1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.352 2006/05/22 17:43:10 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.
27 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
28 other effects) broke the use of negated acl sub-conditions.
30 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
33 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
34 "Deny" causes Exim to reject the incoming connection with a 554 error.
35 Unfortunately, if there is a major crisis, such as a disk failure,
36 tcp-wrappers gives "deny", whereas what one would like would be some
37 kind of temporary error. A kludge has been added to help with this.
38 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
39 554 error is used if errno is still zero or contains ENOENT (which occurs
40 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
43 PH/03 Add -lutil to the default FreeBSD LIBS setting.
45 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
46 errors. Otherwise a message that provokes a temporary error (when other
47 messages do not) can cause a whole host to time out.
49 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
50 addresses were routed directly to files or pipes from a redirect router.
51 File deliveries just didn't batch; pipe deliveries might have suffered
54 PH/06 A failure to get a lock for a hints database would erroneously always say
55 "Failed to get write lock", even when it was really a read lock.
57 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
58 of 0600. This has been changed to use the value of the lockfile_mode
59 option (which defaults to 0600).
61 PH/08 Applied small patch from the Sieve maintainer.
63 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
64 folder from quota calculations, a direct delivery into this folder messed
65 up the contents of the maildirsize file. This was because the regex was
66 used only to exclude .Trash (or whatever) when the size of the mailbox
67 was calculated. There was no check that a delivery was happening into an
68 excluded directory. This bug has been fixed by ignoring all quota
69 processing for deliveries into excluded directories.
71 PH/10 Added the maildirfolder_create_regex option to appendfile.
77 PH/01 The code for finding all the local interface addresses on a FreeBSD
78 system running IPv6 was broken. This may well have applied to all BSD
79 systems, as well as to others that have similar system calls. The broken
80 code found IPv4 interfaces correctly, but gave incorrect values for the
81 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
82 that it would not match correctly against @[] and not recognize the IPv6
85 PH/02 The ipliteral router was not recognizing addresses of the form user@
86 [ipv6:....] because it didn't know about the "ipv6:" prefix.
88 PH/03 Added disable_ipv6.
90 PH/04 Changed $reply_address to use the raw form of the headers instead of the
91 decoded form, because it is most often used to construct To: headers
92 lines in autoreplies, and the decoded form may well be syntactically
93 invalid. However, $reply_address has leading white space removed, and all
94 newlines turned into spaces so that the autoreply transport does not
97 PH/05 If group was specified without a user on a router, and no group or user
98 was specified on a transport, the group from the router was ignored.
100 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
101 for visible compile-time settings that can be used to change these
102 numbers, for those that want even more. Backwards compatibility with old
103 spool files has been maintained. However, going back to a previous Exim
104 release will lost any variables that are in spool files.
106 PH/07 Two small changes when running in the test harness: increase delay when
107 passing a TCP/IP connection to a new process, in case the original
108 process has to generate a bounce, and remove special handling of
109 127.0.0.2 (sic), which is no longer necessary.
111 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
112 be the same on different OS.
114 PH/09 Moved a debug statement in filter processing to avoid a race problem when
117 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
118 whether --show-vars was specified or not
120 JJ/02 exipick: Added support for new ACL variable spool format introduced
123 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
124 syntactically invalid From: or Reply-to: line, and a filter used this to
125 generate an autoreply, and therefore failed to obtain an address for the
126 autoreply, Exim could try to deliver to a non-existent relative file
127 name, causing unrelated and misleading errors. What now happens is that
128 it logs this as a hard delivery error, but does not attempt to create a
131 PH/11 The exinext utility has a -C option for testing purposes, but although
132 the given file was scanned by exinext itself; it wasn't being passed on
133 when Exim was called.
135 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
136 an end-of-file indication when reading a command response.
138 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
139 compiled. In many other places in Exim, IPv6 addresses are always
140 recognized, so I have changed this. It also means that IPv4 domain
141 literals of the form [IPV4:n.n.n.n] are now always recognized.
143 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
144 used if the router is not running as root, for example, when verifying at
145 ACL time, or when using -bh. The debugging output from this situation was
146 non-existent - all you got was a failure to exec. I have made two
149 (a) Failures to set uid/gid, the current directory, or a process leader
150 in a subprocess such as that created by queryprogram now generate
151 suitable debugging ouput when -d is set.
153 (b) The queryprogram router detects when it is not running as root,
154 outputs suitable debugging information if -d is set, and then runs
155 the subprocess without attempting to change uid/gid.
157 PH/15 Minor change to Makefile for building test_host (undocumented testing
160 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
161 additional section of a DNS packet that returns MX or SRV records.
162 Instead, it always explicitly searches for A/AAAA records. This avoids
163 major problems that occur when a DNS server includes only records of one
164 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
165 fixed another bug: if SRV records were looked up and the corresponding
166 address records were *not* found in the additional section, the port
167 values from the SRV records were lost.
169 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
170 using the correct key (the original address) when searching the retry
171 rules in order to find which one to use for generating the retry hint.
173 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
174 adding the default one. Similarly, if it contains a Reply-To: header, the
175 errors_reply_to option, if set, is not used.
177 PH/19 When calculating a retry time, Exim used to measure the "time since
178 failure" by looking at the "first failed" field in the retry record. Now
179 it does not use this if it is later than than the arrival time of the
180 message. Instead it uses the arrival time. This makes for better
181 behaviour in cases where some deliveries succeed, thus re-setting the
182 "first failed" field. An example is a quota failure for a huge message
183 when small messages continue to be delivered. Without this change, the
184 "time since failure" will always be short, possible causing more frequent
185 delivery attempts for the huge message than are intended.
186 [Note: This change was subsequently modified - see PH/04 for 4.62.]
188 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
189 $1, $2, $3) because the numerical variables can be reset during some
190 expansion items (e.g. "match"), thereby losing the authentication data.
192 PH/21 Make -bV show the size of off_t variables so that the test suite can
193 decide whether to run tests for quotas > 2G.
195 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
196 mailbox_size, and mailbox_filecount in the appendfile transport. If a
197 filecount value is greater than 2G or if a quota value is greater than 2G
198 on a system where the size of off_t is not greater than 4, a panic error
201 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
202 never match. The debug and -bh output now contains an explicit error
203 message indicating a malformed IPv4 address or mask.
205 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
206 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
209 PH/25 Do not write to syslog when running in the test harness. The only
210 occasion when this arises is a failure to open the main or panic logs
211 (for which there is an explicit test).
213 PH/26 Added the /no_tell option to "control=freeze".
215 PH/27 If a host name lookup failed very early in a connection, for example, if
216 the IP address matched host_lookup and the reverse lookup yielded a name
217 that did not have a forward lookup, an error message of the form "no IP
218 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
219 could be logged. Now it outputs the IP address instead of "NULL".
221 PH/28 An enabling patch from MH: add new function child_open_exim2() which
222 allows the sender and the authenticated sender to be set when
223 submitting a message from within Exim. Since child_open_exim() is
224 documented for local_scan(), the new function should be too.
226 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
227 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
228 results in an empty string is now treated as unset.
230 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
232 PH/31 Added sender_verify_fail logging option.
234 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
235 needed by RFC 822 but not by RFC 2822 was commented out. I have now
236 tidied the source and removed it altogether.
238 PH/33 When a queue run was abandoned because the load average was too high, a
239 log line was always written; now it is written only if the queue_run log
240 selector is set. In addition, the log line for abandonment now contains
241 information about the queue run such as the pid. This is always present
242 in "start" and "stop" lines but was omitted from the "abandon" line.
244 PH/34 Omit spaces between a header name and the colon in the error message that
245 is given when verify = headers_syntax fails (if there are lots of them,
246 the message gets confusing).
248 PH/35 Change the default for dns_check_names_pattern to allow slashes within
249 names, as there are now some PTR records that contain slashes. This check
250 is only to protect against broken name servers that fall over on strange
251 characters, so the fact that it applies to all lookups doesn't matter.
253 PH/36 Now that the new test suite is complete, we can remove some of the
254 special code in Exim that was needed for the old test suite. For example,
255 sorting DNS records because real resolvers return them in an arbitrary
256 order. The new test suite's fake resolver always returns records in the
259 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
260 bounces) except when queue_only is set, to avoid logging races between
261 the different processes.
263 PH/38 Panic-die if .include specifies a non-absolute path.
265 PH/39 A tweak to the "H" retry rule from its user.
267 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
268 a label. They prevented compilation on older perls.
270 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
271 a warning to be raised on newish perls.
273 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
274 on queue. Changes to match documented behaviour of showing count of
275 messages matching specified criteria.
277 PH/40 Changed the default ident timeout from 30s to 5s.
279 PH/41 Added support for the use of login_cap features, on those BSD systems
280 that have them, for controlling the resources used by pipe deliveries.
282 PH/42 The content-scanning code uses fopen() to create files in which to put
283 message data. Previously it was not paying any attention to the mode of
284 the files. Exim runs with umask(0) because the rest of the code creates
285 files with open(), and sets the required mode explicitly. Thus, these
286 files were ending up world-writeable. This was not a big issue, because,
287 being within the spool directory, they were not world-accessible. I have
288 created a function called modefopen, which takes an additional mode
289 argument. It sets umask(777), creates the file, chmods it to the required
290 mode, then resets the umask. All the relevant calls to fopen() in the
291 content scanning code have been changed to use this function.
293 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
294 to 24 hours. This avoids potential overflow problems when processing G
295 and H retry rules. I suspect nobody ever tinkers with this value.
297 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
299 PH/45 When the plaintext authenticator is running as a client, the server's
300 challenges are checked to ensure they are valid base64 strings. By
301 default, the authentication attempt is cancelled if an invalid string is
302 received. Setting client_ignore_invalid_base64 true ignores these errors.
303 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
304 they are received. Thus, the responses can be made to depend on the
305 challenges. If an invalid string is ignored, an empty string is placed in
308 PH/46 Messages that are created by the autoreply transport now contains a
309 References: header, in accordance with RFCs 2822 and 3834.
311 PH/47 Added authenticated_sender_force to the smtp transport.
313 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
315 PH/49 Installed latest patch from the Sieve maintainer.
317 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
318 was also set, the appendfile transport was unnecessarily scanning a
319 directory of message files (e.g. for maildir delivery) to find the count
320 of files (along with the size), even though it did not need this
321 information. It now does the scan only if it needs to find either the
322 size of the count of files.
324 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
326 PH/52 Two bugs concerned with error handling when the smtp transport is
329 (i) Exim was not creating retry information for temporary errors given
330 for individual recipients after the DATA command when the smtp transport
331 was used in LMTP mode. This meant that they could be retried too
332 frequently, and not timed out correctly.
334 (ii) Exim was setting the flag that allows error details to be returned
335 for LMTP errors on RCPT commands, but not for LMTP errors for individual
336 recipients that were returned after the DATA command.
338 PH/53 This is related to PH/52, but is more general: for any failing address,
339 when detailed error information was permitted to be returned to the
340 sender, but the error was temporary, then after the final timeout, only
341 "retry timeout exceeded" was returned. Now it returns the full error as
342 well as "retry timeout exceeded".
344 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
345 do this, and (what is worse) MTAs that accept it.
347 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
348 will now be deprecated.
350 PH/56 New os.c-cygwin from the Cygwin maintainer.
352 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
353 formats (previously only available in exim formats via -bpr, -bpru,
354 and -bpra. Now also available in native and exiqgrep formats)
356 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
357 with very large, slow to parse queues
359 JJ/08 exipick: added ! as generic prefix to negate any criteria format
361 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
363 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
364 responses to authentication challenges, though it was showing the
365 challenges; (ii) I've removed the CR characters from the debug output for
368 PH/58 Allow for the insertion of a newline as well as a space when a string
369 is turned into more than one encoded-word during RFC 2047 encoding. The
370 Sieve code now uses this.
372 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
373 data_4xx, lost_connection, tls_required.
375 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
376 message was being sent as an SMTP response.
378 PH/61 Add -l and -k options to exicyclog.
380 PH/62 When verifying, if an address was redirected to one new address, so that
381 verification continued, and the new address failed or deferred after
382 having set something in $address_data, the value of $address_data was not
383 passed back to the ACL. This was different to the case when no
384 redirection occurred. The value is now passed back in both cases.
386 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
387 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
388 use login_cap.h, so on its own it isn't the distinguishing feature. The
389 new name refers directly to the setclassresources() function.
391 PH/65 Added configuration files for NetBSD3.
393 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
395 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
396 is preferred over IPv4.
398 PH/68 The bounce_return_message and bounce_return_body options were not being
399 honoured for bounces generated during the reception of non-SMTP messages.
400 In particular, this applied to messages rejected by the ACL. This bug has
401 been fixed. However, if bounce_return_message is true and bounce_return_
402 body is false, the headers that are returned for a non-SMTP message
403 include only those that have been read before the error was detected.
404 (In the case of an ACL rejection, they have all been read.)
406 PH/69 The HTML version of the specification is now built in a directory called
407 spec_html instead of spec.html, because the latter looks like a path with
408 a MIME-type, and this confuses some software.
410 PH/70 Catch two compiler warnings in sieve.c.
412 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
413 function verify_get_ident() calls ip_connect() to connect a socket, but
414 if the "connect()" function timed out, ip_connect() used to close the
415 socket. However, verify_get_ident() also closes the socket later, and in
416 between Exim writes to the log, which may get opened at this point. When
417 the socket was closed in ip_connect(), the log could get the same file
418 descriptor number as the socket. This naturally causes chaos. The fix is
419 not to close the socket in ip_connect(); the socket should be closed by
420 the function that creates it. There was only one place in the code where
421 this was missing, in the iplookup router, which I don't think anybody now
422 uses, but I've fixed it anyway.
424 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
425 well as to direct DNS lookups. Otherwise the handling of names in host
426 lists is inconsistent and therefore confusing.
432 PH/01 Two changes to the default runtime configuration:
434 (1) Move the checks for relay_from_hosts and authenticated clients from
435 after to before the (commented out) DNS black list checks.
437 (2) Add control=submission to the relay_from_hosts and authenticated
438 clients checks, on the grounds that messages accepted by these
439 statements are most likely to be submissions.
441 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
443 (1) Generate an error if the third argument for the ${prvs expansion is
446 (2) Treat a missing third argument of ${prvscheck as if it were an empty
449 (3) Reset the variables that are obtained from the first argument of
450 ${prvscheck and used in the second argument before leaving the code,
451 because their memory is reclaimed, so using them afterwards may do
454 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
455 one (it's much easier than Tom thought :-).
457 (5) Because of (4), we can now allow for the use of $prvscheck_result
458 inside the third argument.
460 PH/03 For some reason, the default setting of PATH when running a command from
461 a pipe transport was just "/usr/bin". I have changed it to
464 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
465 anything to be listed in the output from -bV.
467 PH/05 When a filter generated an autoreply, the entire To: header line was
468 quoted in the delivery log line, like this:
470 => >A.N.Other <ano@some.domain> <original@ddress> ...
472 This has been changed so that it extracts the operative address. There
473 may be more than one such address. If so, they are comma-separated, like
476 => >ano@some.domain,ona@other.domain <original@ddress> ...
478 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
479 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
480 not being looked up in the rDNS to get a host name, Exim was showing the
481 IP address twice in Received: lines, even though the IP addresses were
482 identical. For example:
484 Received: from [1.2.3.4] (helo=[1.2.3.4])
486 However, if the real host name was known, it was omitting the HELO data
487 if it matched the actual IP address. This has been tidied up so that it
488 doesn't show the same IP address twice.
490 PH/07 When both +timestamp and +memory debugging was on, the value given by
491 $tod_xxx expansions could be wrong, because the tod_stamp() function was
492 called by the debug printing, thereby overwriting the timestamp buffer.
493 Debugging no longer uses the tod_stamp() function when +timestamp is set.
495 PH/08 When the original message was included in an autoreply transport, it
496 always said "this is a copy of the message, including all the headers",
497 even if body_only or headers_only was set. It now gives an appropriate
500 PH/09 Applied a patch from the Sieve maintainer which:
502 o fixes some comments
503 o adds the (disabled) notify extension core
504 o adds some debug output for the result of if/elsif tests
505 o points to the current vacation draft in the documentation
506 and documents the missing references header update
510 o fixes a bug in processing the envelope test (when testing
511 multiple envelope elements, the last element determinted the
514 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
515 Electronic Mail") by including:
517 Auto-submitted: auto-generated
519 in the messages that it generates (bounce messages and others, such as
520 warnings). In the case of bounce messages for non-SMTP mesages, there was
521 also a typo: it was using "Auto_submitted" (underscore instead of
522 hyphen). Since every message generated by Exim is necessarily in response
523 to another message, thes have all been changed to:
525 Auto-Submitted: auto-replied
527 in accordance with these statements in the RFC:
529 The auto-replied keyword:
531 - SHOULD be used on messages sent in direct response to another
532 message by an automatic process,
534 - MUST NOT be used on manually-generated messages,
536 - MAY be used on Delivery Status Notifications (DSNs) and Message
537 Disposition Notifications (MDNs),
539 - MUST NOT be used on messages generated by automatic or periodic
540 processes, except for messages which are automatic responses to
543 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
544 to the default Received: header definition.
546 PH/12 Added log selector acl_warn_skipped (default on).
548 PH/13 After a successful wildlsearch lookup, discard the values of numeric
549 variables because (a) they are in the wrong storage pool and (b) even if
550 they were copied, it wouldn't work properly because of the caching.
552 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
553 checking when decoding. Apparently there are clients that generate
554 overlong encoded strings. Why am I not surprised?
556 PH/15 If the first argument of "${if match_address" was not empty, but did not
557 contain an "@" character, Exim crashed. Now it writes a panic log message
558 and treats the condition as false.
560 PH/16 In autoreply, treat an empty string for "once" the same as unset.
562 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
563 extension "envelope-auth". The code is finished and in agreement with
564 other implementations, but there is no documentation so far and in fact,
565 nobody wrote the draft yet. This extension is currently #undef'ed, thus
566 not changing the active code.
568 Print executed "if" and "elsif" statements when debugging is used. This
569 helps a great deal to understand what a filter does.
571 Document more things not specified clearly in RFC3028. I had all this
572 sorted out, when out of a sudden new issues came to my mind. Oops."
574 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
577 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
578 canonical form (as documented). However, after a host name lookup from
579 the IP address, check_host() was doing a simple string comparison with
580 addresses acquired from the DNS when checking that the found name did
581 have the original IP as one of its addresses. Since any found IPv6
582 addresses are likely to be in abbreviated form, the comparison could
583 fail. Luckily, there already exists a function for doing the comparison
584 by converting both addresses to binary, so now that is used instead of
587 PH/20 There was another similar case to PH/19, when a complete host name was
588 given in a host list; looking up its IP address could give an abbreviated
589 form, whereas the current host's name might or might not be abbreviated.
590 The same fix has been applied.
596 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
597 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
600 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
601 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
603 PH/03 Typo: missing ".o" in src/pcre/Makefile.
605 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
606 header line, restrict the check to what is listed in RFCs 2369 and 2929.
607 Also, for "Auto-Submitted", treat anything other than "no" as
608 non-personal, in accordance with RFC 3834. (Previously it treated
609 anything starting "auto-" as non-personal.)
611 TF/01 The control=submission/name=... option had a problem with syntax
612 errors if the name included a slash character. The /name= option
613 now slurps the rest of the string, so it can include any characters
614 but it must come last in the list of options (after /sender_retain
617 PH/05 Some modifications to the interface to the fake nameserver for the new
625 TK/01 Added the "success_on_redirect" address verification option. See
626 NewStuff for rationale and an example.
628 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
630 PH/02 Patch to exigrep to allow it to work on syslog lines.
632 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
633 fread() to skip over the body file's header line, because in Cygwin the
634 header line is locked and is inaccessible.
636 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
637 co-exist for some time) to make it clear that it is the Exim ID that is
638 referenced, not the Message-ID: header line.
640 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
641 string_format() function, because snprintf() does not exist on all
644 PH/06 The use of forbid_filter_existstest now also locks out the use of the
645 ${stat: expansion item.
647 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
648 protocol synchronization error", to keep the pedants happy.
650 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
651 well as for IRIX systems, when gcc is being used. See the host.c source
654 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
656 PH/10 Named domain lists were not working if used in a queue_smtp_domains
659 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
660 transport and to the smtp transport in LMTP mode.
662 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
664 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
665 run a filter in a subprocess. This could lead to confusion in subsequent
666 lookups in the parent process. There should also be a search_tidyup() at
667 the end of the subprocess.
669 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
670 only if the host matched helo_try_verify_hosts, which caused the
671 verification to occur when the EHLO/HELO command was issued. The ACL just
672 tested the remembered result. Now, if a previous verification attempt has
673 not happened, "verify = helo" does it there and then.
675 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
677 TK/03 Fix log output including CR from clamd.
679 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
680 exist provoked a memory error which could cause a segfault.
682 PH/15 Installed PCRE 6.2
684 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
686 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
687 of the problem. Specifically, suggested +O2 rather than +O1 for the
690 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
692 PH/20 If a delivery was routed to a non-standard port by means of an SRV
693 record, the port was not correctly logged when the outgoing_port log
694 selector was set (it logged the transort's default port).
696 PH/21 Added support for host-specific ports to manualroute, queryprogram,
697 fallback_hosts, and "hosts" in the smtp transport.
699 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
700 host errors such as "Connection refused".
702 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
703 authentication with radiusclient 0.4.9:
705 - Error returned from rc_read_config was caught wrongly
706 - Username/password not passed on to radius server due to wrong length.
708 The presumption is that some radiusclient API changes for 4.51/PH/17
709 were not taken care of correctly. The code is still untested by me (my
710 Linux distribution still has 0.3.2 of radiusclient), but it was
711 contributed by a Radius user.
713 PH/24 When doing a callout, the value of $domain wasn't set correctly when
714 expanding the "port" option of the smtp transport.
716 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
717 while reading a MIME header. Thanks to Tom Hughes for a patch.
719 PH/24 Include config.h inside local_scan.h so that configuration settings are
722 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
723 that in an ACL for RCPT (for example), you can examine exactly what was
726 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
727 commands, but it was not correctly comparing the address with the actual
728 client host address. Thus, it would show the EHLO address in Received:
729 header lines when this was not necessary.
731 PH/27 Added the % operator to ${eval:}.
733 PH/28 Exim tries to create and chdir to its spool directory when it starts;
734 it should be ignoring failures (because with -C, for example, it has lost
735 privilege). It wasn't ignoring creation failures other than "already
738 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
741 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
742 because an input line was too long, either on its own, or by virtue of
743 too many continuations, the temporary file was not being removed, and the
744 return code was incorrect.
746 PH/31 Missing "BOOL" in function definition in filtertest.c.
748 PH/32 Applied Sieve patches from the maintainer.
750 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
752 PH/33 Added "verify = not_blind".
754 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
755 Local/Makefile (with some defaults set). These are used in built scripts
756 such as exicyclog, but they have never been used in the exim_install
757 script (though there are many overriding facilities there). I have
758 arranged that the exim_install script now takes note of these two
761 PH/35 Installed configuration files for Dragonfly.
763 PH/36 When a locally submitted message by a trusted user did not contain a
764 From: header, and the sender address was obtained from -f or from an SMTP
765 MAIL command, and the trusted user did not use -F to supply a sender
766 name, $originator_name was incorrectly used when constructing a From:
767 header. Furthermore, $originator_name was used for submission mode
768 messages from external hosts without From: headers in a similar way,
769 which is clearly wrong.
771 PH/37 Added control=suppress_local_fixups.
773 PH/38 When log_selector = +received_sender was set, and the addition of the
774 sender made the log line's construction buffer exactly full, or one byte
775 less than full, an overflow happened when the terminating "\n" was
778 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
779 when the result of a list match is failure because a DNS lookup failed.
781 PH/40 RM_COMMAND is now used in the building process.
783 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
784 the "build-* directories that it finds.
786 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
787 address in a domain literal was a prefix of an interface address.
789 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
790 when verifying a sender address, unless rewrite_headers is false.
792 PH/44 Wrote a long comment about why errors_to addresses are verified as
793 recipients, not senders.
795 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
796 the ratelimit ACL was added.
798 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
800 PH/46 Added extra information about PostgreSQL errors to the error string.
802 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
803 avoiding the need to install special zones in a real server. This is
804 backwards compatible; if it can't find the fake resolver, it drops back.
805 Thus, both old and new test suites can be run.
807 TF/02 Added util/ratelimit.pl
809 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
810 clock is set back in time.
812 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
813 Candler <B.Candler@pobox.com>.
815 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
816 OK for addresses that are the result of redirections.
818 PH/48 A number of further additions for the benefit of the new test suite,
819 including a fake gethostbyname() that interfaces to the fake DNS resolver
822 TF/06 The fix for widen_domains has also been applied to qualify_single and
823 search_parents which are the other dnslookup options that can cause
826 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
829 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
831 TF/07 Exim produced the error message "an SRV record indicated no SMTP
832 service" if it encountered an MX record with an empty target hostname.
833 The message is now "an MX or SRV record indicated no SMTP service".
835 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
836 if the DNS of the sending site is misconfigured. This is quite a
837 common situation. This change restores the behaviour of treating a
838 helo verification defer as a failure.
840 PH/51 If self=fail was set on a router, the bounce message did not include the
841 actual error message.
847 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
849 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
850 command itself ended in error, the underlying message about the transport
851 filter timeout was being overwritten with the pipe command error. Now the
852 underlying error message should be appended to the second error message.
854 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
856 PH/02 Reduce the amount of output that "make" produces by default. Full output
857 can still be requested.
859 PH/03 The warning log line about a condition test deferring for a "warn" verb
860 was being output only once per connection, rather than after each
861 occurrence (because it was using the same function as for successful
862 "warn" verbs). This seems wrong, so I have changed it.
864 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
865 it should not have, which might have caused a crash in the right
866 circumstances, but probably never did.
868 PH/04 Installed a modified version of Tony Finch's patch to make submission
869 mode fix the return path as well as the Sender: header line, and to
870 add a /name= option so that you can make the user's friendly name appear
873 TF/03 Added the control = fakedefer ACL modifier.
875 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
876 Mark Lowes for thorough testing.
878 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
880 TK/03 Merged latest SRS patch from Miles Wilton.
882 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
883 with the definition in sysexits.h (which is #included earlier).
884 Fortunately, Exim does not actually use EX_OK. The code used to try to
885 preserve the sysexits.h value, by assumimg that macro definitions were
886 scanned for macro replacements. I have been disabused of this notion,
887 so now the code just undefines EX_OK before #including unistd.h.
889 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
890 in the smtp transport. When a block could not be written in a single
891 write() function, the timeout was being re-applied to each part-write.
892 This seems wrong - if the receiver was accepting one byte at a time it
893 would take for ever. The timeout is now adjusted when this happens. It
894 doesn't have to be particularly precise.
896 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
897 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
899 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
900 without a domain if the check to <postmaster@domain> fails.
902 SC/01 Eximstats: added -xls and the ability to specify output files
903 (patch written by Frank Heydlauf).
905 SC/02 Eximstats: use FileHandles for outputing results.
907 SC/03 Eximstats: allow any combination of xls, txt, and html output.
909 SC/04 Eximstats: fixed display of large numbers with -nvr option
911 SC/05 Eximstats: fixed merging of reports with empty tables.
913 SC/06 Eximstats: added the -include_original_destination flag
915 SC/07 Eximstats: removed tabs and trailing whitespace.
917 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
919 TK/06 MBOX spool code: Add real "From " MBOX separator line
920 so the .eml file is really in mbox format (even though
921 most programs do not really care). Patch from Alex Miller.
923 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
924 The latter is generated from $received_to and is only set if the
925 message has one envelope recipient. SA can use these headers,
926 obviously out-of-the-box. Patch from Alex Miller.
928 PH/08 The ${def test on a variable was returning false if the variable's
929 value was "0", contrary to what the specification has always said!
930 The result should be true unless the variable is empty.
932 PH/09 The syntax error of a character other than { following "${if
933 def:variable_name" (after optional whitespace) was not being diagnosed.
934 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
935 accidental colon was present, for example, could give incorrect results.
937 PH/10 Tidied the code in a number of places where the st_size field of a stat()
938 result is used (not including appendfile, where other changes are about
941 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
942 This involved changing a lot of size variables from int to off_t. It
943 should work with maildirs and everything.
945 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
946 spamd dying while we are connected to it.
948 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
951 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
952 function for BDB 4.3.
954 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
956 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
957 the sources for PCRE have been re-arranged and more files are now
960 PH/15 The code I had for printing potentially long long variables in PH/11
961 above was not the best (it lost precision). The length of off_t variables
962 is now inspected at build time, and an appropriate printing format (%ld
963 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
964 to be "long long int" or "long int". This is needed for the internal
965 formatting function string_vformat().
967 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
968 the configuration file to be ":syslog", then the script "guesses" where
969 the logs files are, rather than using the compiled in default. In our
970 case the guess is not the same as the compiled default, so the script
971 suddenly stopped working when I started to use syslog. The patch checks
972 to see if log_file_path is "". If so, it attempts to read it from exim
973 with no configuration file to get the compiled in version, before it
974 falls back to the previous guessing code."
976 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
977 implementing BATV in an Exim configuration. See NewStuff for the gory
980 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
981 Makefile that are specific to HP-UX.
983 PH/18 If the "use_postmaster" option was set for a recipient callout together
984 with the "random" option, the postmaster address was used as the MAIL
985 FROM address for the random test, but not for the subsequent recipient
986 test. It is now used for both.
988 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
989 patch removes a few documentation additions to RFC 3028, because the
990 latest draft now contains them. It adds the new en;ascii-case comparator
991 and a new error check for 8bit text in MIME parts. Comparator and
992 require names are now matched exactly. I enabled the subaddress
993 extension, but it is not well tested yet (read: it works for me)."
995 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
996 rework some of the code of TK/09 above to avoid the hardwired use of
997 "%lld" and "long long". Replaced the call to snprintf() with a call to
1000 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1001 records point to non-existent hosts", "retry timeout exceeded", and
1002 "retry time not reached for any host after a long failure period".
1004 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1005 experimental DomainKeys support:
1007 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1008 (2) On an error such as an illegally used "control", the wrong name for
1009 the control was given.
1011 These problems did NOT occur unless DomainKeys support was compiled.
1013 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1015 PH/24 Added ${if match_ip condition.
1017 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1018 databases so that it will be absolutely obvious if a crash occurs in the
1019 DB library. This is a regular occurrence (often caused by mis-matched
1022 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1023 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1024 picked up on a user's system that detects such things. There doesn't seem
1025 to be a gcc warning option for this - only an attribute that has to be
1026 put on the function's prototype. It seems that in Fedora Core 4 they have
1027 set this on a number of new functions. No doubt there will be more in due
1030 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1031 specify a transport. However, if an address that was verified by such a
1032 router was the subject of a callout, Exim crashed because it tried to
1033 read the rcpt_include_affixes from the non-existent transport. Now it
1034 just assumes that the setting of that option is false. This bug was
1035 introduced by 4.51/PH/31.
1037 PH/28 Changed -d+all to exclude +memory, because that information is very
1038 rarely of interest, but it makes the output a lot bigger. People tend to
1039 do -d+all out of habit.
1041 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1042 code in os-type was giving problems when libc.so lives in lib64, like on
1045 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1046 aren't the modern standard, and it seems that some systems' include files
1047 don't always have them. Exim was already checking for some of the newer
1048 ones like T_AAAA, and defining it itself. I've added checks for all the
1049 record types that Exim uses.
1051 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1052 not automatically generating a new one, as it is supposed to. This
1053 prevented TLS from working. If the file did exist, but contained invalid
1054 data, a new version was generated, as expected. It was only the case of a
1055 non-existent file that was broken.
1057 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1058 with a change in libdomainkeys > 0.64.
1060 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1061 from DNS. If the selector record carries the flag, it now has
1062 precedence over the domain-wide flag.
1064 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1066 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1067 the use of a port name that isn't defined in /etc/services) occurred, the
1068 message was deferred as in a normal delivery, and thus remained on the
1069 spool, instead of being failed because of the mua_wrapper setting. This
1070 is now fixed, and I tidied up some of the mua_wrapper messages at the
1073 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1074 the messages in a hash of arrays rather than using individual hashes.
1075 This is a bit cleaner and results in dramatic memory savings, albeit
1076 at a slight CPU cost.
1078 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1079 as requested by Marc Sherman.
1081 SC/10 Eximstats: added histograms for user specified patterns as requested
1084 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1086 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1087 fopen() in the content-scanning modules that did not already have it.
1093 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1094 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1096 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1098 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1100 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1101 to test Sieve filters that use "vacation".
1103 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1104 that changes the way the GnuTLS parameters are stored in the cache file.
1105 The new format can be generated externally. For backward compatibility,
1106 if the data in the cache doesn't make sense, Exim assumes it has read an
1107 old-format file, and it generates new data and writes a new file. This
1108 means that you can't go back to an older release without removing the
1111 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1112 work if there are any delivery delays because "one_time" forces the
1113 parent to be marked "delivered", so its unseen clone is never tried
1114 again. For this reason, Exim now forbids the simultaneous setting of
1117 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1118 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1119 entry if you want to know the details. The fix, however, neglected to
1120 consider the case when local delivery batching is involved. The test for
1121 "previously delivered" was not happening when checking to see if an
1122 address could be batched with a previous (undelivered) one; under
1123 certain circumstances this could lead to multiple deliveries to the same
1126 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1127 in its include files, and this causes problems building Exim.
1129 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1130 header_syntax) but Exim was just ignoring anything given after a slash.
1131 In particular, this caused confusion with an attempt to use "verify =
1132 reverse_host_lookup/defer_ok". An error is now given when options are
1133 supplied for verify items that do not have them. (Maybe reverse_host_
1134 lookup should have a defer_ok option, but that's a different point.)
1136 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1137 defined by RFC 821) to 2048, because there were problems with some AUTH
1138 commands, and RFC 1869 says the size should be increased for extended
1139 SMTP commands that take arguments.
1141 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1144 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1145 "unknown" error; now it says that the functionality isn't in the binary.
1147 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1148 an address' error message when a string expansion fails (syntax or
1149 whatever). Otherwise the password may appear in the log. Following change
1150 PH/42 below, there is no longer a chance of it appearing in a bounce
1153 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1155 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1156 of its addresses was ever tried. (Bugzilla bug #2.)
1158 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1159 the result incorrectly in the debug output. (It correctly added a newline
1160 to what was transported.)
1162 TF/01 Added $received_time.
1164 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1165 commented out examples of how to interface to a virus scanner and to
1166 SpamAssassin. Also added commented examples of av_scanner and
1167 spamd_address settings.
1169 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1170 and controls are allowed in which ACLs. There were a couple of minor
1171 errors. Some of the entries in the conditions table (which is a table of
1172 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1173 negation of where the condition IS allowed.
1175 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1177 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1178 header file does not have a version number, so I've had to invent a new
1179 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1180 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1181 radiusclient), but it was contributed by a Radius user.
1183 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1184 files or directories, for OpenSSL.
1186 PH/19 When an Exim process that is running as root has to create an Exim log
1187 file, it does so in a subprocess that runs as exim:exim so as to get the
1188 ownership right at creation (otherwise, other Exim processes might see
1189 the file with the wrong ownership). There was no test for failure of this
1190 fork() call, which would lead to the process getting stuck as it waited
1191 for a non-existent subprocess. Forks do occasionally fail when resources
1192 run out. I reviewed all the other calls to fork(); they all seem to check
1195 PH/20 When checking for unexpected SMTP input at connect time (before writing
1196 the banner), Exim was not dealing correctly with a non-positive return
1197 from the read() function. If the client had disconnected by this time,
1198 the result was a log entry for a synchronization error with an empty
1199 string after "input=" when read() returned zero. If read() returned -1
1200 (an event I could not check), uninitialized data bytes were printed.
1201 There were reports of junk text (parts of files, etc) appearing after
1204 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1206 PH/22 Added support for macro redefinition, and (re)definition in between
1207 driver and ACL definitions.
1209 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1210 forgetting to use the resulting value; it was using the unexpanded value.
1212 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1213 hadn't been configured. The fix is from Juergen Kreileder, who
1214 understands it better than I do:
1216 "Here's what I see happening with three configured cyrus_sasl
1217 authenticators configured (plain, login, cram-md5):
1219 On startup auth_cyrus_sasl_init() gets called for each of these.
1220 This means three calls to sasl_listmech() without a specified mech_list.
1221 => SASL tests which mechs of all available mechs actually work
1222 => three warnings about OTP not working
1223 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1225 With the patch, sasl_listmech() also gets called three times. But now
1226 SASL's mech_list option is set to the server_mech specified in the the
1227 authenticator. Or in other words, the answer from sasl_listmech()
1228 gets limited to just the mech you're testing for (which is different
1230 => the return list contains just 'plain' or 'login', 'cram-md5' or
1231 nothing depending on the value of ob->server_mech.
1233 I've just tested the patch: Authentication still works fine,
1234 unavailable mechs specified in the exim configuration are still
1235 caught, and the auth.log warnings about OTP are gone."
1237 PH/25 When debugging is enabled, the contents of the command line are added
1238 to the debugging output, even when log_selector=+arguments is not
1241 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1242 answer is "GNU", and only if the return is "GNU/something" is the answer
1245 PH/27 $acl_verify_message is now set immediately after the failure of a
1246 verification in an ACL, and so is available in subsequent modifiers. In
1247 particular, the message can be preserved by coding like this:
1249 warn !verify = sender
1250 set acl_m0 = $acl_verify_message
1252 Previously, $acl_verify_message was set only while expanding "message"
1253 and "log_message" when a very denied access.
1255 PH/28 Modified OS/os.c-Linux with
1257 -#ifndef OS_LOAD_AVERAGE
1258 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1260 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1261 nomenclature these days.)
1263 PH/29 Installed patch from the Sieve maintainer that adds the options
1264 sieve_useraddress and sieve_subaddress to the redirect router.
1266 PH/30 In these circumstances:
1267 . Two addresses routed to the same list of hosts;
1268 . First host does not offer TLS;
1269 . First host accepts first address;
1270 . First host gives temporary error to second address;
1271 . Second host offers TLS and a TLS session is established;
1272 . Second host accepts second address.
1273 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1274 and peerdn, if requested) that were in fact used only for the second
1277 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1278 attention to any local part prefix or suffix that was matched by the
1279 router that accepted the address. It now behaves in the same way as it
1280 does for delivery: the affixes are removed from the local part unless
1281 rcpt_include_affixes is set on the transport.
1283 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1284 timeout during the DATA phase of an incoming message.
1286 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1287 applied a patch sanctioned by the Sieve maintainer.
1289 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1290 the uid or gid is negative. A case of a negative gid caused this to be
1291 noticed. The fix allows for either to be negative.
1293 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1294 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1297 PH/36 The change PH/12 above was broken. Fixed it.
1299 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1300 the grounds that routing the same address twice would always produce the
1301 same answer. This might have been true once, but it is certainly no
1302 longer true now. Routing a child address may depend on the previous
1303 routing that produced that child. Some complicated redirection strategies
1304 went wrong when messages had multiple recipients, and made Exim's
1305 behaviour dependent on the order in which the addresses were given.
1307 I have moved the duplicate checking until after the routing is complete.
1308 Exim scans the addresses that are assigned to local and remote
1309 transports, and removes any duplicates. This means that more work will be
1310 done, as duplicates will always all be routed, but duplicates are
1311 presumably rare, so I don't expect this is of any significance.
1313 For deliveries to pipes, files, and autoreplies, the duplicate checking
1314 still happens during the routing process, since they are not going to be
1317 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1318 It corrects a timeout issue with spamd. This is Ian's comment: "The
1319 background is that sometimes spamd either never reads data from a
1320 connection it has accepted, or it never writes response data. The exiscan
1321 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1322 blindly assumes that writes won't block so it may never time out."
1324 PH/39 Allow G after quota size as well as K and M.
1326 PH/40 The value set for $authenticated_id in an authenticator may not contain
1327 binary zeroes or newlines because the value is written to log lines and
1328 to spool files. There was no check on this. Now the value is run through
1329 the string_printing() function so that such characters are converted to
1330 printable escape sequences.
1332 PH/41 $message_linecount is a new variable that contains the total number of
1333 lines in the message. Compare $body_linecount, which is the count for the
1336 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1337 bounce and delay warning messages, except in certain special cases, which
1340 (a) An SMTP error message from a remote host;
1341 (b) A message specified in a :fail: redirection;
1342 (c) A message specified in a "fail" command in a system filter;
1343 (d) A message specified in a FAIL return from the queryprogram router;
1344 (e) A message specified by the cannot_route_message router option.
1346 In these cases only, Exim does include the error details in bounce and
1347 warning messages. There are also a few cases where bland messages such
1348 as "unrouteable address" or "local delivery error" are given.
1350 PH/43 $value is now also set for the "else" part of a ${run expansion.
1352 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1353 being worked on, but at least Exim now implements the latest version to
1356 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1357 process to complete was treated as a delivery failure, a timeout while
1358 writing the message to the pipe was logged, but erroneously treated as a
1359 successful delivery. Such timeouts include transport filter timeouts. For
1360 consistency with the overall process timeout, these timeouts are now
1361 treated as errors, giving rise to delivery failures by default. However,
1362 there is now a new Boolean option for the pipe transport called
1363 timeout_defer, which, if set TRUE, converts the failures into defers for
1364 both kinds of timeout. A transport filter timeout is now identified in
1367 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1368 systems where "make" and "gmake" are different, calling "gmake" at top
1369 level broke things. I've arranged for the value of $(MAKE) to be passed
1370 from the Makefile to this script so that it can call the same version of
1374 A note about Exim versions 4.44 and 4.50
1375 ----------------------------------------
1377 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1378 changes of various kinds. As a consequence, a big documentation update was
1379 needed. This delayed the release for rather longer than seemed good, especially
1380 in the light of a couple of (minor) security issues. Therefore, the changes
1381 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1382 release. So 4.44 and 4.50 are in effect two different branches that both start
1385 I have left the 4.50 change log unchanged; it contains all the changes since
1386 4.43. The change log for 4.44 is below; many of its items are identical to
1387 those for 4.50. This seems to be the most sensible way to preserve the
1388 historical information.
1394 1. Minor wording change to the doc/README.SIEVE file.
1396 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1397 computation of the current number of files was incorrect.
1399 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1400 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1403 4. Give more explanation in the error message when the command for a transport
1404 filter fails to execute.
1406 5. There are several places where Exim runs a non-Exim command in a
1407 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1408 was being done only for the command run by the queryprogram router. It is
1409 now done for all such subprocesses. The other cases are: ${run, transport
1410 filters, and the commands run by the lmtp and pipe transports.
1412 6. Added CONFIGURE_GROUP build-time option.
1414 7. Some older OS have a limit of 256 on the maximum number of file
1415 descriptors. Exim was using setrlimit() to set 1000 as a large value
1416 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1417 systems. I've change it so that if it can't get 1000, it tries for 256.
1419 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1420 was an oversight, and furthermore, ever since the addition of extra
1421 controls (e.g. 4.43/32), the checks on when to allow different forms of
1422 "control" were broken. There should now be diagnostics for all cases when a
1423 control that does not make sense is encountered.
1425 9. Added the /retain_sender option to "control=submission".
1427 10. $recipients is now available in the predata ACL (oversight).
1429 11. Tidy the search cache before the fork to do a delivery from a message
1430 received from the command line. Otherwise the child will trigger a lookup
1431 failure and thereby defer the delivery if it tries to use (for example) a
1432 cached ldap connection that the parent has called unbind on.
1434 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1435 of $address_data from the recipient verification was clobbered by the
1436 sender verification.
1438 13. The value of address_data from a sender verification is now available in
1439 $sender_address_data in subsequent conditions in the ACL statement.
1441 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1443 15. Added a new option "connect=<time>" to callout options, to set a different
1446 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1447 was its contents. (It was OK if the option was not defined at all.)
1449 17. A "Completed" log line is now written for messages that are removed from
1450 the spool by the -Mrm option.
1452 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1453 information about exactly what failed.
1455 19. Added -dd to debug only the daemon process.
1457 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1458 handles timeouts, both on the server side and network timeouts. Renamed the
1459 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1461 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1462 It is now set to "smtps".
1464 22. $host_address is now set to the target address during the checking of
1465 ignore_target_hosts.
1467 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1468 being passed; this would have caused $sender_host_name to have been used if
1469 matching the list had actually called for a host name (not very likely,
1470 since this list is usually IP addresses). A host name is now passed as
1473 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1474 code that specifies a non-restarting handler (typically sigaction() in
1475 modern systems) in an attempt to fix a rare and obscure crash bug.
1477 25. Narrowed the window for a race in the daemon that could cause it to ignore
1478 SIGCHLD signals. This is not a major problem, because they are used only to
1479 wake it up if nothing else does.
1481 26. A malformed maildirsize file could cause Exim to calculate negative values
1482 for the mailbox size or file count. Odd effects could occur as a result.
1483 The maildirsize information is now recalculated if the size or filecount
1486 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1487 support for a long time. Removed HAVE_SYS_VFS_H.
1489 28. Installed the latest version of exipick from John Jetmore.
1491 29. In an address list, if the pattern was not a regular expression, an empty
1492 subject address (from a bounce message) matched only if the pattern was an
1493 empty string. Non-empty patterns were not even tested. This was the wrong
1494 because it is perfectly reasonable to use an empty address as part of a
1495 database query. An empty address is now tested by patterns that are
1496 lookups. However, all the other forms of pattern expect the subject to
1497 contain a local part and a domain, and therefore, for them, an empty
1498 address still always fails if the pattern is not itself empty.
1500 30. Exim went into a mad DNS loop when attempting to do a callout where the
1501 host was specified on an smtp transport, and looking it up yielded more
1502 than one IP address.
1504 31. Re-factored the code for checking spool and log partition space into a
1505 function that finds that data and another that does the check. The former
1506 is then used to implement four new variables: $spool_space, $log_space,
1507 $spool_inodes, and $log_inodes.
1509 32. The RFC2047 encoding function was originally intended for short strings
1510 such as real names; it was not keeping to the 75-character limit for
1511 encoded words that the RFC imposes. It now respects the limit, and
1512 generates multiple encoded words if necessary. To be on the safe side, I
1513 have increased the buffer size for the ${rfc2047: expansion operator from
1516 33. It is now permitted to omit both strings after an "if" condition; if the
1517 condition is true, the result is "true". As before, when the second string
1518 is omitted, a false condition yields an empty string. This makes it less
1519 cumbersome to write custom ACL and router conditions.
1521 34. Failure to deliver a bounce message always caused it to be frozen, even if
1522 there was an errors_to setting on the router. The errors_to setting is now
1525 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1526 canonical form (fully expanded) before being placed in
1527 $sender_host_address.
1529 36. The table in the code that translates DNS record types into text (T_A to
1530 "A" for instance) was missing entries for NS and CNAME. It is just possible
1531 that this could have caused confusion if both these types were looked up
1532 for the same domain, because the text type is used as part of Exim's
1533 per-process caching. But the chance of anyone hitting this buglet seems
1536 37. The dnsdb lookup has been extended in a number of ways.
1538 (1) There is a new type, "zns", which walks up the domain tree until it
1539 finds some nameserver records. It should be used with care.
1541 (2) There is a new type, "mxh", which is like "mx" except that it returns
1542 just the host names, not the priorities.
1544 (3) It is now possible to give a list of domains (or IP addresses) to be
1545 looked up. The behaviour when one of the lookups defers can be
1546 controlled by a keyword.
1548 (4) It is now possible to specify the separator character for use when
1549 multiple records are returned.
1551 38. The dnslists ACL condition has been extended: it is now possible to supply
1552 a list of IP addresses and/or domains to be looked up in a particular DNS
1555 39. Added log_selector=+queue_time_overall.
1557 40. When running the queue in the test harness, wait just a tad after forking a
1558 delivery process, to get repeatability of debugging output.
1560 41. Include certificate and key file names in error message when GnuTLS fails
1561 to set them up, because the GnuTLS error message doesn't include the name
1562 of the failing file when there is a problem reading it.
1564 42. Allow both -bf and -bF in the same test run.
1566 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1568 44. The "Exiscan patch" is now merged into the mainline Exim source.
1570 45. Sometimes the final signoff response after QUIT could fail to get
1571 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1572 < 0 before doing a fflush(). This bug looks as though it goes back to the
1573 introduction of TLS in release 3.20, but "sometimes" must have been rare
1574 because the tests only now provoked it.
1576 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1577 (this can affect the format of dates).
1579 47. exim_tidydb, when checking for the continued existence of a message for
1580 which it has found a message-specific retry record, was not finding
1581 messages that were in split spool directories. Consequently, it was
1582 deleting retry records that should have stayed in existence.
1584 48. Steve fixed some bugs in eximstats.
1586 49. The SPA authentication driver was not abandoning authentication and moving
1587 on to the next authenticator when an expansion was forced to fail,
1588 contradicting the general specification for all authenticators. Instead it
1589 was generating a temporary error. It now behaves as specified.
1591 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1592 (the order specifies the preference for clients). The order is now AES256,
1593 AES128, 3DES, ARCFOUR128.
1595 51. Small patch to Sieve code - explicitly set From: when generating an
1598 52. Exim crashed if a remote delivery caused a very long error message to be
1599 recorded - for instance if somebody sent an entire SpamAssassin report back
1600 as a large number of 550 error lines. This bug was coincidentally fixed by
1601 increasing the size of one of Exim's internal buffers (big_buffer) that
1602 happened as part of the Exiscan merge. However, to be on the safe side, I
1603 have made the code more robust (and fixed the comments that describe what
1606 53. Now that there can be additional text after "Completed" in log lines (if
1607 the queue_time_overall log selector is set), a one-byte patch to exigrep
1608 was needed to allow it to recognize "Completed" as not the last thing in
1611 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1612 patch that reportedly fixes this has been added. I am not expert enough to
1613 create a test for it. This is what the patch creator wrote:
1615 "I found a little strange behaviour of ldap code when working with
1616 Windows 2003 AD Domain, where users was placed in more than one
1617 Organization Units. When I tried to give exim partial DN, the exit code
1618 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1619 But simultaneously result of request was absolutely normal ldap result,
1620 so I produce this patch..."
1622 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1623 REFERENCE, so I have modified the code to exclude the patch when that macro
1626 55. Some experimental protocols are using DNS PTR records for new purposes. The
1627 keys for these records are domain names, not reversed IP addresses. The
1628 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1629 leaves it alone. Component reversal etc. now happens only for IP addresses.
1632 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1634 57. Double the size of the debug message buffer (to 2048) so that more of very
1635 long debug lines gets shown.
1637 58. The exicyclog utility now does better if the number of log files to keep
1638 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1640 59. Two changes related to the smtp_active_hostname option:
1642 (1) $smtp_active_hostname is now available as a variable.
1643 (2) The default for smtp_banner uses $smtp_active_hostname instead
1644 of $primary_hostname.
1646 60. The host_aton() function is supposed to be passed a string that is known
1647 to be a valid IP address. However, in the case of IPv6 addresses, it was
1648 not checking this. This is a hostage to fortune. Exim now panics and dies
1649 if the condition is not met. A case was found where this could be provoked
1650 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1651 components; fortuitously, this particular loophole had already been fixed
1652 by change 4.50/55 above.
1654 If there are any other similar loopholes, the new check in host_aton()
1655 itself should stop them being exploited. The report I received stated that
1656 data on the command line could provoke the exploit when Exim was running as
1657 exim, but did not say which command line option was involved. All I could
1658 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1659 running as the user.
1662 61. There was a buffer overflow vulnerability in the SPA authentication code
1663 (which came originally from the Samba project). I have added a test to the
1664 spa_base64_to_bits() function which I hope fixes it.
1667 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1668 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1670 63. The daemon start-up calls getloadavg() while still root for those OS that
1671 need the first call to be done as root, but it missed one case: when
1672 deliver_queue_load_max is set with deliver_drop_privilege. This is
1673 necessary for the benefit of the queue runner, because there is no re-exec
1674 when deliver_drop_privilege is set.
1676 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1677 This has been fixed.
1679 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1680 was in use, was not putting the data itself into the right store pool;
1681 consequently, it could be overwritten for a subsequent message in the same
1682 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1685 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1687 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1688 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1689 the function were treating the return as a boolean value, which happened to
1690 work because 0=false and not-0=true, but is not correct code.
1692 68. The host_aton() function was not handling scoped IPv6 addresses (those
1693 with, for example, "%eth0" on the end) correctly.
1695 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1696 negated items (that is, ~something) in unsigned ints. Some compilers
1697 apparently mutter when there is no cast.
1699 70. If an address verification called from an ACL failed, and did not produce a
1700 user-specific message (i.e. there was only a "system" message), nothing was
1701 put in $acl_verify_message. In this situation, it now puts the system
1704 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1705 session; change 4.31/43 added the unwanted input to the log line - except
1706 that it did not do this in the start of session case. It now does.
1708 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1709 This is wrong and can cause the other end to generate a synchronization
1710 error if it is another Exim or anything else that does the synchronization
1711 check. A QUIT command is no longer sent after a timeout.
1713 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1714 during host lookups.
1716 74. The defer_ok option of callout verification was not working if it was used
1717 when verifying addresses in header lines, that is, for this case:
1719 verify = header_sender/callout=defer_ok
1721 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1722 those file descriptors could be used for SMTP connections. If anything
1723 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1724 could be sent to the SMTP client, causing chaos. The daemon now opens
1725 stdin, stdout, and stderr to /dev/null when it puts itself into the
1728 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1729 log by default. The user can override this with suitable Perl magic.
1731 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1732 the log message when discard triggers, was not working for the DATA ACL or
1733 for the non-SMTP ACL.
1735 78. Error message wording change in sieve.c.
1737 79. If smtp_accept_max_per_host was set, the number of connections could be
1738 restricted to fewer than expected, because the daemon was trying to set up
1739 a new connection before checking whether the processes handling previous
1740 connections had finished. The check for completed processes is now done
1741 earlier. On busy systems, this bug wouldn't be noticed because something
1742 else would have woken the daemon, and it would have reaped the completed
1745 80. If a message was submitted locally by a user whose login name contained one
1746 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1747 It caused a spool format error. I have fixed the spool reading code. A
1748 related problem was that the "from" clause in the Received: line became
1749 illegal because of the space(s). It is now covered by ${quote_local_part.
1751 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1754 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1756 83. Patch from Sieve maintainer to fix unterminated string problem in
1757 "vacation" handling.
1759 84. Some minor changes to the Linux configuration files to help with other
1760 OS variants using glibc.
1762 85. One more patch for Sieve to update vacation handling to latest spec.
1765 ----------------------------------------------------
1766 See the note above about the 4.44 and 4.50 releases.
1767 ----------------------------------------------------
1773 1. Change 4.43/35 introduced a bug that caused file counts to be
1774 incorrectly computed when quota_filecount was set in an appendfile
1777 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1778 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1781 3. Give more explanation in the error message when the command for a transport
1782 filter fails to execute.
1784 4. There are several places where Exim runs a non-Exim command in a
1785 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1786 was being done only for the command run by the queryprogram router. It is
1787 now done for all such subprocesses. The other cases are: ${run, transport
1788 filters, and the commands run by the lmtp and pipe transports.
1790 5. Some older OS have a limit of 256 on the maximum number of file
1791 descriptors. Exim was using setrlimit() to set 1000 as a large value
1792 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1793 systems. I've change it so that if it can't get 1000, it tries for 256.
1795 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1796 was an oversight, and furthermore, ever since the addition of extra
1797 controls (e.g. 4.43/32), the checks on when to allow different forms of
1798 "control" were broken. There should now be diagnostics for all cases when a
1799 control that does not make sense is encountered.
1801 7. $recipients is now available in the predata ACL (oversight).
1803 8. Tidy the search cache before the fork to do a delivery from a message
1804 received from the command line. Otherwise the child will trigger a lookup
1805 failure and thereby defer the delivery if it tries to use (for example) a
1806 cached ldap connection that the parent has called unbind on.
1808 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1809 of $address_data from the recipient verification was clobbered by the
1810 sender verification.
1812 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1813 was its contents. (It was OK if the option was not defined at all.)
1815 11. A "Completed" log line is now written for messages that are removed from
1816 the spool by the -Mrm option.
1818 12. $host_address is now set to the target address during the checking of
1819 ignore_target_hosts.
1821 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1822 being passed; this would have caused $sender_host_name to have been used if
1823 matching the list had actually called for a host name (not very likely,
1824 since this list is usually IP addresses). A host name is now passed as
1827 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1828 code that specifies a non-restarting handler (typically sigaction() in
1829 modern systems) in an attempt to fix a rare and obscure crash bug.
1831 15. Narrowed the window for a race in the daemon that could cause it to ignore
1832 SIGCHLD signals. This is not a major problem, because they are used only to
1833 wake it up if nothing else does.
1835 16. A malformed maildirsize file could cause Exim to calculate negative values
1836 for the mailbox size or file count. Odd effects could occur as a result.
1837 The maildirsize information is now recalculated if the size or filecount
1840 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1841 support for a long time. Removed HAVE_SYS_VFS_H.
1843 18. Updated exipick to current release from John Jetmore.
1845 19. Allow an empty sender to be matched against a lookup in an address list.
1846 Previously the only cases considered were a regular expression, or an
1849 20. Exim went into a mad DNS lookup loop when doing a callout where the
1850 host was specified on the transport, if the DNS lookup yielded more than
1853 21. The RFC2047 encoding function was originally intended for short strings
1854 such as real names; it was not keeping to the 75-character limit for
1855 encoded words that the RFC imposes. It now respects the limit, and
1856 generates multiple encoded words if necessary. To be on the safe side, I
1857 have increased the buffer size for the ${rfc2047: expansion operator from
1860 22. Failure to deliver a bounce message always caused it to be frozen, even if
1861 there was an errors_to setting on the router. The errors_to setting is now
1864 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1865 canonical form (fully expanded) before being placed in
1866 $sender_host_address.
1868 24. Updated eximstats to version 1.33
1870 25. Include certificate and key file names in error message when GnuTLS fails
1871 to set them up, because the GnuTLS error message doesn't include the name
1872 of the failing file when there is a problem reading it.
1874 26. Expand error message when OpenSSL has problems setting up cert/key files.
1877 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1878 (this can affect the format of dates).
1880 28. exim_tidydb, when checking for the continued existence of a message for
1881 which it has found a message-specific retry record, was not finding
1882 messages that were in split spool directories. Consequently, it was
1883 deleting retry records that should have stayed in existence.
1885 29. eximstats updated to version 1.35
1886 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1887 1.35 - bugfix such that pie charts by volume are generated correctly
1889 30. The SPA authentication driver was not abandoning authentication and moving
1890 on to the next authenticator when an expansion was forced to fail,
1891 contradicting the general specification for all authenticators. Instead it
1892 was generating a temporary error. It now behaves as specified.
1894 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1895 (the order specifies the preference for clients). The order is now AES256,
1896 AES128, 3DES, ARCFOUR128.
1898 31. Small patch to Sieve code - explicitly set From: when generating an
1901 32. Exim crashed if a remote delivery caused a very long error message to be
1902 recorded - for instance if somebody sent an entire SpamAssassin report back
1903 as a large number of 550 error lines. This bug was coincidentally fixed by
1904 increasing the size of one of Exim's internal buffers (big_buffer) that
1905 happened as part of the Exiscan merge. However, to be on the safe side, I
1906 have made the code more robust (and fixed the comments that describe what
1909 33. Some experimental protocols are using DNS PTR records for new purposes. The
1910 keys for these records are domain names, not reversed IP addresses. The
1911 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1912 leaves it alone. Component reversal etc. now happens only for IP addresses.
1915 34. The host_aton() function is supposed to be passed a string that is known
1916 to be a valid IP address. However, in the case of IPv6 addresses, it was
1917 not checking this. This is a hostage to fortune. Exim now panics and dies
1918 if the condition is not met. A case was found where this could be provoked
1919 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1920 components; fortuitously, this particular loophole had already been fixed
1921 by change 4.50/55 or 4.44/33 above.
1923 If there are any other similar loopholes, the new check in host_aton()
1924 itself should stop them being exploited. The report I received stated that
1925 data on the command line could provoke the exploit when Exim was running as
1926 exim, but did not say which command line option was involved. All I could
1927 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1928 running as the user.
1931 35. There was a buffer overflow vulnerability in the SPA authentication code
1932 (which came originally from the Samba project). I have added a test to the
1933 spa_base64_to_bits() function which I hope fixes it.
1936 36. The daemon start-up calls getloadavg() while still root for those OS that
1937 need the first call to be done as root, but it missed one case: when
1938 deliver_queue_load_max is set with deliver_drop_privilege. This is
1939 necessary for the benefit of the queue runner, because there is no re-exec
1940 when deliver_drop_privilege is set.
1942 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1943 was in use, was not putting the data itself into the right store pool;
1944 consequently, it could be overwritten for a subsequent message in the same
1945 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1948 38. Sometimes the final signoff response after QUIT could fail to get
1949 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1950 < 0 before doing a fflush(). This bug looks as though it goes back to the
1951 introduction of TLS in release 3.20, but "sometimes" must have been rare
1952 because the tests only now provoked it.
1958 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1959 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1960 Now it returns an integer. A number of calls were still expecting a T/F
1961 return. Fortuitously, in all cases, the tests worked in OK situations,
1962 which is the norm. However, things would have gone wrong on any write
1963 failures on the smtp file descriptor. This function is used when sending
1964 messages over SMTP and also when doing verify callouts.
1966 2. When Exim is called to do synchronous delivery of a locally submitted
1967 message (the -odf or -odi options), it no longer closes stderr before doing
1970 3. Implemented the mua_wrapper option.
1972 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1974 5. Implemented the functions header_remove(), header_testname(),
1975 header_add_at_position(), and receive_remove_recipient(), and exported them
1978 6. If an ACL "warn" statement specified the addition of headers, Exim already
1979 inserted X-ACL-Warn: at the start if there was no header name. However, it
1980 was not making this test for the second and subsequent header lines if
1981 there were newlines in the string. This meant that an invalid header could
1982 be inserted if Exim was badly configured.
1984 7. Allow an ACL "warn" statement to add header lines at the start or after all
1985 the Received: headers, as well as at the end.
1987 8. Added the rcpt_4xx retry error code.
1989 9. Added postmaster_mailfrom=xxx to callout verification option.
1991 10. Added mailfrom=xxxx to the callout verification option, for verify=
1994 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1995 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1997 12. Inserted some casts to stop certain compilers warning when using pointer
1998 differences as field lengths or precisions in printf-type calls (mostly
1999 affecting debugging statements).
2001 13. Added optional readline() support for -be (dynamically loaded).
2003 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2004 same clock tick as a message's arrival, so that its received time was the
2005 same as the "first fail" time on the retry record, and that message
2006 remained on the queue past the ultimate address timeout, every queue runner
2007 would try a delivery (because it was past the ultimate address timeout) but
2008 after another failure, the ultimate address timeout, which should have then
2009 bounced the address, did not kick in. This was a "< instead of <=" error;
2010 in most cases the first failure would have been in the next clock tick
2011 after the received time, and all would be well.
2013 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2014 being recognized when the domain list was tested by the match_domain
2015 condition in an expansion string.
2017 16. Added the ${str2b64: operator.
2019 17. Exim was always calling setrlimit() to set a large limit for the number of
2020 processes, without checking whether the existing limit was already
2021 adequate. (It did check for the limit on file descriptors.) Furthermore,
2022 errors from getrlimit() and setrlimit() were being ignored. Now they are
2023 logged to the main and panic logs, but Exim does carry on, to try to do its
2024 job under whatever limits there are.
2026 18. Imported PCRE 5.0.
2028 19. Trivial typo in log message " temporarily refused connection" (the leading
2031 20. If the log selector return_path_on_delivery was set and an address was
2032 redirected to /dev/null, the delivery process crashed because it assumed
2033 that a return path would always be set for a "successful" delivery. In this
2034 case, the whole delivery is bypassed as an optimization, and therefore no
2037 21. Internal re-arrangement: the function for sending a challenge and reading
2038 a response while authentication was assuming a zero-terminated challenge
2039 string. It's now changed to take a pointer and a length, to allow for
2040 binary data in such strings.
2042 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2044 23. Exim was not respecting finduser_retries when seeking the login of the
2045 uid under which it was called; it was always trying 10 times. (The default
2046 setting of finduser_retries is zero.) Also, it was sleeping after the final
2047 failure, which is pointless.
2049 24. Implemented tls_on_connect_ports.
2051 25. Implemented acl_smtp_predata.
2053 26. If the domain in control=submission is set empty, Exim assumes that the
2054 authenticated id is a complete email address when it generates From: or
2055 Sender: header lines.
2057 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2058 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2059 chown and chgrp in /bin and hostname in /usr/bin.
2061 28. Exim was keeping the "process log" file open after each use, just as it
2062 does for the main log. This opens the possibility of it remaining open for
2063 long periods when the USR1 signal hits a daemon. Occasional processlog
2064 errors were reported, that could have been caused by this. Anyway, it seems
2065 much more sensible not to leave this file open at all, so that is what now
2068 29. The long-running daemon process does not normally write to the log once it
2069 has entered its main loop, and it closes the log before doing so. This is
2070 so that log files can straightforwardly be renamed and moved. However,
2071 there are a couple of unusual error situations where the daemon does write
2072 log entries, and I had neglected to close the log afterwards.
2074 30. The text of an SMTP error response that was received during a remote
2075 delivery was being truncated at 512 bytes. This is too short for some of
2076 the long messages that one sometimes sees. I've increased the limit to
2079 31. It is now possible to make retry rules that apply only when a message has a
2080 specific sender, in particular, an empty sender.
2082 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2083 it possible to be selective about when SMTP synchronization is enforced.
2085 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2087 32. Implemented hosts_connection_nolog.
2089 33. Added an ACL for QUIT.
2091 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2094 35. Added mailbox_size and mailbox_filecount to appendfile.
2096 36. Added control = no_multiline_responses to ACLs.
2098 37. There was a bug in the logic of the code that waits for the clock to tick
2099 in the case where the clock went backwards by a substantial amount such
2100 that the microsecond fraction of "now" was more than the microsecond
2101 fraction of "then" (but the whole seconds number was less).
2103 38. Added support for the libradius Radius client library this is found on
2104 FreeBSD (previously only the radiusclient library was supported).
2110 1. When certain lookups returned multiple values in the form name=value, the
2111 quoting of the values was not always being done properly. Specifically:
2112 (a) If the value started with a double quote, but contained no whitespace,
2114 (b) If the value contained whitespace other than a space character (i.e.
2115 tabs or newlines or carriage returns) it was not quoted.
2116 This fix has been applied to the mysql and pgsql lookups by writing a
2117 separate quoting function and calling it from the lookup code. The fix
2118 should probably also be applied to nisplus, ibase and oracle lookups, but
2119 since I cannot test any of those, I have not disturbed their existing code.
2121 2. A hit in the callout cache for a specific address caused a log line with no
2122 reason for rejecting RCPT. Now it says "Previous (cached) callout
2123 verification failure".
2125 3. There was an off-by-one bug in the queryprogram router. An over-long
2126 return line was truncated at 256 instead of 255 characters, thereby
2127 overflowing its buffer with the terminating zero. As well as fixing this, I
2128 have increased the buffer size to 1024 (and made a note to document this).
2130 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2131 when Exim is waiting for an SMTP response from a remote server, Exim
2132 restarts its select() call on the socket, thereby resetting its timeout.
2133 This is not a problem when such interrupts are rare. Somebody set up a cron
2134 job to run exiwhat every 2 minutes, which is less than the normal select()
2135 timeout (5 or 10 minutes). This meant that the select() timeout never
2136 kicked in because it was always reset. I have fixed this by comparing the
2137 time when an interrupt arrives with the time at the start of the first call
2138 to select(). If more time than the timeout has elapsed, the interrupt is
2139 treated as a timeout.
2141 5. Some internal re-factoring in preparation for the addition of Sieve
2142 extensions (by MH). In particular, the "personal" test is moved to a
2143 separate function, and given an option for scanning Cc: and Bcc: (which is
2144 not set for Exim filters).
2146 6. When Exim created an email address using the login of the caller as the
2147 local part (e.g. when creating a From: or Sender: header line), it was not
2148 quoting the local part when it contained special characters such as @.
2150 7. Installed new OpenBSD configuration files.
2152 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2153 try to make them clearer.
2155 9. Callout options, other than the timeout value, were being ignored when
2156 verifying sender addresses in header lines. For example, when using
2158 verify = header_sender/callout=no_cache
2160 the cache was (incorrectly) being used.
2162 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2163 only the Cygwin environment.
2165 11. When return_path_on_delivery was set as a log selector, if different remote
2166 addresses in the same message used different return paths and parallel
2167 remote delivery occurred, the wrong values would sometimes be logged.
2168 (Whenever a remote delivery process finished, the return path value from
2169 the most recently started remote delivery process was logged.)
2171 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2172 lines when AUTH and/or TLS are in use. This is the "received protocol"
2173 field. Exim used to use "asmtp" for authenticated SMTP, without any
2174 indication (in the protocol name) for TLS use. Now it follows the RFC and
2175 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2176 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2177 names appear in log lines as well as in Received: header lines.
2179 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2180 extensions, and comparison tests, and to fix some bugs.
2182 14. Changes to the "personal" filter test:
2184 (1) The test was buggy in that it was just doing the equivalent of
2185 "contains" tests on header lines. For example, if a user's address was
2186 anne@some.where, the "personal" test would incorrectly be true for
2188 To: susanne@some.where
2190 This test is now done by extracting each address from the header in turn,
2191 and checking the entire address. Other tests that are part of "personal"
2192 are now done using regular expressions (for example, to check local parts
2193 of addresses in From: header lines).
2195 (2) The list of non-personal local parts in From: addresses has been
2196 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2197 taken from the Sieve specification recommendations.
2199 (3) If the message contains any header line starting with "List-" it is
2200 treated as non-personal.
2202 (4) The test for "circular" in the Subject: header line has been removed
2203 because it now seems ill-conceived.
2205 15. Minor typos in src/EDITME comments corrected.
2207 16. Installed latest exipick from John Jetmore.
2209 17. If headers_add on a router specified a text string that was too long for
2210 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2211 of string_sprintf() is now avoided.
2213 18. $message_body_size was not set (it was always zero) when running the DATA
2214 ACL and the local_scan() function.
2216 19. For the "mail" command in an Exim filter, no default was being set for
2217 the once_repeat time, causing a random time value to be used if "once" was
2218 specified. (If the value happened to be <= 0, no repeat happened.) The
2219 default is now 0s, meaning "never repeat". The "vacation" command was OK
2220 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2221 (I found it when inspecting the code).
2223 20. There is now an overall timeout for performing a callout verification. It
2224 defaults to 4 times the callout timeout, which applies to individual SMTP
2225 commands during the callout. The overall timeout applies when there is more
2226 than one host that can be tried. The timeout is checked before trying the
2227 next host. This prevents very long delays if there are a large number of
2228 hosts and all are timing out (e.g. when the network connections are timing
2229 out). The value of the overall timeout can be changed by specifying an
2230 additional sub-option for "callout", called "maxwait". For example:
2232 verify = sender/callout=5s,maxwait=20s
2234 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2235 to the end before writing, but this should make it even safer).
2237 22. Exim was forgetting that it had advertised PIPELINING for the second and
2238 subsequent messages on an SMTP connection. It was also not resetting its
2239 memory on STARTTLS and an internal HELO.
2241 23. When Exim logs an SMTP synchronization error within a session, it now
2242 records whether PIPELINING has been advertised or not.
2244 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2245 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2246 rather than long int.
2248 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2250 26. Added the never_mail option to autoreply.
2256 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2257 crash if the getsockname() call failed; this can happen if a connection is
2258 closed very soon after it is established. The problem was simply in the
2259 order in which certain operations were done, causing Exim to try to write
2260 to the SMTP stream before it had set up the file descriptor. The bug has
2261 been fixed by making things happen in the correct order.
2267 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2268 before the connection was closed, thus losing the rejection response.
2270 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2271 some early Solaris releases, but causes trouble in current releases where
2272 socklen_t is defined.
2274 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2277 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2280 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2281 that releasing the top of it at the end releases what was used for sub-
2282 expansions (unless the block got too big). However, discard this block if
2283 the first thing is a variable or header, so that we can use its block when
2284 it is dynamic (useful for very large $message_headers, for example).
2286 6. Lookups now cache *every* query, not just the most recent. A new, separate
2287 store pool is used for this. It can be recovered when all lookup caches are
2288 flushed. Lookups now release memory at the end of their result strings.
2289 This has involved some general refactoring of the lookup sources.
2291 7. Some code has been added to the store_xxx() functions to reduce the amount
2292 of flapping under certain conditions.
2294 8. log_incoming_interface used to affect only the <= reception log lines. Now
2295 it causes the local interface and port to be added to several more SMTP log
2296 lines, for example "SMTP connection from", and rejection lines.
2298 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2300 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2302 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2303 could be overwritten at the end of the current message (or the start of a
2304 new message if it was set in a HELO ACL). The value is now preserved for
2305 the duration of the SMTP connection.
2307 12. If a transport had a headers_rewrite setting, and a matching header line
2308 contained an unqualified address, that address was qualified, even if it
2309 did not match any rewriting rules. The underlying bug was that the values
2310 of the flags that permit the existence of unqualified sender and recipient
2311 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2312 non-local messages, and by -bnq for local messages) were not being
2313 preserved with the message after it was received.
2315 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2316 "next input=" as part of the text comprising the host identity instead of
2317 the correct text. The code was using the same buffer for two different
2318 strings. However, depending on which order the printing function evaluated
2319 its arguments, the bug did not always show up. Under Linux, for example, my
2320 test suite worked just fine.
2322 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2323 doesn't work with some older versions of Perl. It has been changed to "my",
2324 which in any case is probably the better facility to use.
2326 15. A really picky compiler found some instances of statements for creating
2327 error messages that either had too many or two few arguments for the format
2330 16. The size of the buffer for calls to the DNS resolver has been increased
2331 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2332 for addresses that have a lot of PTR records. This alleviates a problem; it
2333 does not fully solve it.
2335 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2336 buffer now truncates the list and logs the incident, which is the same
2337 action as happens when Exim is looking up a host name and its aliases.
2338 Previously in this situation something unpredictable would happen;
2339 sometimes it was "internal error: store_reset failed".
2341 18. If a server dropped the connection unexpectedly when an Exim client was
2342 using GnuTLS and trying to read a response, the client delivery process
2343 crashed while trying to generate an error log message.
2345 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2346 string, for example:
2348 warn message = H1: something\nH2: something
2350 the text was added as a single header line from Exim's point of view
2351 though it ended up OK in the delivered message. However, searching for the
2352 second and subsequent header lines using $h_h2: did not work. This has been
2353 fixed. Similarly, if a system filter added multiple headers in this way,
2354 the routers could not see them.
2356 20. Expanded the error message when iplsearch is called with an invalid key to
2357 suggest using net-iplsearch in a host list.
2359 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2360 ACLs are no longer actually imposed (and a message to that effect is
2363 22. If a "gecos" field in a passwd entry contained escaped characters, in
2364 particular, if it contained a \" sequence, Exim got it wrong when building
2365 a From: or a Sender: header from that name. A second bug also caused
2366 incorrect handling when an unquoted " was present following a character
2367 that needed quoting.
2369 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2370 was not being matched caselessly.
2372 24. Arranged for all hyphens in the exim.8 source to be escaped with
2375 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2376 properly. Recipient callout cache records were still being keyed to include
2377 the sender, even when use_sender was set false. This led to far more
2378 callouts that were necessary. The sender is no longer included in the key
2379 when use_sender is false.
2381 26. Added "control = submission" modifier to ACLs.
2383 27. Added the ${base62d: operator to decode base 62 numbers.
2385 28. dnsdb lookups can now access SRV records.
2387 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2388 the configuration file.
2390 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2391 (-v) mode. This makes the output for a verbose queue run more intelligible.
2393 31. Added a use_postmaster feature to recipient callouts.
2395 32. Added the $body_zerocount variable, containing the number of binary zero
2396 bytes in the message body.
2398 33. The time of last modification of the "new" subdirectory is now used as the
2399 "mailbox time last read" when there is a quota error for a maildir
2402 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2404 35. Added +ignore_unknown as a special item in host lists.
2406 36. Code for decoding IPv6 addresses in host lists is now included, even if
2407 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2408 address was recognized as an IP address, but was then not correctly decoded
2409 into binary, causing unexpected and incorrect effects when compared with
2416 1. Very minor rewording of debugging text in manualroute to say "list of
2417 hosts" instead of "hostlist".
2419 2. If verify=header_syntax was set, and a header line with an unqualified
2420 address (no domain) and a large number of spaces between the end of the
2421 name and the colon was received, the reception process suffered a buffer
2422 overflow, and (when I tested it) crashed. This was caused by some obsolete
2423 code that should have been removed. The fix is to remove it!
2425 3. When running in the test harness, delay a bit after writing a bounce
2426 message to get a bit more predictability in the log output.
2428 4. Added a call to search_tidyup() just before forking a reception process. In
2429 theory, someone could use a lookup in the expansion of smtp_accept_max_
2430 per_host which, without the tidyup, could leave open a database connection.
2432 5. Added the variables $recipient_data and $sender_data which get set from a
2433 lookup success in an ACL "recipients" or "senders" condition, or a router
2434 "senders" option, similar to $domain_data and $local_part_data.
2436 6. Moved the writing of debug_print from before to after the "senders" test
2439 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2440 problems for message scanning, either using a data ACL, or using
2441 local_scan() because the Received: header was not generated till after they
2442 were called (in order to set the time as the time of reception completion).
2443 I have revised the way this works. The header is now generated after the
2444 body is received, but before the ACL or local_scan() are called. After they
2445 are run, the timestamp in the header is updated.
2451 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2452 before starting a queue runner without re-exec. This happened only when
2453 deliver_drop_privilege was set or when the Exim user was set to root. The
2454 effect of the bug was that timeouts during subsequent deliveries caused
2455 crashes instead of being properly handled. The handler is now left at its
2456 default (and expected) setting.
2458 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2459 message, again when deliver_drop_privilege is set or Exim is run as root.
2460 The bug described in (1) was not present in this case, but the tidying up
2461 of the other signals was missing. I have made the two cases consistent.
2463 3. The ignore_target_hosts setting on a manualroute router was being ignored
2464 for hosts that were looked up using the /MX notation.
2466 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2469 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2470 operated on the sender address. After changing the $sender_address to <>
2471 for the sender address verify, Exim was re-instated it as the original
2472 (before rewriting) address, but remembering that it had rewritten it, so it
2473 wasn't rewriting it again. This bug also had the effect of breaking the
2474 sender address verification caching when the sender address was rewritten.
2476 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2477 This has been changed so that if the ip literal address matches
2478 ignore_target_hosts, the router declines.
2480 7. Added expansion conditions match_domain, match_address, and match_local_
2481 part (NOT match_host).
2483 8. The placeholder for the Received: header didn't have a length field set.
2485 9. Added code to Exim itself and to exim_lock to test for a specific race
2486 condition that could lead to file corruption when using MBX delivery. The
2487 issue is with the lockfile that is created in /tmp. If this file is removed
2488 after a process has opened it but before that process has acquired a lock,
2489 there is the potential for a second process to recreate the file and also
2490 acquire a lock. This could lead to two Exim processes writing to the file
2491 at the same time. The added code performs the same test as UW imapd; it
2492 checks after acquiring the lock that its file descriptor still refers to
2493 the same named file.
2495 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2496 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2499 11. Added the smtp_active_hostname option. If used, this will typically be made
2500 to depend on the incoming interface address. Because $interface_address is
2501 not set up until the daemon has forked a reception process, error responses
2502 that can happen earlier (such as "too many connections") no longer contain
2505 12. If an expansion in a condition on a "warn" statement fails because a lookup
2506 defers, the "warn" statement is abandoned, and the next ACL statement is
2507 processed. Previously this caused the whole ACL to be aborted.
2509 13. Added the iplsearch lookup type.
2511 14. Added ident_timeout as a log selector.
2513 15. Added tls_certificate_verified as a log selector.
2515 16. Added a global option tls_require_ciphers (compare the smtp transport
2516 option of the same name). This controls incoming TLS connections.
2518 17. I finally figured out how to make tls_require_ciphers do a similar thing
2519 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2520 before starting the TLS session.
2522 18. Tabs are now shown as \t in -bP output.
2524 19. If the log selector return_path_on_delivery was set, Exim crashed when
2525 bouncing a message because it had too many Received: header lines.
2527 20. If two routers both had headers_remove settings, and the first one included
2528 a superfluous trailing colon, the final name in the first list and the
2529 first name in the second list were incorrectly joined into one item (with a
2530 colon in the middle).
2536 1. Added -C and -D options to the exinext utility, mainly to make it easier
2537 to include in the automated testing, but these could be helpful when
2538 multiple configurations are in use.
2540 2. The exinext utility was not formatting the output nicely when there was
2541 an alternate port involved in the retry record key, nor when there was a
2542 message id as well (for retries that were specific to a specific message
2543 and a specific host). It was also confused by IPv6 addresses, because of
2544 the additional colons they contain. I have fixed the IPv4 problem, and
2545 patched it up to do a reasonable job for IPv6.
2547 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2548 delivery, the log line now contains "pipelined" if PIPELINING was used.
2550 4. An SMTP transport process used to panic and die if the bind() call to set
2551 an explicit outgoing interface failed. This has been changed; it is now
2552 treated in the same way as a connect() failure.
2554 5. A reference to $sender_host_name in the part of a conditional expansion
2555 that was being skipped was still causing a DNS lookup. This no longer
2558 6. The def: expansion condition was not recognizing references to header lines
2559 that used bh_ and bheader_.
2561 7. Added the _cache feature to named lists.
2563 8. The code for checking quota_filecount in the appendfile transport was
2564 allowing one more file than it should have been.
2566 9. For compatibility with Sendmail, the command line option
2574 and sets the incoming protocol and host name (for trusted callers). The
2575 host name and its colon can be omitted when only the protocol is to be set.
2576 Note the Exim already has two private options, -pd and -ps, that refer to
2577 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2578 "s", but I don't think that's a major issue.
2580 10. A number of refactoring changes to the code, none of which should affect
2583 (a) The number of logging options was getting close to filling up the
2584 32-bit word that was used as a bit map. I have split them into two classes:
2585 those that are passed in the argument to log_write(), and those that are
2586 only ever tested independently outside of that function. These are now in
2587 separate 32-bit words, so there is plenty of room for expansion again.
2588 There is no change in the user interface or the logging behaviour.
2590 (b) When building, for example, log lines, the code previously used a
2591 macro that called string_cat() twice, in order to add two strings. This is
2592 not really sufficiently general. Furthermore, there was one instance where
2593 it was actually wrong because one of the argument was used twice, and in
2594 one call a function was used. (As it happened, calling the function twice
2595 did not affect the overall behaviour.) The macro has been replaced by a
2596 function that can join an arbitrary number of extra strings onto a growing
2599 (c) The code for expansion conditions now uses a table and a binary chop
2600 instead of a serial search (which was left over from when there were very
2601 few conditions). Also, it now recognizes conditions like "pam" even when
2602 the relevant support is not compiled in: a suitably worded error message is
2603 given if an attempt is made to use such a condition.
2605 11. Added ${time_interval:xxxxx}.
2607 12. A bug was causing one of the ddress fields not to be passed back correctly
2608 from remote delivery subprocesses. The field in question was not being
2609 subsequently used, so this caused to problems in practice.
2611 13. Added new log selectors queue_time and deliver_time.
2613 14. Might have fixed a bug in maildirsizefile handling that threw up
2614 "unexpected character" debug warnings, and recalculated the data
2615 unnecessarily. In any case, I expanded the warning message to give more
2618 15. Added the message "Restricted characters in address" to the statements in
2619 the default ACL that block characters like @ and % in local parts.
2621 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2622 Three changes have been made:
2624 (a) There was a serious bug; a negative response to MAIL caused the whole
2625 recipient domain to be cached as invalid, thereby blocking all messages
2626 to all local parts at the same domain, from all senders. This bug has
2627 been fixed. The domain is no longer cached after a negative response to
2628 MAIL if the sender used is not empty.
2630 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2633 (c) A new callout option, "use_sender" has been added for people who want
2634 the modified behaviour.
2640 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2643 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2644 indeed breaks things for older releases.
2646 3. Added additional logging to the case where there is a problem reading data
2647 from a filter that is running in a subprocess using a pipe, in order to
2648 try to track down a specific problem.
2650 4. Testing facility fudge: when running in the test harness and attempting
2651 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2652 getting "No route to host". Convert this to a timeout.
2654 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2657 6. Some OS don't have socklen_t but use size_t instead. This affects the
2658 fifth argument of getsockopt() amongst other things. This is now
2659 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2660 can be set for individual OS. I have set it for SunOS5, OSF1, and
2661 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2662 some earlier ones do not.
2664 7. Change 4.30/15 was not doing the test caselessly.
2666 8. The standard form for an IPv6 address literal was being rejected by address
2667 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2668 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2669 this, as well as the form without the "IPv6" on the front (but only when
2670 address literals are enabled, of course).
2672 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2674 10. Exim crashed if a message with an empty sender address specified by -f
2675 encountered a router with an errors_to setting. This could be provoked only
2676 by a command such as
2680 where an empty string was supplied; "<>" did not hit this bug.
2682 11. Installed PCRE release 4.5.
2684 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2685 remained set. It is now erased.
2687 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2688 times from message ids (which are base 36 rather than the normal 62).
2690 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2691 were being counted as actual protocol errors, and logged if the log
2692 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2693 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2694 and DATA following a set of rejected RCPTs do not count as protocol errors.
2695 In other words, Exim assumes they were pipelined, though this may not
2696 actually be the case. Of course, in all cases the client gets an
2697 appropriate error code.
2699 15. If a lookup fails in an ACL condition, a message about the failure may
2700 be available; it is used if testing the ACL cannot continue, because most
2701 such messages specify what the cause of the deferral is. However, some
2702 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2703 that caused an old message to be retained and used if a later statement
2704 caused a defer, replacing the real cause of the deferral.
2706 16. If an IP address had so many PTR records that the DNS lookup buffer
2707 was not large enough to hold them, Exim could crash while trying to process
2708 the truncated data. It now detects and logs this case.
2710 17. Further to 4.21/58, another change has been made: if (and only if) the
2711 first line of a message (the first header line) ends with CRLF, a bare LF
2712 in a subsequent header line has a space inserted after it, so as not to
2713 terminate the header.
2715 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2716 unnecessarily, used atoi() instead of strtol(), and didn't check the
2717 termination when getting file sizes from file names by regex.
2719 19. Completely re-implemented the support for maildirsize files, in the light
2720 of a number of problems with the previous contributed implementation
2721 (4.30/29). In particular:
2723 . If the quota is zero, the maildirsize file is maintained, but no quota is
2726 . If the maildir directory does not exist, it is created before any attempt
2727 to write a maildirsize file.
2729 . The quota value in the file is just a cache; if the quota is changed in
2730 the transport, the new value overrides.
2732 . A regular expression is available for excluding directories from the
2735 20. The autoreply transport checks the characters in options that define the
2736 message's headers; it allows continued headers, but it was checking with
2737 isspace() after an embedded newline instead of explicitly looking for a
2740 21. If all the "regular" hosts to which an address was routed had passed their
2741 expiry times, and had not reached their retry times, the address was
2742 bounced, even if fallback hosts were defined. Now Exim should go on to try
2745 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2746 equivalent code in the SMTP transport. Some hosts send humungous responses
2747 to HELO/EHLO, more than 1024 it seems.
2749 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2750 gives compiler warnings in some environments. I've now done it "properly",
2753 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2754 (because of problems with the built-in one) was declared to return uschar *
2755 instead of char *, causing compiler failure.
2757 25. Fixed a file descriptor leak when processing alias/forward files.
2759 26. Fixed a minor format string issue in dbfn.c.
2761 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2763 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2764 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2767 29. When a sender address is verified, it is cached, to save repeating the test
2768 when there is more than one recipient in a message. However, when the
2769 verification involves a callout, it is possible for different callout
2770 options to be set for different recipients. It is too complicated to keep
2771 track of this in the cache, so now Exim always runs a verification when a
2772 callout is required, relying on the callout cache for the optimization.
2773 The overhead is duplication of the address routing, but this should not be
2776 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2777 to be verified with callout=postmaster, and the main callout worked but the
2778 postmaster check failed, the verification correctly failed. However, if a
2779 subsequent RCPT command asked for sender verification *without* the
2780 postmaster check, incorrect caching caused this verification also to fail,
2783 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2784 it was not caching the DNS options (qualify_single, search_parents) that
2785 were used when the lookup failed. A subsequent lookup with different
2786 options therefore always gave the same answer, though there were cases
2787 where it should not have. (Example: a "domains = !$mx_any" option on a
2788 dnslookup router: the "domains" option is always processed without any
2789 widening, but the router might have qualify_single set.) Now Exim uses the
2790 cached value only when the same options are set.
2792 32. Added John Jetmore's "exipick" utility to the distribution.
2794 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2795 than a timeout (e.g. a certificate is required, and is not provided), an
2796 Exim server now closes the connection immediately. Previously it waited for
2797 the client to close - but if the client is SSL, it seems that they each
2798 wait for each other, leading to a delay before one of them times out.
2800 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2801 maintained 0.8.x compatibility because I don't think many are using it, and
2802 it is clearly obsolete.
2804 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2807 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2808 client certificate was expired. A simple patch fixes this, though I don't
2809 understand the full logic of why the verify callback is called multiple
2812 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2813 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2814 which causes problems with some clients (such as the Certicom SSL Plus
2815 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2816 disables the coutermeasure allowing Eudora to connect."
2818 38. Exim was not checking that a write() to a log file succeeded. This could
2819 lead to Bad Things if a log got too big, in particular if it hit a file
2820 size limit. Exim now panics and dies if it cannot write to a log file, just
2821 as it does if it cannot open a log file.
2823 39. Modified OS/Makefile-Linux so that it now contains
2825 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2827 The two -D definitions ensure that Exim is compiled with large file
2828 support, which makes it possible to handle log files that are bigger than
2831 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2832 instance) a domain was checked against a named list that involved a lookup,
2833 causing $domain_data to be set, then another domain was checked against the
2834 same list, then the first domain was re-checked, the value of $domain_data
2835 after the final check could be wrong. In particular, if the second check
2836 failed, it could be set empty. This bug probably also applied to
2839 41. The strip_trailing_dot option was not being applied to the address given
2840 with the -f command-line option.
2842 42. The code for reading a message's header from the spool was incrementing
2843 $received_count, but never initializing it. This meant that the value was
2844 incorrect (doubled) while delivering a message in the same process in which
2845 it was received. In the most common configuration of Exim, this never
2846 happens - a fresh exec is done - but it can happen when
2847 deliver_drop_privilege is set.
2849 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2850 it now includes up to 150 characters of the unexpected data in the log
2853 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2854 and building data strings. The size of both of these buffers was 10 000
2855 bytes - far larger than anybody would *ever* want, thought I. Needless to
2856 say, somebody hit the limit. I have increased the maximum line length to
2857 20 000 and the maximum data length of concatenated lines to 100 000. I have
2858 also fixed two bugs, because there was no checking on these buffers. Tsk,
2859 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2860 buffer is too small.
2862 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2863 lsearch lookups. Now it does.
2865 46. When parsing a route_list item in a manualroute router, a fixed-length
2866 buffer was used for the list of hosts. I made this 1024 bytes long,
2867 thinking that nobody would ever have a list of hosts that long. Wrong.
2868 Somebody had a whole pile of complicated expansion conditions, and the
2869 string was silently truncated, leading to an expansion error. It turns out
2870 that it is easier to change to an unlimited length (owing to other changes
2871 that have happened since this code was originally written) than to build
2872 structure for giving a limitation error. The length of the item that
2873 expands into the list of hosts is now unlimited.
2875 47. The lsearch lookup could not handle data where the length of text line was
2876 more than 4095 characters. Such lines were truncated, leading to shortened
2877 data being returned. It should now handle lines of any length.
2879 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2880 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2883 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2884 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2885 no longer happens in comment lines. A list of replacements is now placed
2886 at the head of all of the source files, except those whose only change is
2887 to replace PERL_COMMAND in the very first #! line.
2889 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2890 messages on the queue, with a bottom-up merge sort, using code contributed
2891 by Michael Haardt. This should make operations like -bp somewhat faster on
2892 large queues. It won't affect queue runners, except when queue_run_in_order
2895 51. Installed eximstats 1.31 in the distribution.
2897 52. Added support for SRV lookups to the dnslookup router.
2899 53. If an ACL referred to $message_body or $message_body_end, the value was not
2900 reset for any messages that followed in the same SMTP session.
2902 54. The store-handling optimization for building very long strings was not
2903 differentiating between the different store pools. I don't think this
2904 actually made any difference in practice, but I've tidied it.
2906 55. While running the routers to verify a sender address, $sender_address
2907 was still set to the sender address. This is wrong, because when routing to
2908 send a bounce to the sender, it would be empty. Therefore, I have changed
2909 it so that, while verifying a sender address, $sender_address is set to <>.
2910 (There is no change to what happens when verifying a recipient address.)
2912 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2913 target A or AAAA records (if not already returned) without resetting the
2914 qualify_single or search_parents options of the DNS resolver. These are
2915 inappropriate in this case because the targets of MX and SRV records must
2916 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2917 target that, when qualified, matched something in the local domain. These
2918 two options are now turned off when doing these lookups.
2920 57. It seems that at least some releases of Reiserfs (which does not have the
2921 concept of a fixed number of inodes) returns zero and not -1 for the
2922 number of available inodes. This interacted badly with check_spool_inodes,
2923 which assumed that -1 was the "no such thing" setting. What I have done is
2924 to check that the total number of inodes is greater than zero before doing
2925 the test of how many are available.
2927 58. When a "warn" ACL statement has a log_message modifier, the message is
2928 remembered, and not repeated. This is to avoid a lot of repetition when a
2929 message has many recipients that cause the same warning to be written.
2930 Howewer, Exim was preserving the list of already written lines for an
2931 entire SMTP session, which doesn't seem right. The memory is now reset if a
2932 new message is started.
2934 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2935 debugging output unless log_rewrite was also set.
2937 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2938 of (int)(handle) when we know that handle contains (void *)(-1).
2940 61. The Exim daemon panic-logs an error return when it closes the incoming
2941 connection. However "connection reset by peer" seems to be common, and
2942 isn't really an error worthy of noting specially, so that particular error
2945 62. When Exim is trying to find all the local interfaces, it used to panic and
2946 die if the ioctl to get the interface flags failed. However, it seems that
2947 on at least one OS (Solaris 9) it is possible to have an interface that is
2948 included in the list of interfaces, but for which you get a failure error
2949 for this call. This happens when the interface is not "plumbed" into a
2950 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2951 failure of the "get flags" call assumes that the interface is down.
2953 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2954 makes life easier for people who are doing arithmetic on fields extracted
2955 from dates, where you often get leading zeros that should not be
2956 interpreted as octal.
2958 64. Added qualify_domain to the redirect router, to override the global
2961 65. If a pathologically long header line contained very many addresses (the
2962 report of this problem mentioned 10 000) and each of them was rewritten,
2963 Exim could use up a very large amount of memory. (It kept on making new
2964 copies of the header line as it rewrote, and never released the old ones.)
2965 At the expense of a bit more processing, the header rewriting function has
2966 been changed so that it no longer eats memory in this way.
2968 66. The generation of the Received: header has been moved from the time that a
2969 message starts to be received, to the time that it finishes. The timestamp
2970 in the Received: header should now be very close to that of the <= log
2971 line. There are two side-effects of this change:
2973 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2974 logged header lines no longer include the local Received: line, because
2975 it has not yet been created. The same applies to a copy of the message
2976 that is returned to a non-SMTP sender when a message is rejected.
2978 (b) When a filter file is tested using -bf, no additional Received: header
2979 is added to the test message. After some thought, I decided that this
2982 This change does not affect the value of $received_for. It is still set
2983 after address rewriting, but before local_scan() is called.
2985 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2987 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2988 gave an unhelpful panic error message, and a defer error. I have managed to
2989 change this behaviour so that it now rejects any supplied certificate,
2990 which seems right, as the list of acceptable certificates is empty.
2992 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2993 gave an unhelpful defer error. I have not managed to make this reject any
2994 supplied certificates, but the error message it gives is "no certificate
2995 supplied", which is not helpful.
2997 70. exigrep's output now also includes lines that are not associated with any
2998 message, but which match the given pattern. Implemented by a patch from
2999 Martin Sluka, which also tidied up the Perl a bit.
3001 71. Recipient callout verification, like sender verification, was using <> in
3002 the MAIL FROM command. This isn't really the right thing, since the actual
3003 sender may affect whether the remote host accepts the recipient or not. I
3004 have changed it to use the actual sender in the callout; this means that
3005 the cache record is now keyed on a recipient/sender pair, not just the
3006 recipient address. There doesn't seem to be a real danger of callout loops,
3007 since a callout by the remote host to check the sender would use <>.
3008 [SEE ABOVE: changed after hitting problems.]
3010 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3011 temporary errors. However, in the case of such a code being given after
3012 the end of a data transmission (i.e. after ".") Exim was failing to write
3013 a retry record for the message. (Yes, there was some broken host that was
3014 actually sending 8xx at this point.)
3016 73. An unknown lookup type in a host list could cause Exim to panic-die when
3017 the list was checked. (An example that provoked this was putting <; in the
3018 middle of a list instead of at the start.) If this happened during a DATA
3019 ACL check, a -D file could be left lying around. This kind of configuration
3020 error no longer causes Exim to die; instead it causes a defer errror. The
3021 incident is still logged to the main and panic logs.
3023 74. Buglet left over from Exim 3 conversion. The message "too many messages
3024 in one connection" was written to the rejectlog but not the mainlog, except
3025 when address rewriting (yes!) was being logged.
3027 75. Added write_rejectlog option.
3029 76. When a system filter was run not as root (that is, when system_filter_user
3030 was set), the values of the $n variables were not being returned to the
3031 main process; thus, they were not subsequently available in the $sn
3034 77. Added +return_path_on_delivery log selector.
3036 78. A connection timeout was being treated differently from recipients deferred
3037 when testing hosts_max_try with a message that was older than the host's
3038 retry timeout. (The host should not be counted, thus allowing all hosts to
3039 be tried at least once before bouncing.) This may have been the cause of an
3040 occasionally reported bug whereby a message would remain on the queue
3041 longer than the retry timeout, but would be bounced if a delivery was
3042 forced. I say "may" because I never totally pinned down the problem;
3043 setting up timeout/retry tests is difficult. See also the next item.
3045 79. The ultimate address timeout was not being applied to errors that involved
3046 a combination of host plus message (for example, a timeout on a MAIL
3047 command). When an address resolved to a number of possible hosts, and they
3048 were not all tried for each delivery (e.g. because of hosts_max_try), a
3049 message could remain on the queue longer than the retry timeout.
3051 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3054 81. Fixed an obscure SMTP outgoing bug which required at least the following
3055 conditions: (a) there was another message waiting for the same server;
3056 (b) the server returned 5xx to all RCPT commands in the first message so
3057 that the message was not completed; (c) the server dropped the connection
3058 or gave a negative response to the RSET that Exim sends to abort the
3059 transaction. The observed case was a dropped connection after DATA that had
3060 been sent in pipelining mode. That is, the server had advertised PIPELINING
3061 but was not implementing it correctly. The effect of the bug was incorrect
3062 behaviour, such as trying another host, and this could lead to a crash.
3068 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3069 and daemon.c were passed as pointers to ints; they should have been
3070 pointers to socklen_t variables (which are typically unsigned ints).
3072 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3075 3. Fixed a really odd bug that affected only the testing scheme; patching a
3076 certain fixed string in the binary changed the value of another string that
3077 happened to be identical to the end of the original first string.
3079 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3080 name", it returns that address as the IP address. On some operating
3081 systems (e.g. Solaris), it also passes back the IP address string as the
3082 "host name". However, on others (e.g. Linux), it passes back an empty
3083 string. Exim wasn't checking for this, and was changing the host name to an
3084 empty string, assuming it had been canonicized.
3086 5. Although rare, it is permitted to have more than one PTR record for a given
3087 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3088 all the names associated with an address, because they do in Solaris.
3089 However, it seems that they do not in Linux for data that comes from the
3090 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3091 I found this out when I moved to a new Linux workstation and tried to run
3092 the Exim test suite.
3094 To get round this problem I have changed the code so that it now does its
3095 own call to the DNS to look up PTR records when searching for a host name.
3096 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3097 addresses that are only in /etc/hosts are still found.
3099 This behaviour is, however, controlled by an option called host_lookup_
3100 order, which defaults to "bydns:byaddr". If people want to use the other
3101 order, or indeed, just use one or the other means of lookup, they can
3102 specify it in this variable.
3104 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3105 some operating systems, this comes back from gethostbyaddr() as an empty
3106 string, and this is what Exim used to test for. However, it seems that in
3107 other systems, "." is yielded. Exim now tests for this case too.
3109 7. The values of check_spool_space and check_log_space are now held internally
3110 as a number of kilobytes instead of an absolute number of bytes. If a
3111 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3112 kilobyte. This means that much larger values can be stored.
3114 8. Exim monitor: an attempt to get the action menu when not actually pointing
3115 at a message produces an empty menu entitled "No message selected". This
3116 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3117 no entries in it ("Shell widget menu has zero width and/or height"). So I
3118 have added a single, blank menu entry in this case.
3120 9. Added ${quote_local_part.
3122 10. MIME decoding is now applied to the contents of Subject: header lines when
3125 11. Now that a reference to $sender_host_address automatically causes a reverse
3126 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3127 host lookup before query-style lookups in lists that might use this
3128 variable. This has therefore been abolished, and the "net-" prefix is no
3129 longer necessary for query-style lookups.
3131 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3132 have been a typo for LFLAGS, so it has been changed.
3134 13. The install script calls Exim with "-C /dev/null" in order to find the
3135 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3136 to be output. Howeve, since Exim outputs its version number before the
3137 error, it didn't break the script. It just looked ugly. I fixed this by
3138 always allowing "-C /dev/null" if the caller is root.
3140 14. Ignore overlarge ACL variable number when reading spool file - insurance
3141 against a later release with more variables having written the file.
3143 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3144 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3145 this, as well as the form without the "IPv6" on the front.
3147 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3148 OS/Makefile-Darwin file.
3150 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3151 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3153 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3154 variable. However, some OS use ints for inodes. I've added cast to long int
3155 to get rid of the compiler warning.
3157 19. I had forgotten to lock out "/../" in configuration file names when
3158 ALT_CONFIG_PREFIX was set.
3160 20. Routers used for verification do not need to specify transports. However,
3161 if such a router generated a host list, and callout was configured, Exim
3162 crashed, because it could not find a port number from the (non-existent)
3163 transport. It now assumes port 25 in this circumstance.
3165 21. Added the -t option to exigrep.
3167 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3168 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3169 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3170 them is set, LOOKUP_LSEARCH is forced.
3172 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3174 24. Added sender and host information to the "rejected by local_scan()" log
3175 line; previously there was no indication of these.
3177 25. Added .include_if_exists.
3179 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3180 Linux. It turns out that not all file systems support this. Apparently some
3181 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3182 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3183 not supported on the file descriptor, is now ignored when Exim is trying to
3184 sync a directory. This applies only to Linux.
3186 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3188 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3189 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3192 29. Added support for maildirsize files from supplied patch (modified a bit).
3194 30. The use of :fail: followed by an empty string could lead Exim to respond to
3195 sender verification failures with (e.g.):
3197 550 Verification failed for <xxx>
3198 550 Sender verify failed
3200 where the first response line was missing the '-' that indicates it is not
3201 the final line of the response.
3203 31. The loop for finding the name of the user that called Exim had a hardwired
3204 limit of 10; it now uses the value of finduser_retries, which is used for
3205 all other user lookups.
3207 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3210 33. Exim was neglecting to zero errno before one call of strtol() when
3211 expanding a string and expecting an integer value. On some systems this
3212 resulted in spurious "integer overflow" errors. Also, it was casting the
3213 result into an int without checking.
3215 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3216 did not work. The code looks as if it has *never* worked, though it appears
3217 to have been documented since at least releast 1.62. I have made it work.
3219 35. The "timeout_DNS" error in retry rules, also documented since at least
3220 1.62, also never worked. As it isn't clear exactly what this means, and
3221 clearly it isn't a major issue, I have abolished the feature by treating it
3222 as "timeout", and writing a warning to the main and panic logs.
3224 36. The display of retry rules for -brt wasn't always showing the error code
3227 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3228 timeout_connect_A, timeout_connect_MX.
3230 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3231 to the empty sender.
3233 39. The daemon was not analysing the content of -oX till after it had closed
3234 stderr and disconnected from the controlling terminal. This meant that any
3235 syntax errors were only noted on the panic log, and the return code from
3236 the command was 0. By re-arranging the code a little, I've made the
3237 decoding happen first, so such errors now appear on stderr, and the return
3238 code is 1. However, the actual setting up of the sockets still happens in
3239 the disconnected process, so errors there are still only recorded on the
3242 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3243 connections (as happens on some IP stacks) was logged at start up time as
3244 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3245 it from "IPv6 and IPv4", which means that two separate sockets are being
3248 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3249 says whether AF_INET or AF_INET6 was passed as an argument.
3251 42. Exiwhat output was messed up when time zones were included in log
3254 43. Exiwhat now gives more information about the daemon's listening ports,
3255 and whether -tls-on-connect was used.
3257 44. The "port" option of the smtp transport is now expanded.
3259 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3260 silently ignored. Now an error message is written to the main and panic
3263 46. There's a new ACL modifier called "logwrite" which writes to a log file
3264 as soon as it is encountered.
3266 47. Added $local_user_uid and $local_user_gid at routing time.
3268 48. Exim crashed when trying to verify a sender address that was being
3271 49. Exim was recognizing only a space character after ".include". It now also
3272 recognizes a tab character.
3274 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3275 extracting the relevant information from the specification. The man page no
3276 longer contains scrambled data for the -d option, and I've added a section
3277 at the front about calling Exim under different names.
3279 51. Added "extra_headers" argument to the "mail" command in filter files.
3281 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3284 53. Installed eximstats 1.29.
3286 54. Added transport_filter_timeout as a generic transport option.
3288 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3289 Cc: header lines. This was required by RFC 822, but it not required by RFC
3292 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3293 incoming messages that did not have them. Now it does so only if the
3294 message originates locally, that is, if there is no associated remote host
3295 address. When Resent- header lines are present, this applies to the Resent-
3296 lines rather than the non-Resent- lines.
3298 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3299 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3301 58. Messages for configuration errors now include the name of the main
3302 configuration file - useful now that there may be more than one file in a
3303 list (.included file names were always shown).
3305 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3306 for those rare installations that do not start the daemon as root or run it
3307 setuid root. I've cut out the call to initgroups() if the daemon is not
3310 60. The Exim user and group can now be bound into the binary as text strings
3311 that are looked up at the start of Exim's processing.
3313 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3315 62. Added $mailstore_basename variable.
3317 63. Installed patch to sieve.c from Michael Haardt.
3319 64. When Exim failed to open the panic log after failing to open the main log,
3320 the original message it was trying to log was written to stderr and debug
3321 output, but if they were not available (the usual case in production), it
3322 was lost. Now it is written to syslog before the two lines that record the
3323 failures to open the logs.
3325 65. Users' Exim filters run in subprocesses under the user's uid. It is
3326 possible for a "deliver" command or an alias in a "personal" command to
3327 provoke an address rewrite. If logging of address rewriting is configured,
3328 this fails because the process is not running as root or exim. There may be
3329 a better way of dealing with this, but for the moment (because 4.30 needs
3330 to be released), I have disabled address rewrite logging when running a
3331 filter in a non-root, non-exim process.
3337 1. The buildconfig auxiliary program wasn't quoting the value set for
3338 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3339 not defined. This bug was masked in 4.22 by the effect that was fixed in
3342 2. Some messages that were rejected after a message id was allocated were
3343 shown as "incomplete" by exigrep. It no longer does this for messages that
3344 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3346 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3347 have allow_domain_literals set, the ID did not get logged in the <= line.
3348 Domain literals are now always recognized in Message-ID: header lines.
3350 4. The first argument for a ${extract expansion item is the key name or field
3351 number. Leading and trailing spaces in this item were not being ignored,
3352 causing some misleading effects.
3354 5. When deliver_drop_privilege was set, single queue runner processes started
3355 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3356 same command in the process it spins off) were not dropping privilege.
3358 6. When the daemon running as "exim" started a queue runner, it always
3359 re-executed Exim in the spun-off process. This is a waste of effort when
3360 deliver_drop_privilege is set. The new process now just calls the
3361 queue-runner function directly.
3367 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3370 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3371 ignored. Though the use of -oP was forcing the writing of a pid file, it
3372 was always written to the default place.
3374 3. If the message "no IP address found for host xxxx" is generated during
3375 incoming verification, it is now followed by identification of the incoming
3376 connection (so you can more easily find what provoked it).
3378 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3380 5. Added some features to "harden" Exim a bit more against certain attacks:
3382 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3383 be put in Local/Makefile. This is like the never_users runtime option,
3384 but it cannot be overridden. The default setting is "root".
3386 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3387 prefix string with which any file named in a -C command line option
3390 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3391 is retained for -C and -D only if the caller of Exim is root. Without
3392 it, the exim user may also use -C and -D and retain privilege.
3394 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3395 command line option is disabled.
3397 6. Macro names set by the -D option must start with an upper case letter, just
3398 like macro names defined in the configuration file.
3400 7. Added "dereference=" facility to LDAP.
3402 8. Two instances of the typo "uknown" in the source files are fixed.
3404 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3405 the Configure-Makefile script screwed up while processing it.
3407 10. Incorporated PCRE 4.4.
3409 11. The SMTP synchronization check was not operating right at the start of an
3410 SMTP session. For example, it could not catch a HELO sent before the client
3411 waited for the greeting. There is now a check for outstanding input at the
3412 point when the greeting is written. Because of the duplex, asynchronous
3413 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3414 way, but not yet received, when the check is performed.
3416 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3417 on TCP/IP sockets, because this apparently causes some broken clients to
3420 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3421 unchanged) from the Cygwin maintainer.
3423 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3424 was supported instead of testing for mbx. Effectively a typo.
3426 15. The spa authenticator server code was not checking that the input it
3427 received was valid base64.
3429 16. The debug output line for the "set" modifier in ACLs was not showing the
3430 name of the variable that was being set.
3432 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3434 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3435 lookup on its own. Something else was needed to trigger the lookup. For
3436 example, a match in host_lookup or the need for a host name in a host list.
3437 Now, if $sender_host_name is referenced and the host name has not yet been
3438 looked up, a lookup is performed. If the lookup fails, the variable remains
3439 empty, and $host_lookup_failed is set to "1".
3441 19. Added "eqi" as a case-independent comparison operator.
3443 20. The saslauthd authentication condition could segfault if neither service
3444 nor realm was specified.
3446 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3447 error message that was logged was misleading, and incoming SMTP
3448 connections were dropped. The message is now more accurate, and temporary
3449 errors are given to SMTP connections.
3451 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3452 (except RSET) with a 421 error, until QUIT is received. However, it was
3453 failing to send a response to QUIT.
3455 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3456 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3457 ACL is now run after the helo_try_verify_hosts code.
3459 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3460 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3461 case-independent, so other case variants are also recognized). Apparently
3462 some systems use these upper case variants.
3464 25. If more than two messages were waiting for the same host, and a transport
3465 filter was specified for the transport, Exim sent two messages over the
3466 same TCP/IP connection, and then failed with "socket operation on non-
3467 socket" when it tried to send the third.
3469 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3471 27. The extern definition of crypt16() in expand.c was not being excluded when
3472 the OS had its own crypt16() function.
3474 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3475 as a preferred synonym for return_size_limit, both as an option and as an
3478 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3480 30. Changed the default configuration ACL to relax the local part checking rule
3481 for addresses that are not in any local domains. For these addresses,
3482 slashes and pipe symbols are allowed within local parts, but the sequence
3483 /../ is explicitly forbidden.
3485 31. SPA server authentication was not clearing the challenge buffer before
3488 32. log_message in a "warn" ACL statement was writing to the reject log as
3489 well as to the main log, which contradicts the documentation and doesn't
3490 seem right (because no rejection is happening). So I have stopped it.
3492 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3493 However, I am unable to do any testing of this.
3495 34. Fixed an infelicity in the appendfile transport. When checking directories
3496 for a mailbox, to see if any needed to be created, it was accidentally
3497 using path names with one or more superfluous leading slashes; tracing
3498 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3500 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3501 contents are added to the log line that is written for every discarded
3502 recipient. (Previously a log_message setting was ignored.)
3504 36. The ${quote: operator now quotes the string if it is empty.
3506 37. The install script runs exim in order to find its version number. If for
3507 some reason other than non-existence or emptiness, which it checks, it
3508 could not run './exim', it was installing it with an empty version number,
3509 i.e. as "exim-". This error state is now caught, and the installation is
3512 38. An argument was missing from the function that creates an error message
3513 when Exim fails to connect to the socket for saslauthd authentication.
3514 This could cause Exim to crash, or give a corrupted message.
3516 39. Added isip, isip4, and isip6 to ${if conditions.
3518 40. The ACL variables $acl_xx are now saved with the message, and can be
3519 accessed later in routers, transports, and filters.
3521 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3522 strings in the file are not string-expanded.
3524 42. If a MAIL command specified a SIZE value that was too large to fit into an
3525 int variable, the check against message_size_limit failed. Such values are
3526 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3527 day this will have to be increased, but I don't think I want to be around
3528 when emails are that large.
3535 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3536 iconv() is not standard in FreeBSD.
3538 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3539 IPv6 enabled. The observed symptom was a segmentation fault on return from
3540 the function os_common_find_running_interfaces() in src/os.c.
3542 3. In the check_special_case() function in daemon.c I had used "errno" as an
3543 argument name, which causes warnings on some systems. This was basically a
3544 typo, since it was named "eno" in the comments!
3546 4. The code that waits for the clock to tick (at a resolution of some fraction
3547 of a second) so as to ensure message-id uniqueness was always waiting for
3548 at least one whole tick, when it could have waited for less. [This is
3549 almost certainly not relevant at current processor speeds, where it is
3550 unlikely to ever wait at all. But we try to future-proof.]
3552 5. The function that sleeps for a time interval that includes fractions of a
3553 second contained a race. It did not block SIGALRM between setting the
3554 timer, and suspending (a couple of lines later). If the interval was short
3555 and the sigsuspend() was delayed until after it had expired, the suspension
3556 never ended. On busy systems this could lead to processes getting stuck for
3559 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3560 process, before it forks any delivery processes. The open lookup caching
3561 mechanism meant that the open file or database connection was passed into
3562 the delivery process. The problem was that delivery processes always tidy
3563 up cached lookup data. This could cause a problem for the next delivery
3564 process started by the queue runner, because the external queue runner
3565 process does not know about the closure. So the next delivery process
3566 still has data in the lookup cache. In the case of a file lookup, there was
3567 no problem because closing a file descriptor in a subprocess doesn't affect
3568 the parent. However, if the lookup was caching a connection to a database,
3569 the connection was closed, and the second delivery process was likely to
3570 see errors such as "PGSQL: query failed: server closed the connection
3571 unexpectedly". The problem has been fixed by closing all cached lookups
3572 in a queue runner before running a delivery process.
3574 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3575 seem to have the "const" qualifier which it has on other OS. I've
3578 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3579 *of the same type* (client or server) with the same public name that an
3580 error should be diagnosed.
3582 9. When Exim looked up a host name for an IP address, but failed to find the
3583 original IP address when looking up the host name (a safety check), it
3584 output the message "<ip address> does not match any IP for NULL", which was
3585 confusing, to say the least. The bug was that the host name should have
3586 appeared instead of "NULL".
3588 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3589 user that is built into the binary, and the -C or -D options is used, root
3590 privilege is dropped before the configuration file is read. In addition,
3591 logging is switched to stderr instead of the normal log files. If the
3592 configuration then re-defines the Exim user, the unprivileged environment
3593 is probably not what is expected, so Exim logs a panic warning message (but
3596 However, if deliver_drop_privilege is set, the unprivileged state may well
3597 be exactly what is intended, so the warning has been cut out in that case,
3598 and Exim is allowed to try to write to its normal log files.
3604 1. smtp_return_error_details was not giving details for temporary sender
3605 or receiver verification errors.
3607 2. Diagnose a configuration error if two authenticators have the same public
3610 3. Exim used not to create the message log file for a message until the first
3611 delivery attempt. This could be confusing when incoming messages were held
3612 for policy or load reasons. The message log file is now created at the time
3613 the message is received, and an initial "Received" line is written to it.
3615 4. The automatically generated man page for command line options had a minor
3616 bug that caused no ill effects; however, a more serious problem was that
3617 the procedure for building the man page automatically didn't always
3618 operate. Consequently, release 4.20 contains an out-of-date version. This
3619 shouldn't happen again.
3621 5. When building Exim with embedded Perl support, the script that builds the
3622 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3623 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3625 6. The freeze_tell option was not being used for messages that were frozen on
3626 arrival, either by an ACL or by local_scan().
3628 7. Added the smtp_incomplete_transaction log selector.
3630 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3631 was accepting AUTH without a new EHLO.
3633 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3634 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3637 10. Logging of TCP/IP connections (when configured) now happens in the main
3638 daemon process instead of the child process, so that the TCP/IP connection
3639 count is more accurate (but it can never be perfect).
3641 11. The use of "drop" in a nested ACL was not being handled correctly in the
3642 outer ACL. Now, if condition failure induced by the nested "drop" causes
3643 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3644 or "require"), the connection is dropped.
3646 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3647 that yield "discard" can now be used with an "accept" or a "discard" verb,
3648 but an error is generated for any others (because I can't see a useful way
3649 to define what should happen).
3651 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3652 are now processed in the same way as lines in the Exim configuration file.
3653 In particular, continuation lines are supported.
3655 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3657 15. Added -ti meaning -t -i.
3659 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3660 domains, and warn by logging if others are found.
3662 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3663 returned from GIFCONF when Exim is trying to find the list of interfaces on
3664 a host. The code in os.c has been modified to copy each ifreq to an aligned
3665 structure in all cases.
3667 Also, in some cases, the returned ifreq's were being copied to a 'struct
3668 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3669 means the last couple of bytes of an IPv6 address could be chopped if the
3670 ifreq contained only a normal sockaddr (14 bytes storage).
3672 18. Named domain lists were not supported in the hosts_treat_as_local option.
3673 An entry such as +xxxx was not recognized, and was treated as a literal
3676 19. Ensure that header lines added by a DATA ACL are included in the reject log
3677 if the ACL subsequently rejects the message.
3679 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3680 MD5 (which is deprecated).
3682 21. When testing a filter file using -bf, Exim was writing a message when it
3683 took the sender from a "From " line in the message, but it was not doing so
3684 when it took $return_path from a Return-Path: header line. It now does.
3686 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3687 with a valid header line field name (a series of printing characters
3688 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3690 23. Changed "disc" in the source to "disk" to conform to the documentation and
3691 the book and for uniformity.
3693 24. Ignore Sendmail's -Ooption=value command line item.
3695 25. When execve() failed while trying to run a command in a pipe transport,
3696 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3697 could be confused with a return value of 69 from the command itself. This
3698 has been changed to 127, the value the shell returns if it is asked to run
3699 a non-existent command. The wording for the related log line suggests a
3700 non-existent command as the problem.
3702 26. If received_header_text expands to an empty string, do not add a Received:
3703 header line to the message. (Well, it adds a token one on the spool, but
3704 marks it "old" so that it doesn't get used or transmitted.)
3706 27. Installed eximstats 1.28 (addition of -nt option).
3708 28. There was no check for failure on the call to getsockname() in the daemon
3709 code. This can fail if there is a shortage of resources on the system, with
3710 ENOMEM, for example. A temporary error is now given on failure.
3712 29. Contrary to the C standard, it seems that in some environments, the
3713 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3714 program. Exim now does this explicitly; it affects the formatting of
3715 timestamps using strftime().
3717 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3718 complaints. I've now initialized all the variables, to avoid this.
3720 32. Header lines added by a system filter were not being "seen" during
3721 transport-time rewrites.
3723 33. The info_callback() function passed to OpenSSL is set up with type void
3724 (*)(SSL *, int, int), as described somewhere. However, when calling the
3725 function (actually a macro) that sets it up, the type void(*)() is
3726 expected. I've put in a cast to prevent warnings from picky compilers.
3728 34. If a DNS black list lookup found a CNAME record, but there were no A
3729 records associated with the domain it pointed at, Exim crashed.
3731 35. If a DNS black list lookup returned more than one A record, Exim ignored
3732 all but the first. It now scans all returned addresses if a particular IP
3733 value is being sought. In this situation, the contents of the
3734 $dnslist_value variable are a list of all the addresses, separated by a
3737 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3738 to accept a host name and all its aliases if the forward lookup for any of
3739 them yielded the IP address of the incoming connection. Now it accepts only
3740 those names whose forward lookup yields the correct IP address. Any other
3741 names are discarded. This closes a loophole whereby a rogue DNS
3742 administrator could create reverse DNS records to break through a
3743 wildcarded host restriction in an ACL.
3745 37. If a user filter or a system filter that ran in a subprocess used any of
3746 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3747 the wrong values were passed to the pipe command ($thisaddress had the
3748 value of $0, $0 had the value of $1, etc). This bug was introduced by
3749 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3751 38. Improved the line breaking for long SMTP error messages from ACLs.
3752 Previously, if there was no break point between 40 and 75 characters, Exim
3753 left the rest of the message alone. Two changes have been made: (a) I've
3754 reduced the minimum length to 35 characters; (b) if it can't find a break
3755 point between 35 and 75 characters, it looks ahead and uses the first one
3756 that it finds. This may give the occasional overlong line, but at least the
3757 remaining text gets split now.
3759 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3760 file descriptors might be low, and that setting 1000 would always raise it.
3761 It turns out that in some environments, the limit is already over 1000 and
3762 that lowering it causes trouble. So now Exim takes care not to decrease it.
3764 40. When delivering a message, the value of $return_path is set to $sender_
3765 address at the start of routing (routers may change the value). By an
3766 oversight, this default was not being set up when an address was tested by
3767 -bt or -bv, which affected the outcome if any router or filter referred to
3770 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3771 log only when "message" or "log_message" are set. However, if one of the
3772 conditions was an address verification, or a call to a nested ACL, the
3773 messages generated by the underlying test were being passed through. This
3774 no longer happens. The underlying message is available in $acl_verify_
3775 message for both "message" and "log_message" expansions, so it can be
3776 passed through if needed.
3778 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3779 new expansion $bh_ to give the encoded byte string without charset
3780 translation. Translation happens only if iconv() is available; HAVE_ICONV
3781 indicates this at build time. HEADERS_CHARSET gives the charset to
3782 translate to; headers_charset can change it in the configuration, and
3783 "headers charset" can change it in an individual filter file.
3785 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3786 that creates RFC 2047 encoded "words" labels them as that charset instead
3787 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3788 expansion operator; (ii) when Exim creates a From: line for a local
3789 message; (iii) when a header line is rewritten to include a "phrase" part.
3791 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3792 buggy, causing it to skip the first lines of messages whose message ID
3793 ended in 'D'. This would not have bitten before Exim release 4.14, because
3794 message IDs were unlikely to end in 'D' before then. The effect was to have
3795 incorrect size information for certain domains.
3797 45. #include "config.h" was missing at the start of the crypt16.c module. This
3798 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3801 46. If there was a timeout during a "random" callout check, Exim treated it as
3802 a failure of the random address, and carried on sending RSET and the real
3803 address. If the delay was just some slowness somewhere, the response to the
3804 original RCPT would be taken as a response to RSET and so on, causing
3805 mayhem of various kinds.
3807 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3808 when I implemented it. It didn't allow for the fact that some option values
3809 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3810 the right test for positive values.
3812 48. Domain names in DNS records are case-independent. Exim always looks them up
3813 in lower case. Some resolvers return domain names in exactly the case they
3814 appear in the zone file, that is, they may contain uppercase letters. Not
3815 all resolvers do this - some return always lower case. Exim was treating a
3816 change of case by a resolver as a change of domain, similar to a widening
3817 of a domain abbreviation. This triggered its re-routing code and so it was
3818 trying to route what was effectively the same domain again. This normally
3819 caused routing to fail (because the router wouldn't handle the domain
3820 twice). Now Exim checks for this case specially, and just changes the
3821 casing of the domain that it ultimately uses when it transmits the message
3824 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3827 50. If a filter generated a file delivery with a non-absolute name (possible if
3828 no home directory exists for the router), the forbid_file option was not
3831 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3832 the existing equality matching.
3834 52. Exim was using ints instead of ino_t variables in some places where it was
3835 dealing with inode numbers.
3837 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3838 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3839 called TMPDIR, and if it finds it is different, it changes its value.
3841 54. The smtp_printf() function is now made available to local_scan() so
3842 additional output lines can be written before returning. There is also an
3843 smtp_fflush() function to enable the detection of a dropped connection.
3844 The variables smtp_input and smtp_batched_input are exported to
3847 55. Changed the default runtime configuration: the message "Unknown user"
3848 has been removed from the ACL, and instead placed on the localuser router,
3849 using the cannot_route_message feature. This means that any verification
3850 failures that generate their own messages won't get overridden. Similarly,
3851 the "Unrouteable address" message that was in the ACL for unverifiable
3852 relay addresses has also been removed.
3854 56. Added hosts_avoid_esmtp to the smtp transport.
3856 57. The exicyclog script was not checking for the esoteric option
3857 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3858 will work only if exicyclog is run under the appropriate euid.
3860 58. Following a discussion on the list, the rules by which Exim recognises line
3861 endings on incoming messages have been changed. The -dropcr and drop_cr
3862 options are now no-ops, retained only for backwards compatibility. The
3863 following line terminators are recognized: LF CRLF CR. However, special
3864 processing applies to CR:
3866 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3867 nor a local message in the state where . is a terminator.
3869 (ii) If a bare CR is encountered in a header line, an extra space is added
3870 after the line terminator so as not to end the header. The reasoning
3871 behind this is that bare CRs in header lines are most likely either
3872 to be mistakes, or people trying to play silly games.
3874 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3875 was being incorrectly given as 18 bytes larger than it should have been.
3876 This is a VOB (very old bug).
3878 60. This may never have affected anything current, but just in case it has:
3879 When the local host is found other than at the start of a list of hosts,
3880 the local host, those with the same MX, and any that follow, are discarded.
3881 When the list in question was part of a longer list of hosts, the following
3882 hosts (not currently being processed) were also being discarded. This no
3883 longer happens. I'm not sure if this situation could ever has previously
3886 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3889 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3890 header. This is something that is recommended in a new Internet Draft, and
3891 is something that is documented as being done by Sendmail. There are two
3892 possible values. For messages generated by the autoreply transport, Exim
3895 Auto-Submitted: auto-replied
3897 whereas for all other generated messages (e.g. bounces) it adds
3899 Auto-Submitted: auto-generated
3901 63. The "personal" condition in filters now includes a test for the
3902 Auto-Submitted: header. If it contains the string "auto-" the message it
3903 not considered personal.
3905 64. Added rcpt_include_affixes as a generic transport option.
3907 65. Added queue_only_override (default true).
3909 66. Added the syslog_duplication option.
3911 67. If what should have been the first header line of a message consisted of
3912 a space followed by a colon, Exim was mis-interpreting it as a header line.
3913 It isn't of course - it is syntactically invalid and should therefore be
3914 treated as the start of the message body. The misbehaviour could have
3915 caused a number of strange effects, including loss of data in subsequent
3916 header lines, and spool format errors.
3918 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3919 client host had authenticated. This control can now be exercised by an ACL
3920 for more flexibility.
3922 69. By default, callouts do not happen when testing with -bh. There is now a
3923 variant, -bhc, which does actually run the callout code, including
3924 consulting and updating the callout cache.
3926 70. Added support for saslauthd authentication, courtesy of Alexander
3929 71. If statvfs() failed on the spool or log directories while checking their
3930 size for availability, Exim confusingly gave the error "space shortage".
3931 Furthermore, in debugging mode it crashed with a floating point exception.
3932 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3933 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3934 serious problem, Exim now writes to the main and panic logs when this
3935 happens, with details of the failure. It then refuses to accept the
3936 incoming message, giving the message "spool directory problem" or "log
3937 directory problem" with a 421 code for SMTP messages.
3939 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3940 0, 1, and 2 exist, because some OS complain for execs without them (see
3941 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3942 descriptors. However, the code omitted to check that the open succeeded,
3943 causing mysterious errors if for some reason the permissions on /dev/null
3944 got screwed. Now Exim writes a message to the main and panic logs, and
3945 bombs out if it can't open /dev/null.
3947 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3948 interact so that it is all more flexible. It is supposed to remain
3949 backwards compatible. Also added extra_local_interfaces.
3951 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3952 to bomb out with an assertion failure - to the client this appears as a
3953 connection drop. This problem occurs in the part of the code that was taken
3954 from the Samba project. Fortunately, the assertion is in a very simple
3955 function, so I have fixed this by reproducing the function inline in the
3956 one place where it is called, and arranging for authentication to fail
3957 instead of killing the process with assert().
3959 75. The SPA client code was not working when the server requested OEM rather
3960 than Unicode encoding.
3962 76. Added code to make require_files with a specific uid setting more usable in
3963 the case where statting the file as root fails - usually a non-root-mounted
3964 NFS file system. When this happens and the failure is EACCES, Exim now
3965 forks a subprocess and does the per-uid checking as the relevant uid.
3967 77. Added process_log_path.
3969 78. If log_file_path was not explicitly set, a setting of check_log_space or
3970 check_log_inodes was ignored.
3972 79. If a space check for the spool or log partitions fails, the incident is now
3973 logged. Of course, in the latter case the data may get lost...
3975 80. Added the %p formatting code to string_format() so that it can be used to
3976 print addresses in debug_print(). Adjusted all the address printing in the
3977 debugging in store.c to use %p rather than %d.
3979 81. There was a concern that a line of code in smtp_in.c could overflow a
3980 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3981 initially expressed, the concern was not well-founded, because trailing
3982 spaces are removed early. However, if the trailing spaces were followed by
3983 a NULL, they did not get removed, so the overflow was possible. Two fixes
3986 (a) I re-wrote the offending code in a cleaner fashion.
3987 (b) If an incoming SMTP command contains a NULL character, it is rejected
3990 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3991 runs initgroups(), so that if the Exim user is in any additional groups,
3992 they will be used during message reception.
3998 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob