates DNS record types into text (T_A to
1532 "A" for instance) was missing entries for NS and CNAME. It is just possible
1533 that this could have caused confusion if both these types were looked up
1534 for the same domain, because the text type is used as part of Exim's
1535 per-process caching. But the chance of anyone hitting this buglet seems
1538 37. The dnsdb lookup has been extended in a number of ways.
1540 (1) There is a new type, "zns", which walks up the domain tree until it
1541 finds some nameserver records. It should be used with care.
1543 (2) There is a new type, "mxh", which is like "mx" except that it returns
1544 just the host names, not the priorities.
1546 (3) It is now possible to give a list of domains (or IP addresses) to be
1547 looked up. The behaviour when one of the lookups defers can be
1548 controlled by a keyword.
1550 (4) It is now possible to specify the separator character for use when
1551 multiple records are returned.
1553 38. The dnslists ACL condition has been extended: it is now possible to supply
1554 a list of IP addresses and/or domains to be looked up in a particular DNS
1557 39. Added log_selector=+queue_time_overall.
1559 40. When running the queue in the test harness, wait just a tad after forking a
1560 delivery process, to get repeatability of debugging output.
1562 41. Include certificate and key file names in error message when GnuTLS fails
1563 to set them up, because the GnuTLS error message doesn't include the name
1564 of the failing file when there is a problem reading it.
1566 42. Allow both -bf and -bF in the same test run.
1568 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1570 44. The "Exiscan patch" is now merged into the mainline Exim source.
1572 45. Sometimes the final signoff response after QUIT could fail to get
1573 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1574 < 0 before doing a fflush(). This bug looks as though it goes back to the
1575 introduction of TLS in release 3.20, but "sometimes" must have been rare
1576 because the tests only now provoked it.
1578 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1579 (this can affect the format of dates).
1581 47. exim_tidydb, when checking for the continued existence of a message for
1582 which it has found a message-specific retry record, was not finding
1583 messages that were in split spool directories. Consequently, it was
1584 deleting retry records that should have stayed in existence.
1586 48. Steve fixed some bugs in eximstats.
1588 49. The SPA authentication driver was not abandoning authentication and moving
1589 on to the next authenticator when an expansion was forced to fail,
1590 contradicting the general specification for all authenticators. Instead it
1591 was generating a temporary error. It now behaves as specified.
1593 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1594 (the order specifies the preference for clients). The order is now AES256,
1595 AES128, 3DES, ARCFOUR128.
1597 51. Small patch to Sieve code - explicitly set From: when generating an
1600 52. Exim crashed if a remote delivery caused a very long error message to be
1601 recorded - for instance if somebody sent an entire SpamAssassin report back
1602 as a large number of 550 error lines. This bug was coincidentally fixed by
1603 increasing the size of one of Exim's internal buffers (big_buffer) that
1604 happened as part of the Exiscan merge. However, to be on the safe side, I
1605 have made the code more robust (and fixed the comments that describe what
1608 53. Now that there can be additional text after "Completed" in log lines (if
1609 the queue_time_overall log selector is set), a one-byte patch to exigrep
1610 was needed to allow it to recognize "Completed" as not the last thing in
1613 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1614 patch that reportedly fixes this has been added. I am not expert enough to
1615 create a test for it. This is what the patch creator wrote:
1617 "I found a little strange behaviour of ldap code when working with
1618 Windows 2003 AD Domain, where users was placed in more than one
1619 Organization Units. When I tried to give exim partial DN, the exit code
1620 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1621 But simultaneously result of request was absolutely normal ldap result,
1622 so I produce this patch..."
1624 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1625 REFERENCE, so I have modified the code to exclude the patch when that macro
1628 55. Some experimental protocols are using DNS PTR records for new purposes. The
1629 keys for these records are domain names, not reversed IP addresses. The
1630 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1631 leaves it alone. Component reversal etc. now happens only for IP addresses.
1634 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1636 57. Double the size of the debug message buffer (to 2048) so that more of very
1637 long debug lines gets shown.
1639 58. The exicyclog utility now does better if the number of log files to keep
1640 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1642 59. Two changes related to the smtp_active_hostname option:
1644 (1) $smtp_active_hostname is now available as a variable.
1645 (2) The default for smtp_banner uses $smtp_active_hostname instead
1646 of $primary_hostname.
1648 60. The host_aton() function is supposed to be passed a string that is known
1649 to be a valid IP address. However, in the case of IPv6 addresses, it was
1650 not checking this. This is a hostage to fortune. Exim now panics and dies
1651 if the condition is not met. A case was found where this could be provoked
1652 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1653 components; fortuitously, this particular loophole had already been fixed
1654 by change 4.50/55 above.
1656 If there are any other similar loopholes, the new check in host_aton()
1657 itself should stop them being exploited. The report I received stated that
1658 data on the command line could provoke the exploit when Exim was running as
1659 exim, but did not say which command line option was involved. All I could
1660 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1661 running as the user.
1664 61. There was a buffer overflow vulnerability in the SPA authentication code
1665 (which came originally from the Samba project). I have added a test to the
1666 spa_base64_to_bits() function which I hope fixes it.
1669 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1670 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1672 63. The daemon start-up calls getloadavg() while still root for those OS that
1673 need the first call to be done as root, but it missed one case: when
1674 deliver_queue_load_max is set with deliver_drop_privilege. This is
1675 necessary for the benefit of the queue runner, because there is no re-exec
1676 when deliver_drop_privilege is set.
1678 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1679 This has been fixed.
1681 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1682 was in use, was not putting the data itself into the right store pool;
1683 consequently, it could be overwritten for a subsequent message in the same
1684 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1687 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1689 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1690 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1691 the function were treating the return as a boolean value, which happened to
1692 work because 0=false and not-0=true, but is not correct code.
1694 68. The host_aton() function was not handling scoped IPv6 addresses (those
1695 with, for example, "%eth0" on the end) correctly.
1697 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1698 negated items (that is, ~something) in unsigned ints. Some compilers
1699 apparently mutter when there is no cast.
1701 70. If an address verification called from an ACL failed, and did not produce a
1702 user-specific message (i.e. there was only a "system" message), nothing was
1703 put in $acl_verify_message. In this situation, it now puts the system
1706 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1707 session; change 4.31/43 added the unwanted input to the log line - except
1708 that it did not do this in the start of session case. It now does.
1710 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1711 This is wrong and can cause the other end to generate a synchronization
1712 error if it is another Exim or anything else that does the synchronization
1713 check. A QUIT command is no longer sent after a timeout.
1715 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1716 during host lookups.
1718 74. The defer_ok option of callout verification was not working if it was used
1719 when verifying addresses in header lines, that is, for this case:
1721 verify = header_sender/callout=defer_ok
1723 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1724 those file descriptors could be used for SMTP connections. If anything
1725 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1726 could be sent to the SMTP client, causing chaos. The daemon now opens
1727 stdin, stdout, and stderr to /dev/null when it puts itself into the
1730 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1731 log by default. The user can override this with suitable Perl magic.
1733 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1734 the log message when discard triggers, was not working for the DATA ACL or
1735 for the non-SMTP ACL.
1737 78. Error message wording change in sieve.c.
1739 79. If smtp_accept_max_per_host was set, the number of connections could be
1740 restricted to fewer than expected, because the daemon was trying to set up
1741 a new connection before checking whether the processes handling previous
1742 connections had finished. The check for completed processes is now done
1743 earlier. On busy systems, this bug wouldn't be noticed because something
1744 else would have woken the daemon, and it would have reaped the completed
1747 80. If a message was submitted locally by a user whose login name contained one
1748 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1749 It caused a spool format error. I have fixed the spool reading code. A
1750 related problem was that the "from" clause in the Received: line became
1751 illegal because of the space(s). It is now covered by ${quote_local_part.
1753 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1756 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1758 83. Patch from Sieve maintainer to fix unterminated string problem in
1759 "vacation" handling.
1761 84. Some minor changes to the Linux configuration files to help with other
1762 OS variants using glibc.
1764 85. One more patch for Sieve to update vacation handling to latest spec.
1767 ----------------------------------------------------
1768 See the note above about the 4.44 and 4.50 releases.
1769 ----------------------------------------------------
1775 1. Change 4.43/35 introduced a bug that caused file counts to be
1776 incorrectly computed when quota_filecount was set in an appendfile
1779 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1780 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1783 3. Give more explanation in the error message when the command for a transport
1784 filter fails to execute.
1786 4. There are several places where Exim runs a non-Exim command in a
1787 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1788 was being done only for the command run by the queryprogram router. It is
1789 now done for all such subprocesses. The other cases are: ${run, transport
1790 filters, and the commands run by the lmtp and pipe transports.
1792 5. Some older OS have a limit of 256 on the maximum number of file
1793 descriptors. Exim was using setrlimit() to set 1000 as a large value
1794 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1795 systems. I've change it so that if it can't get 1000, it tries for 256.
1797 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1798 was an oversight, and furthermore, ever since the addition of extra
1799 controls (e.g. 4.43/32), the checks on when to allow different forms of
1800 "control" were broken. There should now be diagnostics for all cases when a
1801 control that does not make sense is encountered.
1803 7. $recipients is now available in the predata ACL (oversight).
1805 8. Tidy the search cache before the fork to do a delivery from a message
1806 received from the command line. Otherwise the child will trigger a lookup
1807 failure and thereby defer the delivery if it tries to use (for example) a
1808 cached ldap connection that the parent has called unbind on.
1810 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1811 of $address_data from the recipient verification was clobbered by the
1812 sender verification.
1814 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1815 was its contents. (It was OK if the option was not defined at all.)
1817 11. A "Completed" log line is now written for messages that are removed from
1818 the spool by the -Mrm option.
1820 12. $host_address is now set to the target address during the checking of
1821 ignore_target_hosts.
1823 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1824 being passed; this would have caused $sender_host_name to have been used if
1825 matching the list had actually called for a host name (not very likely,
1826 since this list is usually IP addresses). A host name is now passed as
1829 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1830 code that specifies a non-restarting handler (typically sigaction() in
1831 modern systems) in an attempt to fix a rare and obscure crash bug.
1833 15. Narrowed the window for a race in the daemon that could cause it to ignore
1834 SIGCHLD signals. This is not a major problem, because they are used only to
1835 wake it up if nothing else does.
1837 16. A malformed maildirsize file could cause Exim to calculate negative values
1838 for the mailbox size or file count. Odd effects could occur as a result.
1839 The maildirsize information is now recalculated if the size or filecount
1842 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1843 support for a long time. Removed HAVE_SYS_VFS_H.
1845 18. Updated exipick to current release from John Jetmore.
1847 19. Allow an empty sender to be matched against a lookup in an address list.
1848 Previously the only cases considered were a regular expression, or an
1851 20. Exim went into a mad DNS lookup loop when doing a callout where the
1852 host was specified on the transport, if the DNS lookup yielded more than
1855 21. The RFC2047 encoding function was originally intended for short strings
1856 such as real names; it was not keeping to the 75-character limit for
1857 encoded words that the RFC imposes. It now respects the limit, and
1858 generates multiple encoded words if necessary. To be on the safe side, I
1859 have increased the buffer size for the ${rfc2047: expansion operator from
1862 22. Failure to deliver a bounce message always caused it to be frozen, even if
1863 there was an errors_to setting on the router. The errors_to setting is now
1866 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1867 canonical form (fully expanded) before being placed in
1868 $sender_host_address.
1870 24. Updated eximstats to version 1.33
1872 25. Include certificate and key file names in error message when GnuTLS fails
1873 to set them up, because the GnuTLS error message doesn't include the name
1874 of the failing file when there is a problem reading it.
1876 26. Expand error message when OpenSSL has problems setting up cert/key files.
1879 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1880 (this can affect the format of dates).
1882 28. exim_tidydb, when checking for the continued existence of a message for
1883 which it has found a message-specific retry record, was not finding
1884 messages that were in split spool directories. Consequently, it was
1885 deleting retry records that should have stayed in existence.
1887 29. eximstats updated to version 1.35
1888 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1889 1.35 - bugfix such that pie charts by volume are generated correctly
1891 30. The SPA authentication driver was not abandoning authentication and moving
1892 on to the next authenticator when an expansion was forced to fail,
1893 contradicting the general specification for all authenticators. Instead it
1894 was generating a temporary error. It now behaves as specified.
1896 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1897 (the order specifies the preference for clients). The order is now AES256,
1898 AES128, 3DES, ARCFOUR128.
1900 31. Small patch to Sieve code - explicitly set From: when generating an
1903 32. Exim crashed if a remote delivery caused a very long error message to be
1904 recorded - for instance if somebody sent an entire SpamAssassin report back
1905 as a large number of 550 error lines. This bug was coincidentally fixed by
1906 increasing the size of one of Exim's internal buffers (big_buffer) that
1907 happened as part of the Exiscan merge. However, to be on the safe side, I
1908 have made the code more robust (and fixed the comments that describe what
1911 33. Some experimental protocols are using DNS PTR records for new purposes. The
1912 keys for these records are domain names, not reversed IP addresses. The
1913 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1914 leaves it alone. Component reversal etc. now happens only for IP addresses.
1917 34. The host_aton() function is supposed to be passed a string that is known
1918 to be a valid IP address. However, in the case of IPv6 addresses, it was
1919 not checking this. This is a hostage to fortune. Exim now panics and dies
1920 if the condition is not met. A case was found where this could be provoked
1921 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1922 components; fortuitously, this particular loophole had already been fixed
1923 by change 4.50/55 or 4.44/33 above.
1925 If there are any other similar loopholes, the new check in host_aton()
1926 itself should stop them being exploited. The report I received stated that
1927 data on the command line could provoke the exploit when Exim was running as
1928 exim, but did not say which command line option was involved. All I could
1929 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1930 running as the user.
1933 35. There was a buffer overflow vulnerability in the SPA authentication code
1934 (which came originally from the Samba project). I have added a test to the
1935 spa_base64_to_bits() function which I hope fixes it.
1938 36. The daemon start-up calls getloadavg() while still root for those OS that
1939 need the first call to be done as root, but it missed one case: when
1940 deliver_queue_load_max is set with deliver_drop_privilege. This is
1941 necessary for the benefit of the queue runner, because there is no re-exec
1942 when deliver_drop_privilege is set.
1944 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1945 was in use, was not putting the data itself into the right store pool;
1946 consequently, it could be overwritten for a subsequent message in the same
1947 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1950 38. Sometimes the final signoff response after QUIT could fail to get
1951 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1952 < 0 before doing a fflush(). This bug looks as though it goes back to the
1953 introduction of TLS in release 3.20, but "sometimes" must have been rare
1954 because the tests only now provoked it.
1960 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1961 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1962 Now it returns an integer. A number of calls were still expecting a T/F
1963 return. Fortuitously, in all cases, the tests worked in OK situations,
1964 which is the norm. However, things would have gone wrong on any write
1965 failures on the smtp file descriptor. This function is used when sending
1966 messages over SMTP and also when doing verify callouts.
1968 2. When Exim is called to do synchronous delivery of a locally submitted
1969 message (the -odf or -odi options), it no longer closes stderr before doing
1972 3. Implemented the mua_wrapper option.
1974 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1976 5. Implemented the functions header_remove(), header_testname(),
1977 header_add_at_position(), and receive_remove_recipient(), and exported them
1980 6. If an ACL "warn" statement specified the addition of headers, Exim already
1981 inserted X-ACL-Warn: at the start if there was no header name. However, it
1982 was not making this test for the second and subsequent header lines if
1983 there were newlines in the string. This meant that an invalid header could
1984 be inserted if Exim was badly configured.
1986 7. Allow an ACL "warn" statement to add header lines at the start or after all
1987 the Received: headers, as well as at the end.
1989 8. Added the rcpt_4xx retry error code.
1991 9. Added postmaster_mailfrom=xxx to callout verification option.
1993 10. Added mailfrom=xxxx to the callout verification option, for verify=
1996 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1997 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1999 12. Inserted some casts to stop certain compilers warning when using pointer
2000 differences as field lengths or precisions in printf-type calls (mostly
2001 affecting debugging statements).
2003 13. Added optional readline() support for -be (dynamically loaded).
2005 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2006 same clock tick as a message's arrival, so that its received time was the
2007 same as the "first fail" time on the retry record, and that message
2008 remained on the queue past the ultimate address timeout, every queue runner
2009 would try a delivery (because it was past the ultimate address timeout) but
2010 after another failure, the ultimate address timeout, which should have then
2011 bounced the address, did not kick in. This was a "< instead of <=" error;
2012 in most cases the first failure would have been in the next clock tick
2013 after the received time, and all would be well.
2015 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2016 being recognized when the domain list was tested by the match_domain
2017 condition in an expansion string.
2019 16. Added the ${str2b64: operator.
2021 17. Exim was always calling setrlimit() to set a large limit for the number of
2022 processes, without checking whether the existing limit was already
2023 adequate. (It did check for the limit on file descriptors.) Furthermore,
2024 errors from getrlimit() and setrlimit() were being ignored. Now they are
2025 logged to the main and panic logs, but Exim does carry on, to try to do its
2026 job under whatever limits there are.
2028 18. Imported PCRE 5.0.
2030 19. Trivial typo in log message " temporarily refused connection" (the leading
2033 20. If the log selector return_path_on_delivery was set and an address was
2034 redirected to /dev/null, the delivery process crashed because it assumed
2035 that a return path would always be set for a "successful" delivery. In this
2036 case, the whole delivery is bypassed as an optimization, and therefore no
2039 21. Internal re-arrangement: the function for sending a challenge and reading
2040 a response while authentication was assuming a zero-terminated challenge
2041 string. It's now changed to take a pointer and a length, to allow for
2042 binary data in such strings.
2044 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2046 23. Exim was not respecting finduser_retries when seeking the login of the
2047 uid under which it was called; it was always trying 10 times. (The default
2048 setting of finduser_retries is zero.) Also, it was sleeping after the final
2049 failure, which is pointless.
2051 24. Implemented tls_on_connect_ports.
2053 25. Implemented acl_smtp_predata.
2055 26. If the domain in control=submission is set empty, Exim assumes that the
2056 authenticated id is a complete email address when it generates From: or
2057 Sender: header lines.
2059 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2060 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2061 chown and chgrp in /bin and hostname in /usr/bin.
2063 28. Exim was keeping the "process log" file open after each use, just as it
2064 does for the main log. This opens the possibility of it remaining open for
2065 long periods when the USR1 signal hits a daemon. Occasional processlog
2066 errors were reported, that could have been caused by this. Anyway, it seems
2067 much more sensible not to leave this file open at all, so that is what now
2070 29. The long-running daemon process does not normally write to the log once it
2071 has entered its main loop, and it closes the log before doing so. This is
2072 so that log files can straightforwardly be renamed and moved. However,
2073 there are a couple of unusual error situations where the daemon does write
2074 log entries, and I had neglected to close the log afterwards.
2076 30. The text of an SMTP error response that was received during a remote
2077 delivery was being truncated at 512 bytes. This is too short for some of
2078 the long messages that one sometimes sees. I've increased the limit to
2081 31. It is now possible to make retry rules that apply only when a message has a
2082 specific sender, in particular, an empty sender.
2084 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2085 it possible to be selective about when SMTP synchronization is enforced.
2087 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2089 32. Implemented hosts_connection_nolog.
2091 33. Added an ACL for QUIT.
2093 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2096 35. Added mailbox_size and mailbox_filecount to appendfile.
2098 36. Added control = no_multiline_responses to ACLs.
2100 37. There was a bug in the logic of the code that waits for the clock to tick
2101 in the case where the clock went backwards by a substantial amount such
2102 that the microsecond fraction of "now" was more than the microsecond
2103 fraction of "then" (but the whole seconds number was less).
2105 38. Added support for the libradius Radius client library this is found on
2106 FreeBSD (previously only the radiusclient library was supported).
2112 1. When certain lookups returned multiple values in the form name=value, the
2113 quoting of the values was not always being done properly. Specifically:
2114 (a) If the value started with a double quote, but contained no whitespace,
2116 (b) If the value contained whitespace other than a space character (i.e.
2117 tabs or newlines or carriage returns) it was not quoted.
2118 This fix has been applied to the mysql and pgsql lookups by writing a
2119 separate quoting function and calling it from the lookup code. The fix
2120 should probably also be applied to nisplus, ibase and oracle lookups, but
2121 since I cannot test any of those, I have not disturbed their existing code.
2123 2. A hit in the callout cache for a specific address caused a log line with no
2124 reason for rejecting RCPT. Now it says "Previous (cached) callout
2125 verification failure".
2127 3. There was an off-by-one bug in the queryprogram router. An over-long
2128 return line was truncated at 256 instead of 255 characters, thereby
2129 overflowing its buffer with the terminating zero. As well as fixing this, I
2130 have increased the buffer size to 1024 (and made a note to document this).
2132 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2133 when Exim is waiting for an SMTP response from a remote server, Exim
2134 restarts its select() call on the socket, thereby resetting its timeout.
2135 This is not a problem when such interrupts are rare. Somebody set up a cron
2136 job to run exiwhat every 2 minutes, which is less than the normal select()
2137 timeout (5 or 10 minutes). This meant that the select() timeout never
2138 kicked in because it was always reset. I have fixed this by comparing the
2139 time when an interrupt arrives with the time at the start of the first call
2140 to select(). If more time than the timeout has elapsed, the interrupt is
2141 treated as a timeout.
2143 5. Some internal re-factoring in preparation for the addition of Sieve
2144 extensions (by MH). In particular, the "personal" test is moved to a
2145 separate function, and given an option for scanning Cc: and Bcc: (which is
2146 not set for Exim filters).
2148 6. When Exim created an email address using the login of the caller as the
2149 local part (e.g. when creating a From: or Sender: header line), it was not
2150 quoting the local part when it contained special characters such as @.
2152 7. Installed new OpenBSD configuration files.
2154 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2155 try to make them clearer.
2157 9. Callout options, other than the timeout value, were being ignored when
2158 verifying sender addresses in header lines. For example, when using
2160 verify = header_sender/callout=no_cache
2162 the cache was (incorrectly) being used.
2164 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2165 only the Cygwin environment.
2167 11. When return_path_on_delivery was set as a log selector, if different remote
2168 addresses in the same message used different return paths and parallel
2169 remote delivery occurred, the wrong values would sometimes be logged.
2170 (Whenever a remote delivery process finished, the return path value from
2171 the most recently started remote delivery process was logged.)
2173 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2174 lines when AUTH and/or TLS are in use. This is the "received protocol"
2175 field. Exim used to use "asmtp" for authenticated SMTP, without any
2176 indication (in the protocol name) for TLS use. Now it follows the RFC and
2177 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2178 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2179 names appear in log lines as well as in Received: header lines.
2181 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2182 extensions, and comparison tests, and to fix some bugs.
2184 14. Changes to the "personal" filter test:
2186 (1) The test was buggy in that it was just doing the equivalent of
2187 "contains" tests on header lines. For example, if a user's address was
2188 anne@some.where, the "personal" test would incorrectly be true for
2190 To: susanne@some.where
2192 This test is now done by extracting each address from the header in turn,
2193 and checking the entire address. Other tests that are part of "personal"
2194 are now done using regular expressions (for example, to check local parts
2195 of addresses in From: header lines).
2197 (2) The list of non-personal local parts in From: addresses has been
2198 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2199 taken from the Sieve specification recommendations.
2201 (3) If the message contains any header line starting with "List-" it is
2202 treated as non-personal.
2204 (4) The test for "circular" in the Subject: header line has been removed
2205 because it now seems ill-conceived.
2207 15. Minor typos in src/EDITME comments corrected.
2209 16. Installed latest exipick from John Jetmore.
2211 17. If headers_add on a router specified a text string that was too long for
2212 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2213 of string_sprintf() is now avoided.
2215 18. $message_body_size was not set (it was always zero) when running the DATA
2216 ACL and the local_scan() function.
2218 19. For the "mail" command in an Exim filter, no default was being set for
2219 the once_repeat time, causing a random time value to be used if "once" was
2220 specified. (If the value happened to be <= 0, no repeat happened.) The
2221 default is now 0s, meaning "never repeat". The "vacation" command was OK
2222 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2223 (I found it when inspecting the code).
2225 20. There is now an overall timeout for performing a callout verification. It
2226 defaults to 4 times the callout timeout, which applies to individual SMTP
2227 commands during the callout. The overall timeout applies when there is more
2228 than one host that can be tried. The timeout is checked before trying the
2229 next host. This prevents very long delays if there are a large number of
2230 hosts and all are timing out (e.g. when the network connections are timing
2231 out). The value of the overall timeout can be changed by specifying an
2232 additional sub-option for "callout", called "maxwait". For example:
2234 verify = sender/callout=5s,maxwait=20s
2236 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2237 to the end before writing, but this should make it even safer).
2239 22. Exim was forgetting that it had advertised PIPELINING for the second and
2240 subsequent messages on an SMTP connection. It was also not resetting its
2241 memory on STARTTLS and an internal HELO.
2243 23. When Exim logs an SMTP synchronization error within a session, it now
2244 records whether PIPELINING has been advertised or not.
2246 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2247 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2248 rather than long int.
2250 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2252 26. Added the never_mail option to autoreply.
2258 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2259 crash if the getsockname() call failed; this can happen if a connection is
2260 closed very soon after it is established. The problem was simply in the
2261 order in which certain operations were done, causing Exim to try to write
2262 to the SMTP stream before it had set up the file descriptor. The bug has
2263 been fixed by making things happen in the correct order.
2269 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2270 before the connection was closed, thus losing the rejection response.
2272 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2273 some early Solaris releases, but causes trouble in current releases where
2274 socklen_t is defined.
2276 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2279 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2282 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2283 that releasing the top of it at the end releases what was used for sub-
2284 expansions (unless the block got too big). However, discard this block if
2285 the first thing is a variable or header, so that we can use its block when
2286 it is dynamic (useful for very large $message_headers, for example).
2288 6. Lookups now cache *every* query, not just the most recent. A new, separate
2289 store pool is used for this. It can be recovered when all lookup caches are
2290 flushed. Lookups now release memory at the end of their result strings.
2291 This has involved some general refactoring of the lookup sources.
2293 7. Some code has been added to the store_xxx() functions to reduce the amount
2294 of flapping under certain conditions.
2296 8. log_incoming_interface used to affect only the <= reception log lines. Now
2297 it causes the local interface and port to be added to several more SMTP log
2298 lines, for example "SMTP connection from", and rejection lines.
2300 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2302 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2304 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2305 could be overwritten at the end of the current message (or the start of a
2306 new message if it was set in a HELO ACL). The value is now preserved for
2307 the duration of the SMTP connection.
2309 12. If a transport had a headers_rewrite setting, and a matching header line
2310 contained an unqualified address, that address was qualified, even if it
2311 did not match any rewriting rules. The underlying bug was that the values
2312 of the flags that permit the existence of unqualified sender and recipient
2313 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2314 non-local messages, and by -bnq for local messages) were not being
2315 preserved with the message after it was received.
2317 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2318 "next input=" as part of the text comprising the host identity instead of
2319 the correct text. The code was using the same buffer for two different
2320 strings. However, depending on which order the printing function evaluated
2321 its arguments, the bug did not always show up. Under Linux, for example, my
2322 test suite worked just fine.
2324 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2325 doesn't work with some older versions of Perl. It has been changed to "my",
2326 which in any case is probably the better facility to use.
2328 15. A really picky compiler found some instances of statements for creating
2329 error messages that either had too many or two few arguments for the format
2332 16. The size of the buffer for calls to the DNS resolver has been increased
2333 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2334 for addresses that have a lot of PTR records. This alleviates a problem; it
2335 does not fully solve it.
2337 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2338 buffer now truncates the list and logs the incident, which is the same
2339 action as happens when Exim is looking up a host name and its aliases.
2340 Previously in this situation something unpredictable would happen;
2341 sometimes it was "internal error: store_reset failed".
2343 18. If a server dropped the connection unexpectedly when an Exim client was
2344 using GnuTLS and trying to read a response, the client delivery process
2345 crashed while trying to generate an error log message.
2347 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2348 string, for example:
2350 warn message = H1: something\nH2: something
2352 the text was added as a single header line from Exim's point of view
2353 though it ended up OK in the delivered message. However, searching for the
2354 second and subsequent header lines using $h_h2: did not work. This has been
2355 fixed. Similarly, if a system filter added multiple headers in this way,
2356 the routers could not see them.
2358 20. Expanded the error message when iplsearch is called with an invalid key to
2359 suggest using net-iplsearch in a host list.
2361 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2362 ACLs are no longer actually imposed (and a message to that effect is
2365 22. If a "gecos" field in a passwd entry contained escaped characters, in
2366 particular, if it contained a \" sequence, Exim got it wrong when building
2367 a From: or a Sender: header from that name. A second bug also caused
2368 incorrect handling when an unquoted " was present following a character
2369 that needed quoting.
2371 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2372 was not being matched caselessly.
2374 24. Arranged for all hyphens in the exim.8 source to be escaped with
2377 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2378 properly. Recipient callout cache records were still being keyed to include
2379 the sender, even when use_sender was set false. This led to far more
2380 callouts that were necessary. The sender is no longer included in the key
2381 when use_sender is false.
2383 26. Added "control = submission" modifier to ACLs.
2385 27. Added the ${base62d: operator to decode base 62 numbers.
2387 28. dnsdb lookups can now access SRV records.
2389 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2390 the configuration file.
2392 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2393 (-v) mode. This makes the output for a verbose queue run more intelligible.
2395 31. Added a use_postmaster feature to recipient callouts.
2397 32. Added the $body_zerocount variable, containing the number of binary zero
2398 bytes in the message body.
2400 33. The time of last modification of the "new" subdirectory is now used as the
2401 "mailbox time last read" when there is a quota error for a maildir
2404 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2406 35. Added +ignore_unknown as a special item in host lists.
2408 36. Code for decoding IPv6 addresses in host lists is now included, even if
2409 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2410 address was recognized as an IP address, but was then not correctly decoded
2411 into binary, causing unexpected and incorrect effects when compared with
2418 1. Very minor rewording of debugging text in manualroute to say "list of
2419 hosts" instead of "hostlist".
2421 2. If verify=header_syntax was set, and a header line with an unqualified
2422 address (no domain) and a large number of spaces between the end of the
2423 name and the colon was received, the reception process suffered a buffer
2424 overflow, and (when I tested it) crashed. This was caused by some obsolete
2425 code that should have been removed. The fix is to remove it!
2427 3. When running in the test harness, delay a bit after writing a bounce
2428 message to get a bit more predictability in the log output.
2430 4. Added a call to search_tidyup() just before forking a reception process. In
2431 theory, someone could use a lookup in the expansion of smtp_accept_max_
2432 per_host which, without the tidyup, could leave open a database connection.
2434 5. Added the variables $recipient_data and $sender_data which get set from a
2435 lookup success in an ACL "recipients" or "senders" condition, or a router
2436 "senders" option, similar to $domain_data and $local_part_data.
2438 6. Moved the writing of debug_print from before to after the "senders" test
2441 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2442 problems for message scanning, either using a data ACL, or using
2443 local_scan() because the Received: header was not generated till after they
2444 were called (in order to set the time as the time of reception completion).
2445 I have revised the way this works. The header is now generated after the
2446 body is received, but before the ACL or local_scan() are called. After they
2447 are run, the timestamp in the header is updated.
2453 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2454 before starting a queue runner without re-exec. This happened only when
2455 deliver_drop_privilege was set or when the Exim user was set to root. The
2456 effect of the bug was that timeouts during subsequent deliveries caused
2457 crashes instead of being properly handled. The handler is now left at its
2458 default (and expected) setting.
2460 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2461 message, again when deliver_drop_privilege is set or Exim is run as root.
2462 The bug described in (1) was not present in this case, but the tidying up
2463 of the other signals was missing. I have made the two cases consistent.
2465 3. The ignore_target_hosts setting on a manualroute router was being ignored
2466 for hosts that were looked up using the /MX notation.
2468 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2471 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2472 operated on the sender address. After changing the $sender_address to <>
2473 for the sender address verify, Exim was re-instated it as the original
2474 (before rewriting) address, but remembering that it had rewritten it, so it
2475 wasn't rewriting it again. This bug also had the effect of breaking the
2476 sender address verification caching when the sender address was rewritten.
2478 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2479 This has been changed so that if the ip literal address matches
2480 ignore_target_hosts, the router declines.
2482 7. Added expansion conditions match_domain, match_address, and match_local_
2483 part (NOT match_host).
2485 8. The placeholder for the Received: header didn't have a length field set.
2487 9. Added code to Exim itself and to exim_lock to test for a specific race
2488 condition that could lead to file corruption when using MBX delivery. The
2489 issue is with the lockfile that is created in /tmp. If this file is removed
2490 after a process has opened it but before that process has acquired a lock,
2491 there is the potential for a second process to recreate the file and also
2492 acquire a lock. This could lead to two Exim processes writing to the file
2493 at the same time. The added code performs the same test as UW imapd; it
2494 checks after acquiring the lock that its file descriptor still refers to
2495 the same named file.
2497 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2498 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2501 11. Added the smtp_active_hostname option. If used, this will typically be made
2502 to depend on the incoming interface address. Because $interface_address is
2503 not set up until the daemon has forked a reception process, error responses
2504 that can happen earlier (such as "too many connections") no longer contain
2507 12. If an expansion in a condition on a "warn" statement fails because a lookup
2508 defers, the "warn" statement is abandoned, and the next ACL statement is
2509 processed. Previously this caused the whole ACL to be aborted.
2511 13. Added the iplsearch lookup type.
2513 14. Added ident_timeout as a log selector.
2515 15. Added tls_certificate_verified as a log selector.
2517 16. Added a global option tls_require_ciphers (compare the smtp transport
2518 option of the same name). This controls incoming TLS connections.
2520 17. I finally figured out how to make tls_require_ciphers do a similar thing
2521 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2522 before starting the TLS session.
2524 18. Tabs are now shown as \t in -bP output.
2526 19. If the log selector return_path_on_delivery was set, Exim crashed when
2527 bouncing a message because it had too many Received: header lines.
2529 20. If two routers both had headers_remove settings, and the first one included
2530 a superfluous trailing colon, the final name in the first list and the
2531 first name in the second list were incorrectly joined into one item (with a
2532 colon in the middle).
2538 1. Added -C and -D options to the exinext utility, mainly to make it easier
2539 to include in the automated testing, but these could be helpful when
2540 multiple configurations are in use.
2542 2. The exinext utility was not formatting the output nicely when there was
2543 an alternate port involved in the retry record key, nor when there was a
2544 message id as well (for retries that were specific to a specific message
2545 and a specific host). It was also confused by IPv6 addresses, because of
2546 the additional colons they contain. I have fixed the IPv4 problem, and
2547 patched it up to do a reasonable job for IPv6.
2549 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2550 delivery, the log line now contains "pipelined" if PIPELINING was used.
2552 4. An SMTP transport process used to panic and die if the bind() call to set
2553 an explicit outgoing interface failed. This has been changed; it is now
2554 treated in the same way as a connect() failure.
2556 5. A reference to $sender_host_name in the part of a conditional expansion
2557 that was being skipped was still causing a DNS lookup. This no longer
2560 6. The def: expansion condition was not recognizing references to header lines
2561 that used bh_ and bheader_.
2563 7. Added the _cache feature to named lists.
2565 8. The code for checking quota_filecount in the appendfile transport was
2566 allowing one more file than it should have been.
2568 9. For compatibility with Sendmail, the command line option
2576 and sets the incoming protocol and host name (for trusted callers). The
2577 host name and its colon can be omitted when only the protocol is to be set.
2578 Note the Exim already has two private options, -pd and -ps, that refer to
2579 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2580 "s", but I don't think that's a major issue.
2582 10. A number of refactoring changes to the code, none of which should affect
2585 (a) The number of logging options was getting close to filling up the
2586 32-bit word that was used as a bit map. I have split them into two classes:
2587 those that are passed in the argument to log_write(), and those that are
2588 only ever tested independently outside of that function. These are now in
2589 separate 32-bit words, so there is plenty of room for expansion again.
2590 There is no change in the user interface or the logging behaviour.
2592 (b) When building, for example, log lines, the code previously used a
2593 macro that called string_cat() twice, in order to add two strings. This is
2594 not really sufficiently general. Furthermore, there was one instance where
2595 it was actually wrong because one of the argument was used twice, and in
2596 one call a function was used. (As it happened, calling the function twice
2597 did not affect the overall behaviour.) The macro has been replaced by a
2598 function that can join an arbitrary number of extra strings onto a growing
2601 (c) The code for expansion conditions now uses a table and a binary chop
2602 instead of a serial search (which was left over from when there were very
2603 few conditions). Also, it now recognizes conditions like "pam" even when
2604 the relevant support is not compiled in: a suitably worded error message is
2605 given if an attempt is made to use such a condition.
2607 11. Added ${time_interval:xxxxx}.
2609 12. A bug was causing one of the ddress fields not to be passed back correctly
2610 from remote delivery subprocesses. The field in question was not being
2611 subsequently used, so this caused to problems in practice.
2613 13. Added new log selectors queue_time and deliver_time.
2615 14. Might have fixed a bug in maildirsizefile handling that threw up
2616 "unexpected character" debug warnings, and recalculated the data
2617 unnecessarily. In any case, I expanded the warning message to give more
2620 15. Added the message "Restricted characters in address" to the statements in
2621 the default ACL that block characters like @ and % in local parts.
2623 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2624 Three changes have been made:
2626 (a) There was a serious bug; a negative response to MAIL caused the whole
2627 recipient domain to be cached as invalid, thereby blocking all messages
2628 to all local parts at the same domain, from all senders. This bug has
2629 been fixed. The domain is no longer cached after a negative response to
2630 MAIL if the sender used is not empty.
2632 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2635 (c) A new callout option, "use_sender" has been added for people who want
2636 the modified behaviour.
2642 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2645 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2646 indeed breaks things for older releases.
2648 3. Added additional logging to the case where there is a problem reading data
2649 from a filter that is running in a subprocess using a pipe, in order to
2650 try to track down a specific problem.
2652 4. Testing facility fudge: when running in the test harness and attempting
2653 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2654 getting "No route to host". Convert this to a timeout.
2656 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2659 6. Some OS don't have socklen_t but use size_t instead. This affects the
2660 fifth argument of getsockopt() amongst other things. This is now
2661 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2662 can be set for individual OS. I have set it for SunOS5, OSF1, and
2663 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2664 some earlier ones do not.
2666 7. Change 4.30/15 was not doing the test caselessly.
2668 8. The standard form for an IPv6 address literal was being rejected by address
2669 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2670 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2671 this, as well as the form without the "IPv6" on the front (but only when
2672 address literals are enabled, of course).
2674 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2676 10. Exim crashed if a message with an empty sender address specified by -f
2677 encountered a router with an errors_to setting. This could be provoked only
2678 by a command such as
2682 where an empty string was supplied; "<>" did not hit this bug.
2684 11. Installed PCRE release 4.5.
2686 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2687 remained set. It is now erased.
2689 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2690 times from message ids (which are base 36 rather than the normal 62).
2692 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2693 were being counted as actual protocol errors, and logged if the log
2694 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2695 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2696 and DATA following a set of rejected RCPTs do not count as protocol errors.
2697 In other words, Exim assumes they were pipelined, though this may not
2698 actually be the case. Of course, in all cases the client gets an
2699 appropriate error code.
2701 15. If a lookup fails in an ACL condition, a message about the failure may
2702 be available; it is used if testing the ACL cannot continue, because most
2703 such messages specify what the cause of the deferral is. However, some
2704 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2705 that caused an old message to be retained and used if a later statement
2706 caused a defer, replacing the real cause of the deferral.
2708 16. If an IP address had so many PTR records that the DNS lookup buffer
2709 was not large enough to hold them, Exim could crash while trying to process
2710 the truncated data. It now detects and logs this case.
2712 17. Further to 4.21/58, another change has been made: if (and only if) the
2713 first line of a message (the first header line) ends with CRLF, a bare LF
2714 in a subsequent header line has a space inserted after it, so as not to
2715 terminate the header.
2717 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2718 unnecessarily, used atoi() instead of strtol(), and didn't check the
2719 termination when getting file sizes from file names by regex.
2721 19. Completely re-implemented the support for maildirsize files, in the light
2722 of a number of problems with the previous contributed implementation
2723 (4.30/29). In particular:
2725 . If the quota is zero, the maildirsize file is maintained, but no quota is
2728 . If the maildir directory does not exist, it is created before any attempt
2729 to write a maildirsize file.
2731 . The quota value in the file is just a cache; if the quota is changed in
2732 the transport, the new value overrides.
2734 . A regular expression is available for excluding directories from the
2737 20. The autoreply transport checks the characters in options that define the
2738 message's headers; it allows continued headers, but it was checking with
2739 isspace() after an embedded newline instead of explicitly looking for a
2742 21. If all the "regular" hosts to which an address was routed had passed their
2743 expiry times, and had not reached their retry times, the address was
2744 bounced, even if fallback hosts were defined. Now Exim should go on to try
2747 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2748 equivalent code in the SMTP transport. Some hosts send humungous responses
2749 to HELO/EHLO, more than 1024 it seems.
2751 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2752 gives compiler warnings in some environments. I've now done it "properly",
2755 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2756 (because of problems with the built-in one) was declared to return uschar *
2757 instead of char *, causing compiler failure.
2759 25. Fixed a file descriptor leak when processing alias/forward files.
2761 26. Fixed a minor format string issue in dbfn.c.
2763 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2765 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2766 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2769 29. When a sender address is verified, it is cached, to save repeating the test
2770 when there is more than one recipient in a message. However, when the
2771 verification involves a callout, it is possible for different callout
2772 options to be set for different recipients. It is too complicated to keep
2773 track of this in the cache, so now Exim always runs a verification when a
2774 callout is required, relying on the callout cache for the optimization.
2775 The overhead is duplication of the address routing, but this should not be
2778 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2779 to be verified with callout=postmaster, and the main callout worked but the
2780 postmaster check failed, the verification correctly failed. However, if a
2781 subsequent RCPT command asked for sender verification *without* the
2782 postmaster check, incorrect caching caused this verification also to fail,
2785 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2786 it was not caching the DNS options (qualify_single, search_parents) that
2787 were used when the lookup failed. A subsequent lookup with different
2788 options therefore always gave the same answer, though there were cases
2789 where it should not have. (Example: a "domains = !$mx_any" option on a
2790 dnslookup router: the "domains" option is always processed without any
2791 widening, but the router might have qualify_single set.) Now Exim uses the
2792 cached value only when the same options are set.
2794 32. Added John Jetmore's "exipick" utility to the distribution.
2796 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2797 than a timeout (e.g. a certificate is required, and is not provided), an
2798 Exim server now closes the connection immediately. Previously it waited for
2799 the client to close - but if the client is SSL, it seems that they each
2800 wait for each other, leading to a delay before one of them times out.
2802 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2803 maintained 0.8.x compatibility because I don't think many are using it, and
2804 it is clearly obsolete.
2806 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2809 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2810 client certificate was expired. A simple patch fixes this, though I don't
2811 understand the full logic of why the verify callback is called multiple
2814 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2815 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2816 which causes problems with some clients (such as the Certicom SSL Plus
2817 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2818 disables the coutermeasure allowing Eudora to connect."
2820 38. Exim was not checking that a write() to a log file succeeded. This could
2821 lead to Bad Things if a log got too big, in particular if it hit a file
2822 size limit. Exim now panics and dies if it cannot write to a log file, just
2823 as it does if it cannot open a log file.
2825 39. Modified OS/Makefile-Linux so that it now contains
2827 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2829 The two -D definitions ensure that Exim is compiled with large file
2830 support, which makes it possible to handle log files that are bigger than
2833 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2834 instance) a domain was checked against a named list that involved a lookup,
2835 causing $domain_data to be set, then another domain was checked against the
2836 same list, then the first domain was re-checked, the value of $domain_data
2837 after the final check could be wrong. In particular, if the second check
2838 failed, it could be set empty. This bug probably also applied to
2841 41. The strip_trailing_dot option was not being applied to the address given
2842 with the -f command-line option.
2844 42. The code for reading a message's header from the spool was incrementing
2845 $received_count, but never initializing it. This meant that the value was
2846 incorrect (doubled) while delivering a message in the same process in which
2847 it was received. In the most common configuration of Exim, this never
2848 happens - a fresh exec is done - but it can happen when
2849 deliver_drop_privilege is set.
2851 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2852 it now includes up to 150 characters of the unexpected data in the log
2855 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2856 and building data strings. The size of both of these buffers was 10 000
2857 bytes - far larger than anybody would *ever* want, thought I. Needless to
2858 say, somebody hit the limit. I have increased the maximum line length to
2859 20 000 and the maximum data length of concatenated lines to 100 000. I have
2860 also fixed two bugs, because there was no checking on these buffers. Tsk,
2861 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2862 buffer is too small.
2864 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2865 lsearch lookups. Now it does.
2867 46. When parsing a route_list item in a manualroute router, a fixed-length
2868 buffer was used for the list of hosts. I made this 1024 bytes long,
2869 thinking that nobody would ever have a list of hosts that long. Wrong.
2870 Somebody had a whole pile of complicated expansion conditions, and the
2871 string was silently truncated, leading to an expansion error. It turns out
2872 that it is easier to change to an unlimited length (owing to other changes
2873 that have happened since this code was originally written) than to build
2874 structure for giving a limitation error. The length of the item that
2875 expands into the list of hosts is now unlimited.
2877 47. The lsearch lookup could not handle data where the length of text line was
2878 more than 4095 characters. Such lines were truncated, leading to shortened
2879 data being returned. It should now handle lines of any length.
2881 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2882 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2885 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2886 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2887 no longer happens in comment lines. A list of replacements is now placed
2888 at the head of all of the source files, except those whose only change is
2889 to replace PERL_COMMAND in the very first #! line.
2891 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2892 messages on the queue, with a bottom-up merge sort, using code contributed
2893 by Michael Haardt. This should make operations like -bp somewhat faster on
2894 large queues. It won't affect queue runners, except when queue_run_in_order
2897 51. Installed eximstats 1.31 in the distribution.
2899 52. Added support for SRV lookups to the dnslookup router.
2901 53. If an ACL referred to $message_body or $message_body_end, the value was not
2902 reset for any messages that followed in the same SMTP session.
2904 54. The store-handling optimization for building very long strings was not
2905 differentiating between the different store pools. I don't think this
2906 actually made any difference in practice, but I've tidied it.
2908 55. While running the routers to verify a sender address, $sender_address
2909 was still set to the sender address. This is wrong, because when routing to
2910 send a bounce to the sender, it would be empty. Therefore, I have changed
2911 it so that, while verifying a sender address, $sender_address is set to <>.
2912 (There is no change to what happens when verifying a recipient address.)
2914 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2915 target A or AAAA records (if not already returned) without resetting the
2916 qualify_single or search_parents options of the DNS resolver. These are
2917 inappropriate in this case because the targets of MX and SRV records must
2918 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2919 target that, when qualified, matched something in the local domain. These
2920 two options are now turned off when doing these lookups.
2922 57. It seems that at least some releases of Reiserfs (which does not have the
2923 concept of a fixed number of inodes) returns zero and not -1 for the
2924 number of available inodes. This interacted badly with check_spool_inodes,
2925 which assumed that -1 was the "no such thing" setting. What I have done is
2926 to check that the total number of inodes is greater than zero before doing
2927 the test of how many are available.
2929 58. When a "warn" ACL statement has a log_message modifier, the message is
2930 remembered, and not repeated. This is to avoid a lot of repetition when a
2931 message has many recipients that cause the same warning to be written.
2932 Howewer, Exim was preserving the list of already written lines for an
2933 entire SMTP session, which doesn't seem right. The memory is now reset if a
2934 new message is started.
2936 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2937 debugging output unless log_rewrite was also set.
2939 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2940 of (int)(handle) when we know that handle contains (void *)(-1).
2942 61. The Exim daemon panic-logs an error return when it closes the incoming
2943 connection. However "connection reset by peer" seems to be common, and
2944 isn't really an error worthy of noting specially, so that particular error
2947 62. When Exim is trying to find all the local interfaces, it used to panic and
2948 die if the ioctl to get the interface flags failed. However, it seems that
2949 on at least one OS (Solaris 9) it is possible to have an interface that is
2950 included in the list of interfaces, but for which you get a failure error
2951 for this call. This happens when the interface is not "plumbed" into a
2952 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2953 failure of the "get flags" call assumes that the interface is down.
2955 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2956 makes life easier for people who are doing arithmetic on fields extracted
2957 from dates, where you often get leading zeros that should not be
2958 interpreted as octal.
2960 64. Added qualify_domain to the redirect router, to override the global
2963 65. If a pathologically long header line contained very many addresses (the
2964 report of this problem mentioned 10 000) and each of them was rewritten,
2965 Exim could use up a very large amount of memory. (It kept on making new
2966 copies of the header line as it rewrote, and never released the old ones.)
2967 At the expense of a bit more processing, the header rewriting function has
2968 been changed so that it no longer eats memory in this way.
2970 66. The generation of the Received: header has been moved from the time that a
2971 message starts to be received, to the time that it finishes. The timestamp
2972 in the Received: header should now be very close to that of the <= log
2973 line. There are two side-effects of this change:
2975 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2976 logged header lines no longer include the local Received: line, because
2977 it has not yet been created. The same applies to a copy of the message
2978 that is returned to a non-SMTP sender when a message is rejected.
2980 (b) When a filter file is tested using -bf, no additional Received: header
2981 is added to the test message. After some thought, I decided that this
2984 This change does not affect the value of $received_for. It is still set
2985 after address rewriting, but before local_scan() is called.
2987 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2989 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2990 gave an unhelpful panic error message, and a defer error. I have managed to
2991 change this behaviour so that it now rejects any supplied certificate,
2992 which seems right, as the list of acceptable certificates is empty.
2994 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2995 gave an unhelpful defer error. I have not managed to make this reject any
2996 supplied certificates, but the error message it gives is "no certificate
2997 supplied", which is not helpful.
2999 70. exigrep's output now also includes lines that are not associated with any
3000 message, but which match the given pattern. Implemented by a patch from
3001 Martin Sluka, which also tidied up the Perl a bit.
3003 71. Recipient callout verification, like sender verification, was using <> in
3004 the MAIL FROM command. This isn't really the right thing, since the actual
3005 sender may affect whether the remote host accepts the recipient or not. I
3006 have changed it to use the actual sender in the callout; this means that
3007 the cache record is now keyed on a recipient/sender pair, not just the
3008 recipient address. There doesn't seem to be a real danger of callout loops,
3009 since a callout by the remote host to check the sender would use <>.
3010 [SEE ABOVE: changed after hitting problems.]
3012 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3013 temporary errors. However, in the case of such a code being given after
3014 the end of a data transmission (i.e. after ".") Exim was failing to write
3015 a retry record for the message. (Yes, there was some broken host that was
3016 actually sending 8xx at this point.)
3018 73. An unknown lookup type in a host list could cause Exim to panic-die when
3019 the list was checked. (An example that provoked this was putting <; in the
3020 middle of a list instead of at the start.) If this happened during a DATA
3021 ACL check, a -D file could be left lying around. This kind of configuration
3022 error no longer causes Exim to die; instead it causes a defer errror. The
3023 incident is still logged to the main and panic logs.
3025 74. Buglet left over from Exim 3 conversion. The message "too many messages
3026 in one connection" was written to the rejectlog but not the mainlog, except
3027 when address rewriting (yes!) was being logged.
3029 75. Added write_rejectlog option.
3031 76. When a system filter was run not as root (that is, when system_filter_user
3032 was set), the values of the $n variables were not being returned to the
3033 main process; thus, they were not subsequently available in the $sn
3036 77. Added +return_path_on_delivery log selector.
3038 78. A connection timeout was being treated differently from recipients deferred
3039 when testing hosts_max_try with a message that was older than the host's
3040 retry timeout. (The host should not be counted, thus allowing all hosts to
3041 be tried at least once before bouncing.) This may have been the cause of an
3042 occasionally reported bug whereby a message would remain on the queue
3043 longer than the retry timeout, but would be bounced if a delivery was
3044 forced. I say "may" because I never totally pinned down the problem;
3045 setting up timeout/retry tests is difficult. See also the next item.
3047 79. The ultimate address timeout was not being applied to errors that involved
3048 a combination of host plus message (for example, a timeout on a MAIL
3049 command). When an address resolved to a number of possible hosts, and they
3050 were not all tried for each delivery (e.g. because of hosts_max_try), a
3051 message could remain on the queue longer than the retry timeout.
3053 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3056 81. Fixed an obscure SMTP outgoing bug which required at least the following
3057 conditions: (a) there was another message waiting for the same server;
3058 (b) the server returned 5xx to all RCPT commands in the first message so
3059 that the message was not completed; (c) the server dropped the connection
3060 or gave a negative response to the RSET that Exim sends to abort the
3061 transaction. The observed case was a dropped connection after DATA that had
3062 been sent in pipelining mode. That is, the server had advertised PIPELINING
3063 but was not implementing it correctly. The effect of the bug was incorrect
3064 behaviour, such as trying another host, and this could lead to a crash.
3070 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3071 and daemon.c were passed as pointers to ints; they should have been
3072 pointers to socklen_t variables (which are typically unsigned ints).
3074 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3077 3. Fixed a really odd bug that affected only the testing scheme; patching a
3078 certain fixed string in the binary changed the value of another string that
3079 happened to be identical to the end of the original first string.
3081 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3082 name", it returns that address as the IP address. On some operating
3083 systems (e.g. Solaris), it also passes back the IP address string as the
3084 "host name". However, on others (e.g. Linux), it passes back an empty
3085 string. Exim wasn't checking for this, and was changing the host name to an
3086 empty string, assuming it had been canonicized.
3088 5. Although rare, it is permitted to have more than one PTR record for a given
3089 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3090 all the names associated with an address, because they do in Solaris.
3091 However, it seems that they do not in Linux for data that comes from the
3092 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3093 I found this out when I moved to a new Linux workstation and tried to run
3094 the Exim test suite.
3096 To get round this problem I have changed the code so that it now does its
3097 own call to the DNS to look up PTR records when searching for a host name.
3098 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3099 addresses that are only in /etc/hosts are still found.
3101 This behaviour is, however, controlled by an option called host_lookup_
3102 order, which defaults to "bydns:byaddr". If people want to use the other
3103 order, or indeed, just use one or the other means of lookup, they can
3104 specify it in this variable.
3106 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3107 some operating systems, this comes back from gethostbyaddr() as an empty
3108 string, and this is what Exim used to test for. However, it seems that in
3109 other systems, "." is yielded. Exim now tests for this case too.
3111 7. The values of check_spool_space and check_log_space are now held internally
3112 as a number of kilobytes instead of an absolute number of bytes. If a
3113 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3114 kilobyte. This means that much larger values can be stored.
3116 8. Exim monitor: an attempt to get the action menu when not actually pointing
3117 at a message produces an empty menu entitled "No message selected". This
3118 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3119 no entries in it ("Shell widget menu has zero width and/or height"). So I
3120 have added a single, blank menu entry in this case.
3122 9. Added ${quote_local_part.
3124 10. MIME decoding is now applied to the contents of Subject: header lines when
3127 11. Now that a reference to $sender_host_address automatically causes a reverse
3128 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3129 host lookup before query-style lookups in lists that might use this
3130 variable. This has therefore been abolished, and the "net-" prefix is no
3131 longer necessary for query-style lookups.
3133 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3134 have been a typo for LFLAGS, so it has been changed.
3136 13. The install script calls Exim with "-C /dev/null" in order to find the
3137 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3138 to be output. Howeve, since Exim outputs its version number before the
3139 error, it didn't break the script. It just looked ugly. I fixed this by
3140 always allowing "-C /dev/null" if the caller is root.
3142 14. Ignore overlarge ACL variable number when reading spool file - insurance
3143 against a later release with more variables having written the file.
3145 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3146 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3147 this, as well as the form without the "IPv6" on the front.
3149 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3150 OS/Makefile-Darwin file.
3152 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3153 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3155 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3156 variable. However, some OS use ints for inodes. I've added cast to long int
3157 to get rid of the compiler warning.
3159 19. I had forgotten to lock out "/../" in configuration file names when
3160 ALT_CONFIG_PREFIX was set.
3162 20. Routers used for verification do not need to specify transports. However,
3163 if such a router generated a host list, and callout was configured, Exim
3164 crashed, because it could not find a port number from the (non-existent)
3165 transport. It now assumes port 25 in this circumstance.
3167 21. Added the -t option to exigrep.
3169 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3170 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3171 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3172 them is set, LOOKUP_LSEARCH is forced.
3174 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3176 24. Added sender and host information to the "rejected by local_scan()" log
3177 line; previously there was no indication of these.
3179 25. Added .include_if_exists.
3181 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3182 Linux. It turns out that not all file systems support this. Apparently some
3183 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3184 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3185 not supported on the file descriptor, is now ignored when Exim is trying to
3186 sync a directory. This applies only to Linux.
3188 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3190 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3191 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3194 29. Added support for maildirsize files from supplied patch (modified a bit).
3196 30. The use of :fail: followed by an empty string could lead Exim to respond to
3197 sender verification failures with (e.g.):
3199 550 Verification failed for <xxx>
3200 550 Sender verify failed
3202 where the first response line was missing the '-' that indicates it is not
3203 the final line of the response.
3205 31. The loop for finding the name of the user that called Exim had a hardwired
3206 limit of 10; it now uses the value of finduser_retries, which is used for
3207 all other user lookups.
3209 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3212 33. Exim was neglecting to zero errno before one call of strtol() when
3213 expanding a string and expecting an integer value. On some systems this
3214 resulted in spurious "integer overflow" errors. Also, it was casting the
3215 result into an int without checking.
3217 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3218 did not work. The code looks as if it has *never* worked, though it appears
3219 to have been documented since at least releast 1.62. I have made it work.
3221 35. The "timeout_DNS" error in retry rules, also documented since at least
3222 1.62, also never worked. As it isn't clear exactly what this means, and
3223 clearly it isn't a major issue, I have abolished the feature by treating it
3224 as "timeout", and writing a warning to the main and panic logs.
3226 36. The display of retry rules for -brt wasn't always showing the error code
3229 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3230 timeout_connect_A, timeout_connect_MX.
3232 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3233 to the empty sender.
3235 39. The daemon was not analysing the content of -oX till after it had closed
3236 stderr and disconnected from the controlling terminal. This meant that any
3237 syntax errors were only noted on the panic log, and the return code from
3238 the command was 0. By re-arranging the code a little, I've made the
3239 decoding happen first, so such errors now appear on stderr, and the return
3240 code is 1. However, the actual setting up of the sockets still happens in
3241 the disconnected process, so errors there are still only recorded on the
3244 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3245 connections (as happens on some IP stacks) was logged at start up time as
3246 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3247 it from "IPv6 and IPv4", which means that two separate sockets are being
3250 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3251 says whether AF_INET or AF_INET6 was passed as an argument.
3253 42. Exiwhat output was messed up when time zones were included in log
3256 43. Exiwhat now gives more information about the daemon's listening ports,
3257 and whether -tls-on-connect was used.
3259 44. The "port" option of the smtp transport is now expanded.
3261 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3262 silently ignored. Now an error message is written to the main and panic
3265 46. There's a new ACL modifier called "logwrite" which writes to a log file
3266 as soon as it is encountered.
3268 47. Added $local_user_uid and $local_user_gid at routing time.
3270 48. Exim crashed when trying to verify a sender address that was being
3273 49. Exim was recognizing only a space character after ".include". It now also
3274 recognizes a tab character.
3276 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3277 extracting the relevant information from the specification. The man page no
3278 longer contains scrambled data for the -d option, and I've added a section
3279 at the front about calling Exim under different names.
3281 51. Added "extra_headers" argument to the "mail" command in filter files.
3283 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3286 53. Installed eximstats 1.29.
3288 54. Added transport_filter_timeout as a generic transport option.
3290 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3291 Cc: header lines. This was required by RFC 822, but it not required by RFC
3294 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3295 incoming messages that did not have them. Now it does so only if the
3296 message originates locally, that is, if there is no associated remote host
3297 address. When Resent- header lines are present, this applies to the Resent-
3298 lines rather than the non-Resent- lines.
3300 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3301 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3303 58. Messages for configuration errors now include the name of the main
3304 configuration file - useful now that there may be more than one file in a
3305 list (.included file names were always shown).
3307 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3308 for those rare installations that do not start the daemon as root or run it
3309 setuid root. I've cut out the call to initgroups() if the daemon is not
3312 60. The Exim user and group can now be bound into the binary as text strings
3313 that are looked up at the start of Exim's processing.
3315 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3317 62. Added $mailstore_basename variable.
3319 63. Installed patch to sieve.c from Michael Haardt.
3321 64. When Exim failed to open the panic log after failing to open the main log,
3322 the original message it was trying to log was written to stderr and debug
3323 output, but if they were not available (the usual case in production), it
3324 was lost. Now it is written to syslog before the two lines that record the
3325 failures to open the logs.
3327 65. Users' Exim filters run in subprocesses under the user's uid. It is
3328 possible for a "deliver" command or an alias in a "personal" command to
3329 provoke an address rewrite. If logging of address rewriting is configured,
3330 this fails because the process is not running as root or exim. There may be
3331 a better way of dealing with this, but for the moment (because 4.30 needs
3332 to be released), I have disabled address rewrite logging when running a
3333 filter in a non-root, non-exim process.
3339 1. The buildconfig auxiliary program wasn't quoting the value set for
3340 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3341 not defined. This bug was masked in 4.22 by the effect that was fixed in
3344 2. Some messages that were rejected after a message id was allocated were
3345 shown as "incomplete" by exigrep. It no longer does this for messages that
3346 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3348 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3349 have allow_domain_literals set, the ID did not get logged in the <= line.
3350 Domain literals are now always recognized in Message-ID: header lines.
3352 4. The first argument for a ${extract expansion item is the key name or field
3353 number. Leading and trailing spaces in this item were not being ignored,
3354 causing some misleading effects.
3356 5. When deliver_drop_privilege was set, single queue runner processes started
3357 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3358 same command in the process it spins off) were not dropping privilege.
3360 6. When the daemon running as "exim" started a queue runner, it always
3361 re-executed Exim in the spun-off process. This is a waste of effort when
3362 deliver_drop_privilege is set. The new process now just calls the
3363 queue-runner function directly.
3369 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3372 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3373 ignored. Though the use of -oP was forcing the writing of a pid file, it
3374 was always written to the default place.
3376 3. If the message "no IP address found for host xxxx" is generated during
3377 incoming verification, it is now followed by identification of the incoming
3378 connection (so you can more easily find what provoked it).
3380 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3382 5. Added some features to "harden" Exim a bit more against certain attacks:
3384 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3385 be put in Local/Makefile. This is like the never_users runtime option,
3386 but it cannot be overridden. The default setting is "root".
3388 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3389 prefix string with which any file named in a -C command line option
3392 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3393 is retained for -C and -D only if the caller of Exim is root. Without
3394 it, the exim user may also use -C and -D and retain privilege.
3396 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3397 command line option is disabled.
3399 6. Macro names set by the -D option must start with an upper case letter, just
3400 like macro names defined in the configuration file.
3402 7. Added "dereference=" facility to LDAP.
3404 8. Two instances of the typo "uknown" in the source files are fixed.
3406 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3407 the Configure-Makefile script screwed up while processing it.
3409 10. Incorporated PCRE 4.4.
3411 11. The SMTP synchronization check was not operating right at the start of an
3412 SMTP session. For example, it could not catch a HELO sent before the client
3413 waited for the greeting. There is now a check for outstanding input at the
3414 point when the greeting is written. Because of the duplex, asynchronous
3415 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3416 way, but not yet received, when the check is performed.
3418 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3419 on TCP/IP sockets, because this apparently causes some broken clients to
3422 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3423 unchanged) from the Cygwin maintainer.
3425 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3426 was supported instead of testing for mbx. Effectively a typo.
3428 15. The spa authenticator server code was not checking that the input it
3429 received was valid base64.
3431 16. The debug output line for the "set" modifier in ACLs was not showing the
3432 name of the variable that was being set.
3434 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3436 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3437 lookup on its own. Something else was needed to trigger the lookup. For
3438 example, a match in host_lookup or the need for a host name in a host list.
3439 Now, if $sender_host_name is referenced and the host name has not yet been
3440 looked up, a lookup is performed. If the lookup fails, the variable remains
3441 empty, and $host_lookup_failed is set to "1".
3443 19. Added "eqi" as a case-independent comparison operator.
3445 20. The saslauthd authentication condition could segfault if neither service
3446 nor realm was specified.
3448 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3449 error message that was logged was misleading, and incoming SMTP
3450 connections were dropped. The message is now more accurate, and temporary
3451 errors are given to SMTP connections.
3453 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3454 (except RSET) with a 421 error, until QUIT is received. However, it was
3455 failing to send a response to QUIT.
3457 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3458 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3459 ACL is now run after the helo_try_verify_hosts code.
3461 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3462 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3463 case-independent, so other case variants are also recognized). Apparently
3464 some systems use these upper case variants.
3466 25. If more than two messages were waiting for the same host, and a transport
3467 filter was specified for the transport, Exim sent two messages over the
3468 same TCP/IP connection, and then failed with "socket operation on non-
3469 socket" when it tried to send the third.
3471 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3473 27. The extern definition of crypt16() in expand.c was not being excluded when
3474 the OS had its own crypt16() function.
3476 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3477 as a preferred synonym for return_size_limit, both as an option and as an
3480 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3482 30. Changed the default configuration ACL to relax the local part checking rule
3483 for addresses that are not in any local domains. For these addresses,
3484 slashes and pipe symbols are allowed within local parts, but the sequence
3485 /../ is explicitly forbidden.
3487 31. SPA server authentication was not clearing the challenge buffer before
3490 32. log_message in a "warn" ACL statement was writing to the reject log as
3491 well as to the main log, which contradicts the documentation and doesn't
3492 seem right (because no rejection is happening). So I have stopped it.
3494 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3495 However, I am unable to do any testing of this.
3497 34. Fixed an infelicity in the appendfile transport. When checking directories
3498 for a mailbox, to see if any needed to be created, it was accidentally
3499 using path names with one or more superfluous leading slashes; tracing
3500 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3502 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3503 contents are added to the log line that is written for every discarded
3504 recipient. (Previously a log_message setting was ignored.)
3506 36. The ${quote: operator now quotes the string if it is empty.
3508 37. The install script runs exim in order to find its version number. If for
3509 some reason other than non-existence or emptiness, which it checks, it
3510 could not run './exim', it was installing it with an empty version number,
3511 i.e. as "exim-". This error state is now caught, and the installation is
3514 38. An argument was missing from the function that creates an error message
3515 when Exim fails to connect to the socket for saslauthd authentication.
3516 This could cause Exim to crash, or give a corrupted message.
3518 39. Added isip, isip4, and isip6 to ${if conditions.
3520 40. The ACL variables $acl_xx are now saved with the message, and can be
3521 accessed later in routers, transports, and filters.
3523 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3524 strings in the file are not string-expanded.
3526 42. If a MAIL command specified a SIZE value that was too large to fit into an
3527 int variable, the check against message_size_limit failed. Such values are
3528 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3529 day this will have to be increased, but I don't think I want to be around
3530 when emails are that large.
3537 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3538 iconv() is not standard in FreeBSD.
3540 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3541 IPv6 enabled. The observed symptom was a segmentation fault on return from
3542 the function os_common_find_running_interfaces() in src/os.c.
3544 3. In the check_special_case() function in daemon.c I had used "errno" as an
3545 argument name, which causes warnings on some systems. This was basically a
3546 typo, since it was named "eno" in the comments!
3548 4. The code that waits for the clock to tick (at a resolution of some fraction
3549 of a second) so as to ensure message-id uniqueness was always waiting for
3550 at least one whole tick, when it could have waited for less. [This is
3551 almost certainly not relevant at current processor speeds, where it is
3552 unlikely to ever wait at all. But we try to future-proof.]
3554 5. The function that sleeps for a time interval that includes fractions of a
3555 second contained a race. It did not block SIGALRM between setting the
3556 timer, and suspending (a couple of lines later). If the interval was short
3557 and the sigsuspend() was delayed until after it had expired, the suspension
3558 never ended. On busy systems this could lead to processes getting stuck for
3561 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3562 process, before it forks any delivery processes. The open lookup caching
3563 mechanism meant that the open file or database connection was passed into
3564 the delivery process. The problem was that delivery processes always tidy
3565 up cached lookup data. This could cause a problem for the next delivery
3566 process started by the queue runner, because the external queue runner
3567 process does not know about the closure. So the next delivery process
3568 still has data in the lookup cache. In the case of a file lookup, there was
3569 no problem because closing a file descriptor in a subprocess doesn't affect
3570 the parent. However, if the lookup was caching a connection to a database,
3571 the connection was closed, and the second delivery process was likely to
3572 see errors such as "PGSQL: query failed: server closed the connection
3573 unexpectedly". The problem has been fixed by closing all cached lookups
3574 in a queue runner before running a delivery process.
3576 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3577 seem to have the "const" qualifier which it has on other OS. I've
3580 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3581 *of the same type* (client or server) with the same public name that an
3582 error should be diagnosed.
3584 9. When Exim looked up a host name for an IP address, but failed to find the
3585 original IP address when looking up the host name (a safety check), it
3586 output the message "<ip address> does not match any IP for NULL", which was
3587 confusing, to say the least. The bug was that the host name should have
3588 appeared instead of "NULL".
3590 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3591 user that is built into the binary, and the -C or -D options is used, root
3592 privilege is dropped before the configuration file is read. In addition,
3593 logging is switched to stderr instead of the normal log files. If the
3594 configuration then re-defines the Exim user, the unprivileged environment
3595 is probably not what is expected, so Exim logs a panic warning message (but
3598 However, if deliver_drop_privilege is set, the unprivileged state may well
3599 be exactly what is intended, so the warning has been cut out in that case,
3600 and Exim is allowed to try to write to its normal log files.
3606 1. smtp_return_error_details was not giving details for temporary sender
3607 or receiver verification errors.
3609 2. Diagnose a configuration error if two authenticators have the same public
3612 3. Exim used not to create the message log file for a message until the first
3613 delivery attempt. This could be confusing when incoming messages were held
3614 for policy or load reasons. The message log file is now created at the time
3615 the message is received, and an initial "Received" line is written to it.
3617 4. The automatically generated man page for command line options had a minor
3618 bug that caused no ill effects; however, a more serious problem was that
3619 the procedure for building the man page automatically didn't always
3620 operate. Consequently, release 4.20 contains an out-of-date version. This
3621 shouldn't happen again.
3623 5. When building Exim with embedded Perl support, the script that builds the
3624 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3625 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3627 6. The freeze_tell option was not being used for messages that were frozen on
3628 arrival, either by an ACL or by local_scan().
3630 7. Added the smtp_incomplete_transaction log selector.
3632 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3633 was accepting AUTH without a new EHLO.
3635 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3636 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3639 10. Logging of TCP/IP connections (when configured) now happens in the main
3640 daemon process instead of the child process, so that the TCP/IP connection
3641 count is more accurate (but it can never be perfect).
3643 11. The use of "drop" in a nested ACL was not being handled correctly in the
3644 outer ACL. Now, if condition failure induced by the nested "drop" causes
3645 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3646 or "require"), the connection is dropped.
3648 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3649 that yield "discard" can now be used with an "accept" or a "discard" verb,
3650 but an error is generated for any others (because I can't see a useful way
3651 to define what should happen).
3653 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3654 are now processed in the same way as lines in the Exim configuration file.
3655 In particular, continuation lines are supported.
3657 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3659 15. Added -ti meaning -t -i.
3661 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3662 domains, and warn by logging if others are found.
3664 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3665 returned from GIFCONF when Exim is trying to find the list of interfaces on
3666 a host. The code in os.c has been modified to copy each ifreq to an aligned
3667 structure in all cases.
3669 Also, in some cases, the returned ifreq's were being copied to a 'struct
3670 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3671 means the last couple of bytes of an IPv6 address could be chopped if the
3672 ifreq contained only a normal sockaddr (14 bytes storage).
3674 18. Named domain lists were not supported in the hosts_treat_as_local option.
3675 An entry such as +xxxx was not recognized, and was treated as a literal
3678 19. Ensure that header lines added by a DATA ACL are included in the reject log
3679 if the ACL subsequently rejects the message.
3681 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3682 MD5 (which is deprecated).
3684 21. When testing a filter file using -bf, Exim was writing a message when it
3685 took the sender from a "From " line in the message, but it was not doing so
3686 when it took $return_path from a Return-Path: header line. It now does.
3688 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3689 with a valid header line field name (a series of printing characters
3690 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3692 23. Changed "disc" in the source to "disk" to conform to the documentation and
3693 the book and for uniformity.
3695 24. Ignore Sendmail's -Ooption=value command line item.
3697 25. When execve() failed while trying to run a command in a pipe transport,
3698 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3699 could be confused with a return value of 69 from the command itself. This
3700 has been changed to 127, the value the shell returns if it is asked to run
3701 a non-existent command. The wording for the related log line suggests a
3702 non-existent command as the problem.
3704 26. If received_header_text expands to an empty string, do not add a Received:
3705 header line to the message. (Well, it adds a token one on the spool, but
3706 marks it "old" so that it doesn't get used or transmitted.)
3708 27. Installed eximstats 1.28 (addition of -nt option).
3710 28. There was no check for failure on the call to getsockname() in the daemon
3711 code. This can fail if there is a shortage of resources on the system, with
3712 ENOMEM, for example. A temporary error is now given on failure.
3714 29. Contrary to the C standard, it seems that in some environments, the
3715 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3716 program. Exim now does this explicitly; it affects the formatting of
3717 timestamps using strftime().
3719 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3720 complaints. I've now initialized all the variables, to avoid this.
3722 32. Header lines added by a system filter were not being "seen" during
3723 transport-time rewrites.
3725 33. The info_callback() function passed to OpenSSL is set up with type void
3726 (*)(SSL *, int, int), as described somewhere. However, when calling the
3727 function (actually a macro) that sets it up, the type void(*)() is
3728 expected. I've put in a cast to prevent warnings from picky compilers.
3730 34. If a DNS black list lookup found a CNAME record, but there were no A
3731 records associated with the domain it pointed at, Exim crashed.
3733 35. If a DNS black list lookup returned more than one A record, Exim ignored
3734 all but the first. It now scans all returned addresses if a particular IP
3735 value is being sought. In this situation, the contents of the
3736 $dnslist_value variable are a list of all the addresses, separated by a
3739 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3740 to accept a host name and all its aliases if the forward lookup for any of
3741 them yielded the IP address of the incoming connection. Now it accepts only
3742 those names whose forward lookup yields the correct IP address. Any other
3743 names are discarded. This closes a loophole whereby a rogue DNS
3744 administrator could create reverse DNS records to break through a
3745 wildcarded host restriction in an ACL.
3747 37. If a user filter or a system filter that ran in a subprocess used any of
3748 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3749 the wrong values were passed to the pipe command ($thisaddress had the
3750 value of $0, $0 had the value of $1, etc). This bug was introduced by
3751 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3753 38. Improved the line breaking for long SMTP error messages from ACLs.
3754 Previously, if there was no break point between 40 and 75 characters, Exim
3755 left the rest of the message alone. Two changes have been made: (a) I've
3756 reduced the minimum length to 35 characters; (b) if it can't find a break
3757 point between 35 and 75 characters, it looks ahead and uses the first one
3758 that it finds. This may give the occasional overlong line, but at least the
3759 remaining text gets split now.
3761 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3762 file descriptors might be low, and that setting 1000 would always raise it.
3763 It turns out that in some environments, the limit is already over 1000 and
3764 that lowering it causes trouble. So now Exim takes care not to decrease it.
3766 40. When delivering a message, the value of $return_path is set to $sender_
3767 address at the start of routing (routers may change the value). By an
3768 oversight, this default was not being set up when an address was tested by
3769 -bt or -bv, which affected the outcome if any router or filter referred to
3772 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3773 log only when "message" or "log_message" are set. However, if one of the
3774 conditions was an address verification, or a call to a nested ACL, the
3775 messages generated by the underlying test were being passed through. This
3776 no longer happens. The underlying message is available in $acl_verify_
3777 message for both "message" and "log_message" expansions, so it can be
3778 passed through if needed.
3780 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3781 new expansion $bh_ to give the encoded byte string without charset
3782 translation. Translation happens only if iconv() is available; HAVE_ICONV
3783 indicates this at build time. HEADERS_CHARSET gives the charset to
3784 translate to; headers_charset can change it in the configuration, and
3785 "headers charset" can change it in an individual filter file.
3787 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3788 that creates RFC 2047 encoded "words" labels them as that charset instead
3789 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3790 expansion operator; (ii) when Exim creates a From: line for a local
3791 message; (iii) when a header line is rewritten to include a "phrase" part.
3793 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3794 buggy, causing it to skip the first lines of messages whose message ID
3795 ended in 'D'. This would not have bitten before Exim release 4.14, because
3796 message IDs were unlikely to end in 'D' before then. The effect was to have
3797 incorrect size information for certain domains.
3799 45. #include "config.h" was missing at the start of the crypt16.c module. This
3800 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3803 46. If there was a timeout during a "random" callout check, Exim treated it as
3804 a failure of the random address, and carried on sending RSET and the real
3805 address. If the delay was just some slowness somewhere, the response to the
3806 original RCPT would be taken as a response to RSET and so on, causing
3807 mayhem of various kinds.
3809 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3810 when I implemented it. It didn't allow for the fact that some option values
3811 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3812 the right test for positive values.
3814 48. Domain names in DNS records are case-independent. Exim always looks them up
3815 in lower case. Some resolvers return domain names in exactly the case they
3816 appear in the zone file, that is, they may contain uppercase letters. Not
3817 all resolvers do this - some return always lower case. Exim was treating a
3818 change of case by a resolver as a change of domain, similar to a widening
3819 of a domain abbreviation. This triggered its re-routing code and so it was
3820 trying to route what was effectively the same domain again. This normally
3821 caused routing to fail (because the router wouldn't handle the domain
3822 twice). Now Exim checks for this case specially, and just changes the
3823 casing of the domain that it ultimately uses when it transmits the message
3826 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3829 50. If a filter generated a file delivery with a non-absolute name (possible if
3830 no home directory exists for the router), the forbid_file option was not
3833 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3834 the existing equality matching.
3836 52. Exim was using ints instead of ino_t variables in some places where it was
3837 dealing with inode numbers.
3839 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3840 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3841 called TMPDIR, and if it finds it is different, it changes its value.
3843 54. The smtp_printf() function is now made available to local_scan() so
3844 additional output lines can be written before returning. There is also an
3845 smtp_fflush() function to enable the detection of a dropped connection.
3846 The variables smtp_input and smtp_batched_input are exported to
3849 55. Changed the default runtime configuration: the message "Unknown user"
3850 has been removed from the ACL, and instead placed on the localuser router,
3851 using the cannot_route_message feature. This means that any verification
3852 failures that generate their own messages won't get overridden. Similarly,
3853 the "Unrouteable address" message that was in the ACL for unverifiable
3854 relay addresses has also been removed.
3856 56. Added hosts_avoid_esmtp to the smtp transport.
3858 57. The exicyclog script was not checking for the esoteric option
3859 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3860 will work only if exicyclog is run under the appropriate euid.
3862 58. Following a discussion on the list, the rules by which Exim recognises line
3863 endings on incoming messages have been changed. The -dropcr and drop_cr
3864 options are now no-ops, retained only for backwards compatibility. The
3865 following line terminators are recognized: LF CRLF CR. However, special
3866 processing applies to CR:
3868 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3869 nor a local message in the state where . is a terminator.
3871 (ii) If a bare CR is encountered in a header line, an extra space is added
3872 after the line terminator so as not to end the header. The reasoning
3873 behind this is that bare CRs in header lines are most likely either
3874 to be mistakes, or people trying to play silly games.
3876 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3877 was being incorrectly given as 18 bytes larger than it should have been.
3878 This is a VOB (very old bug).
3880 60. This may never have affected anything current, but just in case it has:
3881 When the local host is found other than at the start of a list of hosts,
3882 the local host, those with the same MX, and any that follow, are discarded.
3883 When the list in question was part of a longer list of hosts, the following
3884 hosts (not currently being processed) were also being discarded. This no
3885 longer happens. I'm not sure if this situation could ever has previously
3888 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3891 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3892 header. This is something that is recommended in a new Internet Draft, and
3893 is something that is documented as being done by Sendmail. There are two
3894 possible values. For messages generated by the autoreply transport, Exim
3897 Auto-Submitted: auto-replied
3899 whereas for all other generated messages (e.g. bounces) it adds
3901 Auto-Submitted: auto-generated
3903 63. The "personal" condition in filters now includes a test for the
3904 Auto-Submitted: header. If it contains the string "auto-" the message it
3905 not considered personal.
3907 64. Added rcpt_include_affixes as a generic transport option.
3909 65. Added queue_only_override (default true).
3911 66. Added the syslog_duplication option.
3913 67. If what should have been the first header line of a message consisted of
3914 a space followed by a colon, Exim was mis-interpreting it as a header line.
3915 It isn't of course - it is syntactically invalid and should therefore be
3916 treated as the start of the message body. The misbehaviour could have
3917 caused a number of strange effects, including loss of data in subsequent
3918 header lines, and spool format errors.
3920 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3921 client host had authenticated. This control can now be exercised by an ACL
3922 for more flexibility.
3924 69. By default, callouts do not happen when testing with -bh. There is now a
3925 variant, -bhc, which does actually run the callout code, including
3926 consulting and updating the callout cache.
3928 70. Added support for saslauthd authentication, courtesy of Alexander
3931 71. If statvfs() failed on the spool or log directories while checking their
3932 size for availability, Exim confusingly gave the error "space shortage".
3933 Furthermore, in debugging mode it crashed with a floating point exception.
3934 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3935 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3936 serious problem, Exim now writes to the main and panic logs when this
3937 happens, with details of the failure. It then refuses to accept the
3938 incoming message, giving the message "spool directory problem" or "log
3939 directory problem" with a 421 code for SMTP messages.
3941 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3942 0, 1, and 2 exist, because some OS complain for execs without them (see
3943 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3944 descriptors. However, the code omitted to check that the open succeeded,
3945 causing mysterious errors if for some reason the permissions on /dev/null
3946 got screwed. Now Exim writes a message to the main and panic logs, and
3947 bombs out if it can't open /dev/null.
3949 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3950 interact so that it is all more flexible. It is supposed to remain
3951 backwards compatible. Also added extra_local_interfaces.
3953 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3954 to bomb out with an assertion failure - to the client this appears as a
3955 connection drop. This problem occurs in the part of the code that was taken
3956 from the Samba project. Fortunately, the assertion is in a very simple
3957 function, so I have fixed this by reproducing the function inline in the
3958 one place where it is called, and arranging for authentication to fail
3959 instead of killing the process with assert().
3961 75. The SPA client code was not working when the server requested OEM rather
3962 than Unicode encoding.
3964 76. Added code to make require_files with a specific uid setting more usable in
3965 the case where statting the file as root fails - usually a non-root-mounted
3966 NFS file system. When this happens and the failure is EACCES, Exim now
3967 forks a subprocess and does the per-uid checking as the relevant uid.
3969 77. Added process_log_path.
3971 78. If log_file_path was not explicitly set, a setting of check_log_space or
3972 check_log_inodes was ignored.
3974 79. If a space check for the spool or log partitions fails, the incident is now
3975 logged. Of course, in the latter case the data may get lost...
3977 80. Added the %p formatting code to string_format() so that it can be used to
3978 print addresses in debug_print(). Adjusted all the address printing in the
3979 debugging in store.c to use %p rather than %d.
3981 81. There was a concern that a line of code in smtp_in.c could overflow a
3982 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3983 initially expressed, the concern was not well-founded, because trailing
3984 spaces are removed early. However, if the trailing spaces were followed by
3985 a NULL, they did not get removed, so the overflow was possible. Two fixes
3988 (a) I re-wrote the offending code in a cleaner fashion.
3989 (b) If an incoming SMTP command contains a NULL character, it is rejected
3992 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3993 runs initgroups(), so that if the Exim user is in any additional groups,
3994 they will be used during message reception.
4000 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob