1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.350 2006/05/05 14:31:24 steve Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
10 other effects) broke the use of negated acl sub-conditions.
12 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
15 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
16 "Deny" causes Exim to reject the incoming connection with a 554 error.
17 Unfortunately, if there is a major crisis, such as a disk failure,
18 tcp-wrappers gives "deny", whereas what one would like would be some
19 kind of temporary error. A kludge has been added to help with this.
20 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
21 554 error is used if errno is still zero or contains ENOENT (which occurs
22 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
25 PH/03 Add -lutil to the default FreeBSD LIBS setting.
27 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
28 errors. Otherwise a message that provokes a temporary error (when other
29 messages do not) can cause a whole host to time out.
31 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
32 addresses were routed directly to files or pipes from a redirect router.
33 File deliveries just didn't batch; pipe deliveries might have suffered
36 PH/06 A failure to get a lock for a hints database would erroneously always say
37 "Failed to get write lock", even when it was really a read lock.
39 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
40 of 0600. This has been changed to use the value of the lockfile_mode
41 option (which defaults to 0600).
43 PH/08 Applied small patch from the Sieve maintainer.
45 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
46 folder from quota calculations, a direct delivery into this folder messed
47 up the contents of the maildirsize file. This was because the regex was
48 used only to exclude .Trash (or whatever) when the size of the mailbox
49 was calculated. There was no check that a delivery was happening into an
50 excluded directory. This bug has been fixed by ignoring all quota
51 processing for deliveries into excluded directories.
53 PH/10 Added the maildirfolder_create_regex option to appendfile.
55 SC/01 Use a glob alias rather than an array ref in eximstats generated
56 parser. This improves both readability and performance.
58 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
59 Don't display local sender or destination tables in eximstats unless
60 there is data to show.
61 Added average volumes into the eximstats top table text output.
63 SC/03 Collect data on the number of addresses (recipients) as well
64 as the number of messages in eximstats.
70 PH/01 The code for finding all the local interface addresses on a FreeBSD
71 system running IPv6 was broken. This may well have applied to all BSD
72 systems, as well as to others that have similar system calls. The broken
73 code found IPv4 interfaces correctly, but gave incorrect values for the
74 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
75 that it would not match correctly against @[] and not recognize the IPv6
78 PH/02 The ipliteral router was not recognizing addresses of the form user@
79 [ipv6:....] because it didn't know about the "ipv6:" prefix.
81 PH/03 Added disable_ipv6.
83 PH/04 Changed $reply_address to use the raw form of the headers instead of the
84 decoded form, because it is most often used to construct To: headers
85 lines in autoreplies, and the decoded form may well be syntactically
86 invalid. However, $reply_address has leading white space removed, and all
87 newlines turned into spaces so that the autoreply transport does not
90 PH/05 If group was specified without a user on a router, and no group or user
91 was specified on a transport, the group from the router was ignored.
93 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
94 for visible compile-time settings that can be used to change these
95 numbers, for those that want even more. Backwards compatibility with old
96 spool files has been maintained. However, going back to a previous Exim
97 release will lost any variables that are in spool files.
99 PH/07 Two small changes when running in the test harness: increase delay when
100 passing a TCP/IP connection to a new process, in case the original
101 process has to generate a bounce, and remove special handling of
102 127.0.0.2 (sic), which is no longer necessary.
104 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
105 be the same on different OS.
107 PH/09 Moved a debug statement in filter processing to avoid a race problem when
110 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
111 whether --show-vars was specified or not
113 JJ/02 exipick: Added support for new ACL variable spool format introduced
116 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
117 syntactically invalid From: or Reply-to: line, and a filter used this to
118 generate an autoreply, and therefore failed to obtain an address for the
119 autoreply, Exim could try to deliver to a non-existent relative file
120 name, causing unrelated and misleading errors. What now happens is that
121 it logs this as a hard delivery error, but does not attempt to create a
124 PH/11 The exinext utility has a -C option for testing purposes, but although
125 the given file was scanned by exinext itself; it wasn't being passed on
126 when Exim was called.
128 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
129 an end-of-file indication when reading a command response.
131 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
132 compiled. In many other places in Exim, IPv6 addresses are always
133 recognized, so I have changed this. It also means that IPv4 domain
134 literals of the form [IPV4:n.n.n.n] are now always recognized.
136 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
137 used if the router is not running as root, for example, when verifying at
138 ACL time, or when using -bh. The debugging output from this situation was
139 non-existent - all you got was a failure to exec. I have made two
142 (a) Failures to set uid/gid, the current directory, or a process leader
143 in a subprocess such as that created by queryprogram now generate
144 suitable debugging ouput when -d is set.
146 (b) The queryprogram router detects when it is not running as root,
147 outputs suitable debugging information if -d is set, and then runs
148 the subprocess without attempting to change uid/gid.
150 PH/15 Minor change to Makefile for building test_host (undocumented testing
153 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
154 additional section of a DNS packet that returns MX or SRV records.
155 Instead, it always explicitly searches for A/AAAA records. This avoids
156 major problems that occur when a DNS server includes only records of one
157 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
158 fixed another bug: if SRV records were looked up and the corresponding
159 address records were *not* found in the additional section, the port
160 values from the SRV records were lost.
162 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
163 using the correct key (the original address) when searching the retry
164 rules in order to find which one to use for generating the retry hint.
166 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
167 adding the default one. Similarly, if it contains a Reply-To: header, the
168 errors_reply_to option, if set, is not used.
170 PH/19 When calculating a retry time, Exim used to measure the "time since
171 failure" by looking at the "first failed" field in the retry record. Now
172 it does not use this if it is later than than the arrival time of the
173 message. Instead it uses the arrival time. This makes for better
174 behaviour in cases where some deliveries succeed, thus re-setting the
175 "first failed" field. An example is a quota failure for a huge message
176 when small messages continue to be delivered. Without this change, the
177 "time since failure" will always be short, possible causing more frequent
178 delivery attempts for the huge message than are intended.
179 [Note: This change was subsequently modified - see PH/04 for 4.62.]
181 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
182 $1, $2, $3) because the numerical variables can be reset during some
183 expansion items (e.g. "match"), thereby losing the authentication data.
185 PH/21 Make -bV show the size of off_t variables so that the test suite can
186 decide whether to run tests for quotas > 2G.
188 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
189 mailbox_size, and mailbox_filecount in the appendfile transport. If a
190 filecount value is greater than 2G or if a quota value is greater than 2G
191 on a system where the size of off_t is not greater than 4, a panic error
194 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
195 never match. The debug and -bh output now contains an explicit error
196 message indicating a malformed IPv4 address or mask.
198 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
199 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
202 PH/25 Do not write to syslog when running in the test harness. The only
203 occasion when this arises is a failure to open the main or panic logs
204 (for which there is an explicit test).
206 PH/26 Added the /no_tell option to "control=freeze".
208 PH/27 If a host name lookup failed very early in a connection, for example, if
209 the IP address matched host_lookup and the reverse lookup yielded a name
210 that did not have a forward lookup, an error message of the form "no IP
211 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
212 could be logged. Now it outputs the IP address instead of "NULL".
214 PH/28 An enabling patch from MH: add new function child_open_exim2() which
215 allows the sender and the authenticated sender to be set when
216 submitting a message from within Exim. Since child_open_exim() is
217 documented for local_scan(), the new function should be too.
219 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
220 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
221 results in an empty string is now treated as unset.
223 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
225 PH/31 Added sender_verify_fail logging option.
227 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
228 needed by RFC 822 but not by RFC 2822 was commented out. I have now
229 tidied the source and removed it altogether.
231 PH/33 When a queue run was abandoned because the load average was too high, a
232 log line was always written; now it is written only if the queue_run log
233 selector is set. In addition, the log line for abandonment now contains
234 information about the queue run such as the pid. This is always present
235 in "start" and "stop" lines but was omitted from the "abandon" line.
237 PH/34 Omit spaces between a header name and the colon in the error message that
238 is given when verify = headers_syntax fails (if there are lots of them,
239 the message gets confusing).
241 PH/35 Change the default for dns_check_names_pattern to allow slashes within
242 names, as there are now some PTR records that contain slashes. This check
243 is only to protect against broken name servers that fall over on strange
244 characters, so the fact that it applies to all lookups doesn't matter.
246 PH/36 Now that the new test suite is complete, we can remove some of the
247 special code in Exim that was needed for the old test suite. For example,
248 sorting DNS records because real resolvers return them in an arbitrary
249 order. The new test suite's fake resolver always returns records in the
252 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
253 bounces) except when queue_only is set, to avoid logging races between
254 the different processes.
256 PH/38 Panic-die if .include specifies a non-absolute path.
258 PH/39 A tweak to the "H" retry rule from its user.
260 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
261 a label. They prevented compilation on older perls.
263 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
264 a warning to be raised on newish perls.
266 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
267 on queue. Changes to match documented behaviour of showing count of
268 messages matching specified criteria.
270 PH/40 Changed the default ident timeout from 30s to 5s.
272 PH/41 Added support for the use of login_cap features, on those BSD systems
273 that have them, for controlling the resources used by pipe deliveries.
275 PH/42 The content-scanning code uses fopen() to create files in which to put
276 message data. Previously it was not paying any attention to the mode of
277 the files. Exim runs with umask(0) because the rest of the code creates
278 files with open(), and sets the required mode explicitly. Thus, these
279 files were ending up world-writeable. This was not a big issue, because,
280 being within the spool directory, they were not world-accessible. I have
281 created a function called modefopen, which takes an additional mode
282 argument. It sets umask(777), creates the file, chmods it to the required
283 mode, then resets the umask. All the relevant calls to fopen() in the
284 content scanning code have been changed to use this function.
286 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
287 to 24 hours. This avoids potential overflow problems when processing G
288 and H retry rules. I suspect nobody ever tinkers with this value.
290 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
292 PH/45 When the plaintext authenticator is running as a client, the server's
293 challenges are checked to ensure they are valid base64 strings. By
294 default, the authentication attempt is cancelled if an invalid string is
295 received. Setting client_ignore_invalid_base64 true ignores these errors.
296 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
297 they are received. Thus, the responses can be made to depend on the
298 challenges. If an invalid string is ignored, an empty string is placed in
301 PH/46 Messages that are created by the autoreply transport now contains a
302 References: header, in accordance with RFCs 2822 and 3834.
304 PH/47 Added authenticated_sender_force to the smtp transport.
306 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
308 PH/49 Installed latest patch from the Sieve maintainer.
310 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
311 was also set, the appendfile transport was unnecessarily scanning a
312 directory of message files (e.g. for maildir delivery) to find the count
313 of files (along with the size), even though it did not need this
314 information. It now does the scan only if it needs to find either the
315 size of the count of files.
317 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
319 PH/52 Two bugs concerned with error handling when the smtp transport is
322 (i) Exim was not creating retry information for temporary errors given
323 for individual recipients after the DATA command when the smtp transport
324 was used in LMTP mode. This meant that they could be retried too
325 frequently, and not timed out correctly.
327 (ii) Exim was setting the flag that allows error details to be returned
328 for LMTP errors on RCPT commands, but not for LMTP errors for individual
329 recipients that were returned after the DATA command.
331 PH/53 This is related to PH/52, but is more general: for any failing address,
332 when detailed error information was permitted to be returned to the
333 sender, but the error was temporary, then after the final timeout, only
334 "retry timeout exceeded" was returned. Now it returns the full error as
335 well as "retry timeout exceeded".
337 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
338 do this, and (what is worse) MTAs that accept it.
340 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
341 will now be deprecated.
343 PH/56 New os.c-cygwin from the Cygwin maintainer.
345 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
346 formats (previously only available in exim formats via -bpr, -bpru,
347 and -bpra. Now also available in native and exiqgrep formats)
349 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
350 with very large, slow to parse queues
352 JJ/08 exipick: added ! as generic prefix to negate any criteria format
354 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
356 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
357 responses to authentication challenges, though it was showing the
358 challenges; (ii) I've removed the CR characters from the debug output for
361 PH/58 Allow for the insertion of a newline as well as a space when a string
362 is turned into more than one encoded-word during RFC 2047 encoding. The
363 Sieve code now uses this.
365 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
366 data_4xx, lost_connection, tls_required.
368 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
369 message was being sent as an SMTP response.
371 PH/61 Add -l and -k options to exicyclog.
373 PH/62 When verifying, if an address was redirected to one new address, so that
374 verification continued, and the new address failed or deferred after
375 having set something in $address_data, the value of $address_data was not
376 passed back to the ACL. This was different to the case when no
377 redirection occurred. The value is now passed back in both cases.
379 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
380 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
381 use login_cap.h, so on its own it isn't the distinguishing feature. The
382 new name refers directly to the setclassresources() function.
384 PH/65 Added configuration files for NetBSD3.
386 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
388 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
389 is preferred over IPv4.
391 PH/68 The bounce_return_message and bounce_return_body options were not being
392 honoured for bounces generated during the reception of non-SMTP messages.
393 In particular, this applied to messages rejected by the ACL. This bug has
394 been fixed. However, if bounce_return_message is true and bounce_return_
395 body is false, the headers that are returned for a non-SMTP message
396 include only those that have been read before the error was detected.
397 (In the case of an ACL rejection, they have all been read.)
399 PH/69 The HTML version of the specification is now built in a directory called
400 spec_html instead of spec.html, because the latter looks like a path with
401 a MIME-type, and this confuses some software.
403 PH/70 Catch two compiler warnings in sieve.c.
405 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
406 function verify_get_ident() calls ip_connect() to connect a socket, but
407 if the "connect()" function timed out, ip_connect() used to close the
408 socket. However, verify_get_ident() also closes the socket later, and in
409 between Exim writes to the log, which may get opened at this point. When
410 the socket was closed in ip_connect(), the log could get the same file
411 descriptor number as the socket. This naturally causes chaos. The fix is
412 not to close the socket in ip_connect(); the socket should be closed by
413 the function that creates it. There was only one place in the code where
414 this was missing, in the iplookup router, which I don't think anybody now
415 uses, but I've fixed it anyway.
417 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
418 well as to direct DNS lookups. Otherwise the handling of names in host
419 lists is inconsistent and therefore confusing.
425 PH/01 Two changes to the default runtime configuration:
427 (1) Move the checks for relay_from_hosts and authenticated clients from
428 after to before the (commented out) DNS black list checks.
430 (2) Add control=submission to the relay_from_hosts and authenticated
431 clients checks, on the grounds that messages accepted by these
432 statements are most likely to be submissions.
434 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
436 (1) Generate an error if the third argument for the ${prvs expansion is
439 (2) Treat a missing third argument of ${prvscheck as if it were an empty
442 (3) Reset the variables that are obtained from the first argument of
443 ${prvscheck and used in the second argument before leaving the code,
444 because their memory is reclaimed, so using them afterwards may do
447 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
448 one (it's much easier than Tom thought :-).
450 (5) Because of (4), we can now allow for the use of $prvscheck_result
451 inside the third argument.
453 PH/03 For some reason, the default setting of PATH when running a command from
454 a pipe transport was just "/usr/bin". I have changed it to
457 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
458 anything to be listed in the output from -bV.
460 PH/05 When a filter generated an autoreply, the entire To: header line was
461 quoted in the delivery log line, like this:
463 => >A.N.Other <ano@some.domain> <original@ddress> ...
465 This has been changed so that it extracts the operative address. There
466 may be more than one such address. If so, they are comma-separated, like
469 => >ano@some.domain,ona@other.domain <original@ddress> ...
471 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
472 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
473 not being looked up in the rDNS to get a host name, Exim was showing the
474 IP address twice in Received: lines, even though the IP addresses were
475 identical. For example:
477 Received: from [1.2.3.4] (helo=[1.2.3.4])
479 However, if the real host name was known, it was omitting the HELO data
480 if it matched the actual IP address. This has been tidied up so that it
481 doesn't show the same IP address twice.
483 PH/07 When both +timestamp and +memory debugging was on, the value given by
484 $tod_xxx expansions could be wrong, because the tod_stamp() function was
485 called by the debug printing, thereby overwriting the timestamp buffer.
486 Debugging no longer uses the tod_stamp() function when +timestamp is set.
488 PH/08 When the original message was included in an autoreply transport, it
489 always said "this is a copy of the message, including all the headers",
490 even if body_only or headers_only was set. It now gives an appropriate
493 PH/09 Applied a patch from the Sieve maintainer which:
495 o fixes some comments
496 o adds the (disabled) notify extension core
497 o adds some debug output for the result of if/elsif tests
498 o points to the current vacation draft in the documentation
499 and documents the missing references header update
503 o fixes a bug in processing the envelope test (when testing
504 multiple envelope elements, the last element determinted the
507 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
508 Electronic Mail") by including:
510 Auto-submitted: auto-generated
512 in the messages that it generates (bounce messages and others, such as
513 warnings). In the case of bounce messages for non-SMTP mesages, there was
514 also a typo: it was using "Auto_submitted" (underscore instead of
515 hyphen). Since every message generated by Exim is necessarily in response
516 to another message, thes have all been changed to:
518 Auto-Submitted: auto-replied
520 in accordance with these statements in the RFC:
522 The auto-replied keyword:
524 - SHOULD be used on messages sent in direct response to another
525 message by an automatic process,
527 - MUST NOT be used on manually-generated messages,
529 - MAY be used on Delivery Status Notifications (DSNs) and Message
530 Disposition Notifications (MDNs),
532 - MUST NOT be used on messages generated by automatic or periodic
533 processes, except for messages which are automatic responses to
536 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
537 to the default Received: header definition.
539 PH/12 Added log selector acl_warn_skipped (default on).
541 PH/13 After a successful wildlsearch lookup, discard the values of numeric
542 variables because (a) they are in the wrong storage pool and (b) even if
543 they were copied, it wouldn't work properly because of the caching.
545 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
546 checking when decoding. Apparently there are clients that generate
547 overlong encoded strings. Why am I not surprised?
549 PH/15 If the first argument of "${if match_address" was not empty, but did not
550 contain an "@" character, Exim crashed. Now it writes a panic log message
551 and treats the condition as false.
553 PH/16 In autoreply, treat an empty string for "once" the same as unset.
555 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
556 extension "envelope-auth". The code is finished and in agreement with
557 other implementations, but there is no documentation so far and in fact,
558 nobody wrote the draft yet. This extension is currently #undef'ed, thus
559 not changing the active code.
561 Print executed "if" and "elsif" statements when debugging is used. This
562 helps a great deal to understand what a filter does.
564 Document more things not specified clearly in RFC3028. I had all this
565 sorted out, when out of a sudden new issues came to my mind. Oops."
567 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
570 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
571 canonical form (as documented). However, after a host name lookup from
572 the IP address, check_host() was doing a simple string comparison with
573 addresses acquired from the DNS when checking that the found name did
574 have the original IP as one of its addresses. Since any found IPv6
575 addresses are likely to be in abbreviated form, the comparison could
576 fail. Luckily, there already exists a function for doing the comparison
577 by converting both addresses to binary, so now that is used instead of
580 PH/20 There was another similar case to PH/19, when a complete host name was
581 given in a host list; looking up its IP address could give an abbreviated
582 form, whereas the current host's name might or might not be abbreviated.
583 The same fix has been applied.
589 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
590 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
593 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
594 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
596 PH/03 Typo: missing ".o" in src/pcre/Makefile.
598 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
599 header line, restrict the check to what is listed in RFCs 2369 and 2929.
600 Also, for "Auto-Submitted", treat anything other than "no" as
601 non-personal, in accordance with RFC 3834. (Previously it treated
602 anything starting "auto-" as non-personal.)
604 TF/01 The control=submission/name=... option had a problem with syntax
605 errors if the name included a slash character. The /name= option
606 now slurps the rest of the string, so it can include any characters
607 but it must come last in the list of options (after /sender_retain
610 PH/05 Some modifications to the interface to the fake nameserver for the new
618 TK/01 Added the "success_on_redirect" address verification option. See
619 NewStuff for rationale and an example.
621 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
623 PH/02 Patch to exigrep to allow it to work on syslog lines.
625 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
626 fread() to skip over the body file's header line, because in Cygwin the
627 header line is locked and is inaccessible.
629 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
630 co-exist for some time) to make it clear that it is the Exim ID that is
631 referenced, not the Message-ID: header line.
633 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
634 string_format() function, because snprintf() does not exist on all
637 PH/06 The use of forbid_filter_existstest now also locks out the use of the
638 ${stat: expansion item.
640 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
641 protocol synchronization error", to keep the pedants happy.
643 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
644 well as for IRIX systems, when gcc is being used. See the host.c source
647 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
649 PH/10 Named domain lists were not working if used in a queue_smtp_domains
652 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
653 transport and to the smtp transport in LMTP mode.
655 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
657 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
658 run a filter in a subprocess. This could lead to confusion in subsequent
659 lookups in the parent process. There should also be a search_tidyup() at
660 the end of the subprocess.
662 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
663 only if the host matched helo_try_verify_hosts, which caused the
664 verification to occur when the EHLO/HELO command was issued. The ACL just
665 tested the remembered result. Now, if a previous verification attempt has
666 not happened, "verify = helo" does it there and then.
668 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
670 TK/03 Fix log output including CR from clamd.
672 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
673 exist provoked a memory error which could cause a segfault.
675 PH/15 Installed PCRE 6.2
677 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
679 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
680 of the problem. Specifically, suggested +O2 rather than +O1 for the
683 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
685 PH/20 If a delivery was routed to a non-standard port by means of an SRV
686 record, the port was not correctly logged when the outgoing_port log
687 selector was set (it logged the transort's default port).
689 PH/21 Added support for host-specific ports to manualroute, queryprogram,
690 fallback_hosts, and "hosts" in the smtp transport.
692 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
693 host errors such as "Connection refused".
695 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
696 authentication with radiusclient 0.4.9:
698 - Error returned from rc_read_config was caught wrongly
699 - Username/password not passed on to radius server due to wrong length.
701 The presumption is that some radiusclient API changes for 4.51/PH/17
702 were not taken care of correctly. The code is still untested by me (my
703 Linux distribution still has 0.3.2 of radiusclient), but it was
704 contributed by a Radius user.
706 PH/24 When doing a callout, the value of $domain wasn't set correctly when
707 expanding the "port" option of the smtp transport.
709 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
710 while reading a MIME header. Thanks to Tom Hughes for a patch.
712 PH/24 Include config.h inside local_scan.h so that configuration settings are
715 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
716 that in an ACL for RCPT (for example), you can examine exactly what was
719 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
720 commands, but it was not correctly comparing the address with the actual
721 client host address. Thus, it would show the EHLO address in Received:
722 header lines when this was not necessary.
724 PH/27 Added the % operator to ${eval:}.
726 PH/28 Exim tries to create and chdir to its spool directory when it starts;
727 it should be ignoring failures (because with -C, for example, it has lost
728 privilege). It wasn't ignoring creation failures other than "already
731 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
734 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
735 because an input line was too long, either on its own, or by virtue of
736 too many continuations, the temporary file was not being removed, and the
737 return code was incorrect.
739 PH/31 Missing "BOOL" in function definition in filtertest.c.
741 PH/32 Applied Sieve patches from the maintainer.
743 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
745 PH/33 Added "verify = not_blind".
747 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
748 Local/Makefile (with some defaults set). These are used in built scripts
749 such as exicyclog, but they have never been used in the exim_install
750 script (though there are many overriding facilities there). I have
751 arranged that the exim_install script now takes note of these two
754 PH/35 Installed configuration files for Dragonfly.
756 PH/36 When a locally submitted message by a trusted user did not contain a
757 From: header, and the sender address was obtained from -f or from an SMTP
758 MAIL command, and the trusted user did not use -F to supply a sender
759 name, $originator_name was incorrectly used when constructing a From:
760 header. Furthermore, $originator_name was used for submission mode
761 messages from external hosts without From: headers in a similar way,
762 which is clearly wrong.
764 PH/37 Added control=suppress_local_fixups.
766 PH/38 When log_selector = +received_sender was set, and the addition of the
767 sender made the log line's construction buffer exactly full, or one byte
768 less than full, an overflow happened when the terminating "\n" was
771 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
772 when the result of a list match is failure because a DNS lookup failed.
774 PH/40 RM_COMMAND is now used in the building process.
776 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
777 the "build-* directories that it finds.
779 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
780 address in a domain literal was a prefix of an interface address.
782 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
783 when verifying a sender address, unless rewrite_headers is false.
785 PH/44 Wrote a long comment about why errors_to addresses are verified as
786 recipients, not senders.
788 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
789 the ratelimit ACL was added.
791 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
793 PH/46 Added extra information about PostgreSQL errors to the error string.
795 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
796 avoiding the need to install special zones in a real server. This is
797 backwards compatible; if it can't find the fake resolver, it drops back.
798 Thus, both old and new test suites can be run.
800 TF/02 Added util/ratelimit.pl
802 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
803 clock is set back in time.
805 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
806 Candler <B.Candler@pobox.com>.
808 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
809 OK for addresses that are the result of redirections.
811 PH/48 A number of further additions for the benefit of the new test suite,
812 including a fake gethostbyname() that interfaces to the fake DNS resolver
815 TF/06 The fix for widen_domains has also been applied to qualify_single and
816 search_parents which are the other dnslookup options that can cause
819 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
822 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
824 TF/07 Exim produced the error message "an SRV record indicated no SMTP
825 service" if it encountered an MX record with an empty target hostname.
826 The message is now "an MX or SRV record indicated no SMTP service".
828 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
829 if the DNS of the sending site is misconfigured. This is quite a
830 common situation. This change restores the behaviour of treating a
831 helo verification defer as a failure.
833 PH/51 If self=fail was set on a router, the bounce message did not include the
834 actual error message.
840 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
842 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
843 command itself ended in error, the underlying message about the transport
844 filter timeout was being overwritten with the pipe command error. Now the
845 underlying error message should be appended to the second error message.
847 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
849 PH/02 Reduce the amount of output that "make" produces by default. Full output
850 can still be requested.
852 PH/03 The warning log line about a condition test deferring for a "warn" verb
853 was being output only once per connection, rather than after each
854 occurrence (because it was using the same function as for successful
855 "warn" verbs). This seems wrong, so I have changed it.
857 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
858 it should not have, which might have caused a crash in the right
859 circumstances, but probably never did.
861 PH/04 Installed a modified version of Tony Finch's patch to make submission
862 mode fix the return path as well as the Sender: header line, and to
863 add a /name= option so that you can make the user's friendly name appear
866 TF/03 Added the control = fakedefer ACL modifier.
868 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
869 Mark Lowes for thorough testing.
871 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
873 TK/03 Merged latest SRS patch from Miles Wilton.
875 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
876 with the definition in sysexits.h (which is #included earlier).
877 Fortunately, Exim does not actually use EX_OK. The code used to try to
878 preserve the sysexits.h value, by assumimg that macro definitions were
879 scanned for macro replacements. I have been disabused of this notion,
880 so now the code just undefines EX_OK before #including unistd.h.
882 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
883 in the smtp transport. When a block could not be written in a single
884 write() function, the timeout was being re-applied to each part-write.
885 This seems wrong - if the receiver was accepting one byte at a time it
886 would take for ever. The timeout is now adjusted when this happens. It
887 doesn't have to be particularly precise.
889 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
890 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
892 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
893 without a domain if the check to <postmaster@domain> fails.
895 SC/01 Eximstats: added -xls and the ability to specify output files
896 (patch written by Frank Heydlauf).
898 SC/02 Eximstats: use FileHandles for outputing results.
900 SC/03 Eximstats: allow any combination of xls, txt, and html output.
902 SC/04 Eximstats: fixed display of large numbers with -nvr option
904 SC/05 Eximstats: fixed merging of reports with empty tables.
906 SC/06 Eximstats: added the -include_original_destination flag
908 SC/07 Eximstats: removed tabs and trailing whitespace.
910 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
912 TK/06 MBOX spool code: Add real "From " MBOX separator line
913 so the .eml file is really in mbox format (even though
914 most programs do not really care). Patch from Alex Miller.
916 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
917 The latter is generated from $received_to and is only set if the
918 message has one envelope recipient. SA can use these headers,
919 obviously out-of-the-box. Patch from Alex Miller.
921 PH/08 The ${def test on a variable was returning false if the variable's
922 value was "0", contrary to what the specification has always said!
923 The result should be true unless the variable is empty.
925 PH/09 The syntax error of a character other than { following "${if
926 def:variable_name" (after optional whitespace) was not being diagnosed.
927 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
928 accidental colon was present, for example, could give incorrect results.
930 PH/10 Tidied the code in a number of places where the st_size field of a stat()
931 result is used (not including appendfile, where other changes are about
934 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
935 This involved changing a lot of size variables from int to off_t. It
936 should work with maildirs and everything.
938 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
939 spamd dying while we are connected to it.
941 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
944 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
945 function for BDB 4.3.
947 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
949 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
950 the sources for PCRE have been re-arranged and more files are now
953 PH/15 The code I had for printing potentially long long variables in PH/11
954 above was not the best (it lost precision). The length of off_t variables
955 is now inspected at build time, and an appropriate printing format (%ld
956 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
957 to be "long long int" or "long int". This is needed for the internal
958 formatting function string_vformat().
960 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
961 the configuration file to be ":syslog", then the script "guesses" where
962 the logs files are, rather than using the compiled in default. In our
963 case the guess is not the same as the compiled default, so the script
964 suddenly stopped working when I started to use syslog. The patch checks
965 to see if log_file_path is "". If so, it attempts to read it from exim
966 with no configuration file to get the compiled in version, before it
967 falls back to the previous guessing code."
969 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
970 implementing BATV in an Exim configuration. See NewStuff for the gory
973 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
974 Makefile that are specific to HP-UX.
976 PH/18 If the "use_postmaster" option was set for a recipient callout together
977 with the "random" option, the postmaster address was used as the MAIL
978 FROM address for the random test, but not for the subsequent recipient
979 test. It is now used for both.
981 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
982 patch removes a few documentation additions to RFC 3028, because the
983 latest draft now contains them. It adds the new en;ascii-case comparator
984 and a new error check for 8bit text in MIME parts. Comparator and
985 require names are now matched exactly. I enabled the subaddress
986 extension, but it is not well tested yet (read: it works for me)."
988 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
989 rework some of the code of TK/09 above to avoid the hardwired use of
990 "%lld" and "long long". Replaced the call to snprintf() with a call to
993 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
994 records point to non-existent hosts", "retry timeout exceeded", and
995 "retry time not reached for any host after a long failure period".
997 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
998 experimental DomainKeys support:
1000 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1001 (2) On an error such as an illegally used "control", the wrong name for
1002 the control was given.
1004 These problems did NOT occur unless DomainKeys support was compiled.
1006 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1008 PH/24 Added ${if match_ip condition.
1010 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1011 databases so that it will be absolutely obvious if a crash occurs in the
1012 DB library. This is a regular occurrence (often caused by mis-matched
1015 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1016 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1017 picked up on a user's system that detects such things. There doesn't seem
1018 to be a gcc warning option for this - only an attribute that has to be
1019 put on the function's prototype. It seems that in Fedora Core 4 they have
1020 set this on a number of new functions. No doubt there will be more in due
1023 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1024 specify a transport. However, if an address that was verified by such a
1025 router was the subject of a callout, Exim crashed because it tried to
1026 read the rcpt_include_affixes from the non-existent transport. Now it
1027 just assumes that the setting of that option is false. This bug was
1028 introduced by 4.51/PH/31.
1030 PH/28 Changed -d+all to exclude +memory, because that information is very
1031 rarely of interest, but it makes the output a lot bigger. People tend to
1032 do -d+all out of habit.
1034 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1035 code in os-type was giving problems when libc.so lives in lib64, like on
1038 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1039 aren't the modern standard, and it seems that some systems' include files
1040 don't always have them. Exim was already checking for some of the newer
1041 ones like T_AAAA, and defining it itself. I've added checks for all the
1042 record types that Exim uses.
1044 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1045 not automatically generating a new one, as it is supposed to. This
1046 prevented TLS from working. If the file did exist, but contained invalid
1047 data, a new version was generated, as expected. It was only the case of a
1048 non-existent file that was broken.
1050 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1051 with a change in libdomainkeys > 0.64.
1053 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1054 from DNS. If the selector record carries the flag, it now has
1055 precedence over the domain-wide flag.
1057 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1059 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1060 the use of a port name that isn't defined in /etc/services) occurred, the
1061 message was deferred as in a normal delivery, and thus remained on the
1062 spool, instead of being failed because of the mua_wrapper setting. This
1063 is now fixed, and I tidied up some of the mua_wrapper messages at the
1066 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1067 the messages in a hash of arrays rather than using individual hashes.
1068 This is a bit cleaner and results in dramatic memory savings, albeit
1069 at a slight CPU cost.
1071 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1072 as requested by Marc Sherman.
1074 SC/10 Eximstats: added histograms for user specified patterns as requested
1077 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1079 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1080 fopen() in the content-scanning modules that did not already have it.
1086 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1087 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1089 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1091 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1093 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1094 to test Sieve filters that use "vacation".
1096 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1097 that changes the way the GnuTLS parameters are stored in the cache file.
1098 The new format can be generated externally. For backward compatibility,
1099 if the data in the cache doesn't make sense, Exim assumes it has read an
1100 old-format file, and it generates new data and writes a new file. This
1101 means that you can't go back to an older release without removing the
1104 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1105 work if there are any delivery delays because "one_time" forces the
1106 parent to be marked "delivered", so its unseen clone is never tried
1107 again. For this reason, Exim now forbids the simultaneous setting of
1110 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1111 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1112 entry if you want to know the details. The fix, however, neglected to
1113 consider the case when local delivery batching is involved. The test for
1114 "previously delivered" was not happening when checking to see if an
1115 address could be batched with a previous (undelivered) one; under
1116 certain circumstances this could lead to multiple deliveries to the same
1119 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1120 in its include files, and this causes problems building Exim.
1122 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1123 header_syntax) but Exim was just ignoring anything given after a slash.
1124 In particular, this caused confusion with an attempt to use "verify =
1125 reverse_host_lookup/defer_ok". An error is now given when options are
1126 supplied for verify items that do not have them. (Maybe reverse_host_
1127 lookup should have a defer_ok option, but that's a different point.)
1129 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1130 defined by RFC 821) to 2048, because there were problems with some AUTH
1131 commands, and RFC 1869 says the size should be increased for extended
1132 SMTP commands that take arguments.
1134 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1137 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1138 "unknown" error; now it says that the functionality isn't in the binary.
1140 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1141 an address' error message when a string expansion fails (syntax or
1142 whatever). Otherwise the password may appear in the log. Following change
1143 PH/42 below, there is no longer a chance of it appearing in a bounce
1146 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1148 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1149 of its addresses was ever tried. (Bugzilla bug #2.)
1151 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1152 the result incorrectly in the debug output. (It correctly added a newline
1153 to what was transported.)
1155 TF/01 Added $received_time.
1157 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1158 commented out examples of how to interface to a virus scanner and to
1159 SpamAssassin. Also added commented examples of av_scanner and
1160 spamd_address settings.
1162 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1163 and controls are allowed in which ACLs. There were a couple of minor
1164 errors. Some of the entries in the conditions table (which is a table of
1165 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1166 negation of where the condition IS allowed.
1168 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1170 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1171 header file does not have a version number, so I've had to invent a new
1172 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1173 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1174 radiusclient), but it was contributed by a Radius user.
1176 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1177 files or directories, for OpenSSL.
1179 PH/19 When an Exim process that is running as root has to create an Exim log
1180 file, it does so in a subprocess that runs as exim:exim so as to get the
1181 ownership right at creation (otherwise, other Exim processes might see
1182 the file with the wrong ownership). There was no test for failure of this
1183 fork() call, which would lead to the process getting stuck as it waited
1184 for a non-existent subprocess. Forks do occasionally fail when resources
1185 run out. I reviewed all the other calls to fork(); they all seem to check
1188 PH/20 When checking for unexpected SMTP input at connect time (before writing
1189 the banner), Exim was not dealing correctly with a non-positive return
1190 from the read() function. If the client had disconnected by this time,
1191 the result was a log entry for a synchronization error with an empty
1192 string after "input=" when read() returned zero. If read() returned -1
1193 (an event I could not check), uninitialized data bytes were printed.
1194 There were reports of junk text (parts of files, etc) appearing after
1197 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1199 PH/22 Added support for macro redefinition, and (re)definition in between
1200 driver and ACL definitions.
1202 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1203 forgetting to use the resulting value; it was using the unexpanded value.
1205 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1206 hadn't been configured. The fix is from Juergen Kreileder, who
1207 understands it better than I do:
1209 "Here's what I see happening with three configured cyrus_sasl
1210 authenticators configured (plain, login, cram-md5):
1212 On startup auth_cyrus_sasl_init() gets called for each of these.
1213 This means three calls to sasl_listmech() without a specified mech_list.
1214 => SASL tests which mechs of all available mechs actually work
1215 => three warnings about OTP not working
1216 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1218 With the patch, sasl_listmech() also gets called three times. But now
1219 SASL's mech_list option is set to the server_mech specified in the the
1220 authenticator. Or in other words, the answer from sasl_listmech()
1221 gets limited to just the mech you're testing for (which is different
1223 => the return list contains just 'plain' or 'login', 'cram-md5' or
1224 nothing depending on the value of ob->server_mech.
1226 I've just tested the patch: Authentication still works fine,
1227 unavailable mechs specified in the exim configuration are still
1228 caught, and the auth.log warnings about OTP are gone."
1230 PH/25 When debugging is enabled, the contents of the command line are added
1231 to the debugging output, even when log_selector=+arguments is not
1234 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1235 answer is "GNU", and only if the return is "GNU/something" is the answer
1238 PH/27 $acl_verify_message is now set immediately after the failure of a
1239 verification in an ACL, and so is available in subsequent modifiers. In
1240 particular, the message can be preserved by coding like this:
1242 warn !verify = sender
1243 set acl_m0 = $acl_verify_message
1245 Previously, $acl_verify_message was set only while expanding "message"
1246 and "log_message" when a very denied access.
1248 PH/28 Modified OS/os.c-Linux with
1250 -#ifndef OS_LOAD_AVERAGE
1251 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1253 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1254 nomenclature these days.)
1256 PH/29 Installed patch from the Sieve maintainer that adds the options
1257 sieve_useraddress and sieve_subaddress to the redirect router.
1259 PH/30 In these circumstances:
1260 . Two addresses routed to the same list of hosts;
1261 . First host does not offer TLS;
1262 . First host accepts first address;
1263 . First host gives temporary error to second address;
1264 . Second host offers TLS and a TLS session is established;
1265 . Second host accepts second address.
1266 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1267 and peerdn, if requested) that were in fact used only for the second
1270 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1271 attention to any local part prefix or suffix that was matched by the
1272 router that accepted the address. It now behaves in the same way as it
1273 does for delivery: the affixes are removed from the local part unless
1274 rcpt_include_affixes is set on the transport.
1276 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1277 timeout during the DATA phase of an incoming message.
1279 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1280 applied a patch sanctioned by the Sieve maintainer.
1282 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1283 the uid or gid is negative. A case of a negative gid caused this to be
1284 noticed. The fix allows for either to be negative.
1286 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1287 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1290 PH/36 The change PH/12 above was broken. Fixed it.
1292 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1293 the grounds that routing the same address twice would always produce the
1294 same answer. This might have been true once, but it is certainly no
1295 longer true now. Routing a child address may depend on the previous
1296 routing that produced that child. Some complicated redirection strategies
1297 went wrong when messages had multiple recipients, and made Exim's
1298 behaviour dependent on the order in which the addresses were given.
1300 I have moved the duplicate checking until after the routing is complete.
1301 Exim scans the addresses that are assigned to local and remote
1302 transports, and removes any duplicates. This means that more work will be
1303 done, as duplicates will always all be routed, but duplicates are
1304 presumably rare, so I don't expect this is of any significance.
1306 For deliveries to pipes, files, and autoreplies, the duplicate checking
1307 still happens during the routing process, since they are not going to be
1310 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1311 It corrects a timeout issue with spamd. This is Ian's comment: "The
1312 background is that sometimes spamd either never reads data from a
1313 connection it has accepted, or it never writes response data. The exiscan
1314 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1315 blindly assumes that writes won't block so it may never time out."
1317 PH/39 Allow G after quota size as well as K and M.
1319 PH/40 The value set for $authenticated_id in an authenticator may not contain
1320 binary zeroes or newlines because the value is written to log lines and
1321 to spool files. There was no check on this. Now the value is run through
1322 the string_printing() function so that such characters are converted to
1323 printable escape sequences.
1325 PH/41 $message_linecount is a new variable that contains the total number of
1326 lines in the message. Compare $body_linecount, which is the count for the
1329 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1330 bounce and delay warning messages, except in certain special cases, which
1333 (a) An SMTP error message from a remote host;
1334 (b) A message specified in a :fail: redirection;
1335 (c) A message specified in a "fail" command in a system filter;
1336 (d) A message specified in a FAIL return from the queryprogram router;
1337 (e) A message specified by the cannot_route_message router option.
1339 In these cases only, Exim does include the error details in bounce and
1340 warning messages. There are also a few cases where bland messages such
1341 as "unrouteable address" or "local delivery error" are given.
1343 PH/43 $value is now also set for the "else" part of a ${run expansion.
1345 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1346 being worked on, but at least Exim now implements the latest version to
1349 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1350 process to complete was treated as a delivery failure, a timeout while
1351 writing the message to the pipe was logged, but erroneously treated as a
1352 successful delivery. Such timeouts include transport filter timeouts. For
1353 consistency with the overall process timeout, these timeouts are now
1354 treated as errors, giving rise to delivery failures by default. However,
1355 there is now a new Boolean option for the pipe transport called
1356 timeout_defer, which, if set TRUE, converts the failures into defers for
1357 both kinds of timeout. A transport filter timeout is now identified in
1360 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1361 systems where "make" and "gmake" are different, calling "gmake" at top
1362 level broke things. I've arranged for the value of $(MAKE) to be passed
1363 from the Makefile to this script so that it can call the same version of
1367 A note about Exim versions 4.44 and 4.50
1368 ----------------------------------------
1370 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1371 changes of various kinds. As a consequence, a big documentation update was
1372 needed. This delayed the release for rather longer than seemed good, especially
1373 in the light of a couple of (minor) security issues. Therefore, the changes
1374 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1375 release. So 4.44 and 4.50 are in effect two different branches that both start
1378 I have left the 4.50 change log unchanged; it contains all the changes since
1379 4.43. The change log for 4.44 is below; many of its items are identical to
1380 those for 4.50. This seems to be the most sensible way to preserve the
1381 historical information.
1387 1. Minor wording change to the doc/README.SIEVE file.
1389 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1390 computation of the current number of files was incorrect.
1392 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1393 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1396 4. Give more explanation in the error message when the command for a transport
1397 filter fails to execute.
1399 5. There are several places where Exim runs a non-Exim command in a
1400 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1401 was being done only for the command run by the queryprogram router. It is
1402 now done for all such subprocesses. The other cases are: ${run, transport
1403 filters, and the commands run by the lmtp and pipe transports.
1405 6. Added CONFIGURE_GROUP build-time option.
1407 7. Some older OS have a limit of 256 on the maximum number of file
1408 descriptors. Exim was using setrlimit() to set 1000 as a large value
1409 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1410 systems. I've change it so that if it can't get 1000, it tries for 256.
1412 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1413 was an oversight, and furthermore, ever since the addition of extra
1414 controls (e.g. 4.43/32), the checks on when to allow different forms of
1415 "control" were broken. There should now be diagnostics for all cases when a
1416 control that does not make sense is encountered.
1418 9. Added the /retain_sender option to "control=submission".
1420 10. $recipients is now available in the predata ACL (oversight).
1422 11. Tidy the search cache before the fork to do a delivery from a message
1423 received from the command line. Otherwise the child will trigger a lookup
1424 failure and thereby defer the delivery if it tries to use (for example) a
1425 cached ldap connection that the parent has called unbind on.
1427 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1428 of $address_data from the recipient verification was clobbered by the
1429 sender verification.
1431 13. The value of address_data from a sender verification is now available in
1432 $sender_address_data in subsequent conditions in the ACL statement.
1434 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1436 15. Added a new option "connect=<time>" to callout options, to set a different
1439 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1440 was its contents. (It was OK if the option was not defined at all.)
1442 17. A "Completed" log line is now written for messages that are removed from
1443 the spool by the -Mrm option.
1445 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1446 information about exactly what failed.
1448 19. Added -dd to debug only the daemon process.
1450 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1451 handles timeouts, both on the server side and network timeouts. Renamed the
1452 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1454 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1455 It is now set to "smtps".
1457 22. $host_address is now set to the target address during the checking of
1458 ignore_target_hosts.
1460 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1461 being passed; this would have caused $sender_host_name to have been used if
1462 matching the list had actually called for a host name (not very likely,
1463 since this list is usually IP addresses). A host name is now passed as
1466 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1467 code that specifies a non-restarting handler (typically sigaction() in
1468 modern systems) in an attempt to fix a rare and obscure crash bug.
1470 25. Narrowed the window for a race in the daemon that could cause it to ignore
1471 SIGCHLD signals. This is not a major problem, because they are used only to
1472 wake it up if nothing else does.
1474 26. A malformed maildirsize file could cause Exim to calculate negative values
1475 for the mailbox size or file count. Odd effects could occur as a result.
1476 The maildirsize information is now recalculated if the size or filecount
1479 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1480 support for a long time. Removed HAVE_SYS_VFS_H.
1482 28. Installed the latest version of exipick from John Jetmore.
1484 29. In an address list, if the pattern was not a regular expression, an empty
1485 subject address (from a bounce message) matched only if the pattern was an
1486 empty string. Non-empty patterns were not even tested. This was the wrong
1487 because it is perfectly reasonable to use an empty address as part of a
1488 database query. An empty address is now tested by patterns that are
1489 lookups. However, all the other forms of pattern expect the subject to
1490 contain a local part and a domain, and therefore, for them, an empty
1491 address still always fails if the pattern is not itself empty.
1493 30. Exim went into a mad DNS loop when attempting to do a callout where the
1494 host was specified on an smtp transport, and looking it up yielded more
1495 than one IP address.
1497 31. Re-factored the code for checking spool and log partition space into a
1498 function that finds that data and another that does the check. The former
1499 is then used to implement four new variables: $spool_space, $log_space,
1500 $spool_inodes, and $log_inodes.
1502 32. The RFC2047 encoding function was originally intended for short strings
1503 such as real names; it was not keeping to the 75-character limit for
1504 encoded words that the RFC imposes. It now respects the limit, and
1505 generates multiple encoded words if necessary. To be on the safe side, I
1506 have increased the buffer size for the ${rfc2047: expansion operator from
1509 33. It is now permitted to omit both strings after an "if" condition; if the
1510 condition is true, the result is "true". As before, when the second string
1511 is omitted, a false condition yields an empty string. This makes it less
1512 cumbersome to write custom ACL and router conditions.
1514 34. Failure to deliver a bounce message always caused it to be frozen, even if
1515 there was an errors_to setting on the router. The errors_to setting is now
1518 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1519 canonical form (fully expanded) before being placed in
1520 $sender_host_address.
1522 36. The table in the code that translates DNS record types into text (T_A to
1523 "A" for instance) was missing entries for NS and CNAME. It is just possible
1524 that this could have caused confusion if both these types were looked up
1525 for the same domain, because the text type is used as part of Exim's
1526 per-process caching. But the chance of anyone hitting this buglet seems
1529 37. The dnsdb lookup has been extended in a number of ways.
1531 (1) There is a new type, "zns", which walks up the domain tree until it
1532 finds some nameserver records. It should be used with care.
1534 (2) There is a new type, "mxh", which is like "mx" except that it returns
1535 just the host names, not the priorities.
1537 (3) It is now possible to give a list of domains (or IP addresses) to be
1538 looked up. The behaviour when one of the lookups defers can be
1539 controlled by a keyword.
1541 (4) It is now possible to specify the separator character for use when
1542 multiple records are returned.
1544 38. The dnslists ACL condition has been extended: it is now possible to supply
1545 a list of IP addresses and/or domains to be looked up in a particular DNS
1548 39. Added log_selector=+queue_time_overall.
1550 40. When running the queue in the test harness, wait just a tad after forking a
1551 delivery process, to get repeatability of debugging output.
1553 41. Include certificate and key file names in error message when GnuTLS fails
1554 to set them up, because the GnuTLS error message doesn't include the name
1555 of the failing file when there is a problem reading it.
1557 42. Allow both -bf and -bF in the same test run.
1559 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1561 44. The "Exiscan patch" is now merged into the mainline Exim source.
1563 45. Sometimes the final signoff response after QUIT could fail to get
1564 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1565 < 0 before doing a fflush(). This bug looks as though it goes back to the
1566 introduction of TLS in release 3.20, but "sometimes" must have been rare
1567 because the tests only now provoked it.
1569 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1570 (this can affect the format of dates).
1572 47. exim_tidydb, when checking for the continued existence of a message for
1573 which it has found a message-specific retry record, was not finding
1574 messages that were in split spool directories. Consequently, it was
1575 deleting retry records that should have stayed in existence.
1577 48. Steve fixed some bugs in eximstats.
1579 49. The SPA authentication driver was not abandoning authentication and moving
1580 on to the next authenticator when an expansion was forced to fail,
1581 contradicting the general specification for all authenticators. Instead it
1582 was generating a temporary error. It now behaves as specified.
1584 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1585 (the order specifies the preference for clients). The order is now AES256,
1586 AES128, 3DES, ARCFOUR128.
1588 51. Small patch to Sieve code - explicitly set From: when generating an
1591 52. Exim crashed if a remote delivery caused a very long error message to be
1592 recorded - for instance if somebody sent an entire SpamAssassin report back
1593 as a large number of 550 error lines. This bug was coincidentally fixed by
1594 increasing the size of one of Exim's internal buffers (big_buffer) that
1595 happened as part of the Exiscan merge. However, to be on the safe side, I
1596 have made the code more robust (and fixed the comments that describe what
1599 53. Now that there can be additional text after "Completed" in log lines (if
1600 the queue_time_overall log selector is set), a one-byte patch to exigrep
1601 was needed to allow it to recognize "Completed" as not the last thing in
1604 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1605 patch that reportedly fixes this has been added. I am not expert enough to
1606 create a test for it. This is what the patch creator wrote:
1608 "I found a little strange behaviour of ldap code when working with
1609 Windows 2003 AD Domain, where users was placed in more than one
1610 Organization Units. When I tried to give exim partial DN, the exit code
1611 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1612 But simultaneously result of request was absolutely normal ldap result,
1613 so I produce this patch..."
1615 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1616 REFERENCE, so I have modified the code to exclude the patch when that macro
1619 55. Some experimental protocols are using DNS PTR records for new purposes. The
1620 keys for these records are domain names, not reversed IP addresses. The
1621 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1622 leaves it alone. Component reversal etc. now happens only for IP addresses.
1625 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1627 57. Double the size of the debug message buffer (to 2048) so that more of very
1628 long debug lines gets shown.
1630 58. The exicyclog utility now does better if the number of log files to keep
1631 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1633 59. Two changes related to the smtp_active_hostname option:
1635 (1) $smtp_active_hostname is now available as a variable.
1636 (2) The default for smtp_banner uses $smtp_active_hostname instead
1637 of $primary_hostname.
1639 60. The host_aton() function is supposed to be passed a string that is known
1640 to be a valid IP address. However, in the case of IPv6 addresses, it was
1641 not checking this. This is a hostage to fortune. Exim now panics and dies
1642 if the condition is not met. A case was found where this could be provoked
1643 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1644 components; fortuitously, this particular loophole had already been fixed
1645 by change 4.50/55 above.
1647 If there are any other similar loopholes, the new check in host_aton()
1648 itself should stop them being exploited. The report I received stated that
1649 data on the command line could provoke the exploit when Exim was running as
1650 exim, but did not say which command line option was involved. All I could
1651 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1652 running as the user.
1655 61. There was a buffer overflow vulnerability in the SPA authentication code
1656 (which came originally from the Samba project). I have added a test to the
1657 spa_base64_to_bits() function which I hope fixes it.
1660 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1661 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1663 63. The daemon start-up calls getloadavg() while still root for those OS that
1664 need the first call to be done as root, but it missed one case: when
1665 deliver_queue_load_max is set with deliver_drop_privilege. This is
1666 necessary for the benefit of the queue runner, because there is no re-exec
1667 when deliver_drop_privilege is set.
1669 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1670 This has been fixed.
1672 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1673 was in use, was not putting the data itself into the right store pool;
1674 consequently, it could be overwritten for a subsequent message in the same
1675 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1678 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1680 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1681 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1682 the function were treating the return as a boolean value, which happened to
1683 work because 0=false and not-0=true, but is not correct code.
1685 68. The host_aton() function was not handling scoped IPv6 addresses (those
1686 with, for example, "%eth0" on the end) correctly.
1688 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1689 negated items (that is, ~something) in unsigned ints. Some compilers
1690 apparently mutter when there is no cast.
1692 70. If an address verification called from an ACL failed, and did not produce a
1693 user-specific message (i.e. there was only a "system" message), nothing was
1694 put in $acl_verify_message. In this situation, it now puts the system
1697 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1698 session; change 4.31/43 added the unwanted input to the log line - except
1699 that it did not do this in the start of session case. It now does.
1701 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1702 This is wrong and can cause the other end to generate a synchronization
1703 error if it is another Exim or anything else that does the synchronization
1704 check. A QUIT command is no longer sent after a timeout.
1706 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1707 during host lookups.
1709 74. The defer_ok option of callout verification was not working if it was used
1710 when verifying addresses in header lines, that is, for this case:
1712 verify = header_sender/callout=defer_ok
1714 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1715 those file descriptors could be used for SMTP connections. If anything
1716 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1717 could be sent to the SMTP client, causing chaos. The daemon now opens
1718 stdin, stdout, and stderr to /dev/null when it puts itself into the
1721 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1722 log by default. The user can override this with suitable Perl magic.
1724 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1725 the log message when discard triggers, was not working for the DATA ACL or
1726 for the non-SMTP ACL.
1728 78. Error message wording change in sieve.c.
1730 79. If smtp_accept_max_per_host was set, the number of connections could be
1731 restricted to fewer than expected, because the daemon was trying to set up
1732 a new connection before checking whether the processes handling previous
1733 connections had finished. The check for completed processes is now done
1734 earlier. On busy systems, this bug wouldn't be noticed because something
1735 else would have woken the daemon, and it would have reaped the completed
1738 80. If a message was submitted locally by a user whose login name contained one
1739 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1740 It caused a spool format error. I have fixed the spool reading code. A
1741 related problem was that the "from" clause in the Received: line became
1742 illegal because of the space(s). It is now covered by ${quote_local_part.
1744 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1747 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1749 83. Patch from Sieve maintainer to fix unterminated string problem in
1750 "vacation" handling.
1752 84. Some minor changes to the Linux configuration files to help with other
1753 OS variants using glibc.
1755 85. One more patch for Sieve to update vacation handling to latest spec.
1758 ----------------------------------------------------
1759 See the note above about the 4.44 and 4.50 releases.
1760 ----------------------------------------------------
1766 1. Change 4.43/35 introduced a bug that caused file counts to be
1767 incorrectly computed when quota_filecount was set in an appendfile
1770 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1771 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1774 3. Give more explanation in the error message when the command for a transport
1775 filter fails to execute.
1777 4. There are several places where Exim runs a non-Exim command in a
1778 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1779 was being done only for the command run by the queryprogram router. It is
1780 now done for all such subprocesses. The other cases are: ${run, transport
1781 filters, and the commands run by the lmtp and pipe transports.
1783 5. Some older OS have a limit of 256 on the maximum number of file
1784 descriptors. Exim was using setrlimit() to set 1000 as a large value
1785 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1786 systems. I've change it so that if it can't get 1000, it tries for 256.
1788 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1789 was an oversight, and furthermore, ever since the addition of extra
1790 controls (e.g. 4.43/32), the checks on when to allow different forms of
1791 "control" were broken. There should now be diagnostics for all cases when a
1792 control that does not make sense is encountered.
1794 7. $recipients is now available in the predata ACL (oversight).
1796 8. Tidy the search cache before the fork to do a delivery from a message
1797 received from the command line. Otherwise the child will trigger a lookup
1798 failure and thereby defer the delivery if it tries to use (for example) a
1799 cached ldap connection that the parent has called unbind on.
1801 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1802 of $address_data from the recipient verification was clobbered by the
1803 sender verification.
1805 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1806 was its contents. (It was OK if the option was not defined at all.)
1808 11. A "Completed" log line is now written for messages that are removed from
1809 the spool by the -Mrm option.
1811 12. $host_address is now set to the target address during the checking of
1812 ignore_target_hosts.
1814 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1815 being passed; this would have caused $sender_host_name to have been used if
1816 matching the list had actually called for a host name (not very likely,
1817 since this list is usually IP addresses). A host name is now passed as
1820 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1821 code that specifies a non-restarting handler (typically sigaction() in
1822 modern systems) in an attempt to fix a rare and obscure crash bug.
1824 15. Narrowed the window for a race in the daemon that could cause it to ignore
1825 SIGCHLD signals. This is not a major problem, because they are used only to
1826 wake it up if nothing else does.
1828 16. A malformed maildirsize file could cause Exim to calculate negative values
1829 for the mailbox size or file count. Odd effects could occur as a result.
1830 The maildirsize information is now recalculated if the size or filecount
1833 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1834 support for a long time. Removed HAVE_SYS_VFS_H.
1836 18. Updated exipick to current release from John Jetmore.
1838 19. Allow an empty sender to be matched against a lookup in an address list.
1839 Previously the only cases considered were a regular expression, or an
1842 20. Exim went into a mad DNS lookup loop when doing a callout where the
1843 host was specified on the transport, if the DNS lookup yielded more than
1846 21. The RFC2047 encoding function was originally intended for short strings
1847 such as real names; it was not keeping to the 75-character limit for
1848 encoded words that the RFC imposes. It now respects the limit, and
1849 generates multiple encoded words if necessary. To be on the safe side, I
1850 have increased the buffer size for the ${rfc2047: expansion operator from
1853 22. Failure to deliver a bounce message always caused it to be frozen, even if
1854 there was an errors_to setting on the router. The errors_to setting is now
1857 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1858 canonical form (fully expanded) before being placed in
1859 $sender_host_address.
1861 24. Updated eximstats to version 1.33
1863 25. Include certificate and key file names in error message when GnuTLS fails
1864 to set them up, because the GnuTLS error message doesn't include the name
1865 of the failing file when there is a problem reading it.
1867 26. Expand error message when OpenSSL has problems setting up cert/key files.
1870 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1871 (this can affect the format of dates).
1873 28. exim_tidydb, when checking for the continued existence of a message for
1874 which it has found a message-specific retry record, was not finding
1875 messages that were in split spool directories. Consequently, it was
1876 deleting retry records that should have stayed in existence.
1878 29. eximstats updated to version 1.35
1879 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1880 1.35 - bugfix such that pie charts by volume are generated correctly
1882 30. The SPA authentication driver was not abandoning authentication and moving
1883 on to the next authenticator when an expansion was forced to fail,
1884 contradicting the general specification for all authenticators. Instead it
1885 was generating a temporary error. It now behaves as specified.
1887 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1888 (the order specifies the preference for clients). The order is now AES256,
1889 AES128, 3DES, ARCFOUR128.
1891 31. Small patch to Sieve code - explicitly set From: when generating an
1894 32. Exim crashed if a remote delivery caused a very long error message to be
1895 recorded - for instance if somebody sent an entire SpamAssassin report back
1896 as a large number of 550 error lines. This bug was coincidentally fixed by
1897 increasing the size of one of Exim's internal buffers (big_buffer) that
1898 happened as part of the Exiscan merge. However, to be on the safe side, I
1899 have made the code more robust (and fixed the comments that describe what
1902 33. Some experimental protocols are using DNS PTR records for new purposes. The
1903 keys for these records are domain names, not reversed IP addresses. The
1904 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1905 leaves it alone. Component reversal etc. now happens only for IP addresses.
1908 34. The host_aton() function is supposed to be passed a string that is known
1909 to be a valid IP address. However, in the case of IPv6 addresses, it was
1910 not checking this. This is a hostage to fortune. Exim now panics and dies
1911 if the condition is not met. A case was found where this could be provoked
1912 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1913 components; fortuitously, this particular loophole had already been fixed
1914 by change 4.50/55 or 4.44/33 above.
1916 If there are any other similar loopholes, the new check in host_aton()
1917 itself should stop them being exploited. The report I received stated that
1918 data on the command line could provoke the exploit when Exim was running as
1919 exim, but did not say which command line option was involved. All I could
1920 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1921 running as the user.
1924 35. There was a buffer overflow vulnerability in the SPA authentication code
1925 (which came originally from the Samba project). I have added a test to the
1926 spa_base64_to_bits() function which I hope fixes it.
1929 36. The daemon start-up calls getloadavg() while still root for those OS that
1930 need the first call to be done as root, but it missed one case: when
1931 deliver_queue_load_max is set with deliver_drop_privilege. This is
1932 necessary for the benefit of the queue runner, because there is no re-exec
1933 when deliver_drop_privilege is set.
1935 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1936 was in use, was not putting the data itself into the right store pool;
1937 consequently, it could be overwritten for a subsequent message in the same
1938 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1941 38. Sometimes the final signoff response after QUIT could fail to get
1942 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1943 < 0 before doing a fflush(). This bug looks as though it goes back to the
1944 introduction of TLS in release 3.20, but "sometimes" must have been rare
1945 because the tests only now provoked it.
1951 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1952 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1953 Now it returns an integer. A number of calls were still expecting a T/F
1954 return. Fortuitously, in all cases, the tests worked in OK situations,
1955 which is the norm. However, things would have gone wrong on any write
1956 failures on the smtp file descriptor. This function is used when sending
1957 messages over SMTP and also when doing verify callouts.
1959 2. When Exim is called to do synchronous delivery of a locally submitted
1960 message (the -odf or -odi options), it no longer closes stderr before doing
1963 3. Implemented the mua_wrapper option.
1965 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1967 5. Implemented the functions header_remove(), header_testname(),
1968 header_add_at_position(), and receive_remove_recipient(), and exported them
1971 6. If an ACL "warn" statement specified the addition of headers, Exim already
1972 inserted X-ACL-Warn: at the start if there was no header name. However, it
1973 was not making this test for the second and subsequent header lines if
1974 there were newlines in the string. This meant that an invalid header could
1975 be inserted if Exim was badly configured.
1977 7. Allow an ACL "warn" statement to add header lines at the start or after all
1978 the Received: headers, as well as at the end.
1980 8. Added the rcpt_4xx retry error code.
1982 9. Added postmaster_mailfrom=xxx to callout verification option.
1984 10. Added mailfrom=xxxx to the callout verification option, for verify=
1987 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1988 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1990 12. Inserted some casts to stop certain compilers warning when using pointer
1991 differences as field lengths or precisions in printf-type calls (mostly
1992 affecting debugging statements).
1994 13. Added optional readline() support for -be (dynamically loaded).
1996 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1997 same clock tick as a message's arrival, so that its received time was the
1998 same as the "first fail" time on the retry record, and that message
1999 remained on the queue past the ultimate address timeout, every queue runner
2000 would try a delivery (because it was past the ultimate address timeout) but
2001 after another failure, the ultimate address timeout, which should have then
2002 bounced the address, did not kick in. This was a "< instead of <=" error;
2003 in most cases the first failure would have been in the next clock tick
2004 after the received time, and all would be well.
2006 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2007 being recognized when the domain list was tested by the match_domain
2008 condition in an expansion string.
2010 16. Added the ${str2b64: operator.
2012 17. Exim was always calling setrlimit() to set a large limit for the number of
2013 processes, without checking whether the existing limit was already
2014 adequate. (It did check for the limit on file descriptors.) Furthermore,
2015 errors from getrlimit() and setrlimit() were being ignored. Now they are
2016 logged to the main and panic logs, but Exim does carry on, to try to do its
2017 job under whatever limits there are.
2019 18. Imported PCRE 5.0.
2021 19. Trivial typo in log message " temporarily refused connection" (the leading
2024 20. If the log selector return_path_on_delivery was set and an address was
2025 redirected to /dev/null, the delivery process crashed because it assumed
2026 that a return path would always be set for a "successful" delivery. In this
2027 case, the whole delivery is bypassed as an optimization, and therefore no
2030 21. Internal re-arrangement: the function for sending a challenge and reading
2031 a response while authentication was assuming a zero-terminated challenge
2032 string. It's now changed to take a pointer and a length, to allow for
2033 binary data in such strings.
2035 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2037 23. Exim was not respecting finduser_retries when seeking the login of the
2038 uid under which it was called; it was always trying 10 times. (The default
2039 setting of finduser_retries is zero.) Also, it was sleeping after the final
2040 failure, which is pointless.
2042 24. Implemented tls_on_connect_ports.
2044 25. Implemented acl_smtp_predata.
2046 26. If the domain in control=submission is set empty, Exim assumes that the
2047 authenticated id is a complete email address when it generates From: or
2048 Sender: header lines.
2050 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2051 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2052 chown and chgrp in /bin and hostname in /usr/bin.
2054 28. Exim was keeping the "process log" file open after each use, just as it
2055 does for the main log. This opens the possibility of it remaining open for
2056 long periods when the USR1 signal hits a daemon. Occasional processlog
2057 errors were reported, that could have been caused by this. Anyway, it seems
2058 much more sensible not to leave this file open at all, so that is what now
2061 29. The long-running daemon process does not normally write to the log once it
2062 has entered its main loop, and it closes the log before doing so. This is
2063 so that log files can straightforwardly be renamed and moved. However,
2064 there are a couple of unusual error situations where the daemon does write
2065 log entries, and I had neglected to close the log afterwards.
2067 30. The text of an SMTP error response that was received during a remote
2068 delivery was being truncated at 512 bytes. This is too short for some of
2069 the long messages that one sometimes sees. I've increased the limit to
2072 31. It is now possible to make retry rules that apply only when a message has a
2073 specific sender, in particular, an empty sender.
2075 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2076 it possible to be selective about when SMTP synchronization is enforced.
2078 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2080 32. Implemented hosts_connection_nolog.
2082 33. Added an ACL for QUIT.
2084 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2087 35. Added mailbox_size and mailbox_filecount to appendfile.
2089 36. Added control = no_multiline_responses to ACLs.
2091 37. There was a bug in the logic of the code that waits for the clock to tick
2092 in the case where the clock went backwards by a substantial amount such
2093 that the microsecond fraction of "now" was more than the microsecond
2094 fraction of "then" (but the whole seconds number was less).
2096 38. Added support for the libradius Radius client library this is found on
2097 FreeBSD (previously only the radiusclient library was supported).
2103 1. When certain lookups returned multiple values in the form name=value, the
2104 quoting of the values was not always being done properly. Specifically:
2105 (a) If the value started with a double quote, but contained no whitespace,
2107 (b) If the value contained whitespace other than a space character (i.e.
2108 tabs or newlines or carriage returns) it was not quoted.
2109 This fix has been applied to the mysql and pgsql lookups by writing a
2110 separate quoting function and calling it from the lookup code. The fix
2111 should probably also be applied to nisplus, ibase and oracle lookups, but
2112 since I cannot test any of those, I have not disturbed their existing code.
2114 2. A hit in the callout cache for a specific address caused a log line with no
2115 reason for rejecting RCPT. Now it says "Previous (cached) callout
2116 verification failure".
2118 3. There was an off-by-one bug in the queryprogram router. An over-long
2119 return line was truncated at 256 instead of 255 characters, thereby
2120 overflowing its buffer with the terminating zero. As well as fixing this, I
2121 have increased the buffer size to 1024 (and made a note to document this).
2123 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2124 when Exim is waiting for an SMTP response from a remote server, Exim
2125 restarts its select() call on the socket, thereby resetting its timeout.
2126 This is not a problem when such interrupts are rare. Somebody set up a cron
2127 job to run exiwhat every 2 minutes, which is less than the normal select()
2128 timeout (5 or 10 minutes). This meant that the select() timeout never
2129 kicked in because it was always reset. I have fixed this by comparing the
2130 time when an interrupt arrives with the time at the start of the first call
2131 to select(). If more time than the timeout has elapsed, the interrupt is
2132 treated as a timeout.
2134 5. Some internal re-factoring in preparation for the addition of Sieve
2135 extensions (by MH). In particular, the "personal" test is moved to a
2136 separate function, and given an option for scanning Cc: and Bcc: (which is
2137 not set for Exim filters).
2139 6. When Exim created an email address using the login of the caller as the
2140 local part (e.g. when creating a From: or Sender: header line), it was not
2141 quoting the local part when it contained special characters such as @.
2143 7. Installed new OpenBSD configuration files.
2145 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2146 try to make them clearer.
2148 9. Callout options, other than the timeout value, were being ignored when
2149 verifying sender addresses in header lines. For example, when using
2151 verify = header_sender/callout=no_cache
2153 the cache was (incorrectly) being used.
2155 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2156 only the Cygwin environment.
2158 11. When return_path_on_delivery was set as a log selector, if different remote
2159 addresses in the same message used different return paths and parallel
2160 remote delivery occurred, the wrong values would sometimes be logged.
2161 (Whenever a remote delivery process finished, the return path value from
2162 the most recently started remote delivery process was logged.)
2164 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2165 lines when AUTH and/or TLS are in use. This is the "received protocol"
2166 field. Exim used to use "asmtp" for authenticated SMTP, without any
2167 indication (in the protocol name) for TLS use. Now it follows the RFC and
2168 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2169 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2170 names appear in log lines as well as in Received: header lines.
2172 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2173 extensions, and comparison tests, and to fix some bugs.
2175 14. Changes to the "personal" filter test:
2177 (1) The test was buggy in that it was just doing the equivalent of
2178 "contains" tests on header lines. For example, if a user's address was
2179 anne@some.where, the "personal" test would incorrectly be true for
2181 To: susanne@some.where
2183 This test is now done by extracting each address from the header in turn,
2184 and checking the entire address. Other tests that are part of "personal"
2185 are now done using regular expressions (for example, to check local parts
2186 of addresses in From: header lines).
2188 (2) The list of non-personal local parts in From: addresses has been
2189 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2190 taken from the Sieve specification recommendations.
2192 (3) If the message contains any header line starting with "List-" it is
2193 treated as non-personal.
2195 (4) The test for "circular" in the Subject: header line has been removed
2196 because it now seems ill-conceived.
2198 15. Minor typos in src/EDITME comments corrected.
2200 16. Installed latest exipick from John Jetmore.
2202 17. If headers_add on a router specified a text string that was too long for
2203 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2204 of string_sprintf() is now avoided.
2206 18. $message_body_size was not set (it was always zero) when running the DATA
2207 ACL and the local_scan() function.
2209 19. For the "mail" command in an Exim filter, no default was being set for
2210 the once_repeat time, causing a random time value to be used if "once" was
2211 specified. (If the value happened to be <= 0, no repeat happened.) The
2212 default is now 0s, meaning "never repeat". The "vacation" command was OK
2213 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2214 (I found it when inspecting the code).
2216 20. There is now an overall timeout for performing a callout verification. It
2217 defaults to 4 times the callout timeout, which applies to individual SMTP
2218 commands during the callout. The overall timeout applies when there is more
2219 than one host that can be tried. The timeout is checked before trying the
2220 next host. This prevents very long delays if there are a large number of
2221 hosts and all are timing out (e.g. when the network connections are timing
2222 out). The value of the overall timeout can be changed by specifying an
2223 additional sub-option for "callout", called "maxwait". For example:
2225 verify = sender/callout=5s,maxwait=20s
2227 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2228 to the end before writing, but this should make it even safer).
2230 22. Exim was forgetting that it had advertised PIPELINING for the second and
2231 subsequent messages on an SMTP connection. It was also not resetting its
2232 memory on STARTTLS and an internal HELO.
2234 23. When Exim logs an SMTP synchronization error within a session, it now
2235 records whether PIPELINING has been advertised or not.
2237 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2238 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2239 rather than long int.
2241 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2243 26. Added the never_mail option to autoreply.
2249 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2250 crash if the getsockname() call failed; this can happen if a connection is
2251 closed very soon after it is established. The problem was simply in the
2252 order in which certain operations were done, causing Exim to try to write
2253 to the SMTP stream before it had set up the file descriptor. The bug has
2254 been fixed by making things happen in the correct order.
2260 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2261 before the connection was closed, thus losing the rejection response.
2263 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2264 some early Solaris releases, but causes trouble in current releases where
2265 socklen_t is defined.
2267 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2270 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2273 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2274 that releasing the top of it at the end releases what was used for sub-
2275 expansions (unless the block got too big). However, discard this block if
2276 the first thing is a variable or header, so that we can use its block when
2277 it is dynamic (useful for very large $message_headers, for example).
2279 6. Lookups now cache *every* query, not just the most recent. A new, separate
2280 store pool is used for this. It can be recovered when all lookup caches are
2281 flushed. Lookups now release memory at the end of their result strings.
2282 This has involved some general refactoring of the lookup sources.
2284 7. Some code has been added to the store_xxx() functions to reduce the amount
2285 of flapping under certain conditions.
2287 8. log_incoming_interface used to affect only the <= reception log lines. Now
2288 it causes the local interface and port to be added to several more SMTP log
2289 lines, for example "SMTP connection from", and rejection lines.
2291 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2293 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2295 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2296 could be overwritten at the end of the current message (or the start of a
2297 new message if it was set in a HELO ACL). The value is now preserved for
2298 the duration of the SMTP connection.
2300 12. If a transport had a headers_rewrite setting, and a matching header line
2301 contained an unqualified address, that address was qualified, even if it
2302 did not match any rewriting rules. The underlying bug was that the values
2303 of the flags that permit the existence of unqualified sender and recipient
2304 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2305 non-local messages, and by -bnq for local messages) were not being
2306 preserved with the message after it was received.
2308 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2309 "next input=" as part of the text comprising the host identity instead of
2310 the correct text. The code was using the same buffer for two different
2311 strings. However, depending on which order the printing function evaluated
2312 its arguments, the bug did not always show up. Under Linux, for example, my
2313 test suite worked just fine.
2315 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2316 doesn't work with some older versions of Perl. It has been changed to "my",
2317 which in any case is probably the better facility to use.
2319 15. A really picky compiler found some instances of statements for creating
2320 error messages that either had too many or two few arguments for the format
2323 16. The size of the buffer for calls to the DNS resolver has been increased
2324 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2325 for addresses that have a lot of PTR records. This alleviates a problem; it
2326 does not fully solve it.
2328 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2329 buffer now truncates the list and logs the incident, which is the same
2330 action as happens when Exim is looking up a host name and its aliases.
2331 Previously in this situation something unpredictable would happen;
2332 sometimes it was "internal error: store_reset failed".
2334 18. If a server dropped the connection unexpectedly when an Exim client was
2335 using GnuTLS and trying to read a response, the client delivery process
2336 crashed while trying to generate an error log message.
2338 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2339 string, for example:
2341 warn message = H1: something\nH2: something
2343 the text was added as a single header line from Exim's point of view
2344 though it ended up OK in the delivered message. However, searching for the
2345 second and subsequent header lines using $h_h2: did not work. This has been
2346 fixed. Similarly, if a system filter added multiple headers in this way,
2347 the routers could not see them.
2349 20. Expanded the error message when iplsearch is called with an invalid key to
2350 suggest using net-iplsearch in a host list.
2352 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2353 ACLs are no longer actually imposed (and a message to that effect is
2356 22. If a "gecos" field in a passwd entry contained escaped characters, in
2357 particular, if it contained a \" sequence, Exim got it wrong when building
2358 a From: or a Sender: header from that name. A second bug also caused
2359 incorrect handling when an unquoted " was present following a character
2360 that needed quoting.
2362 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2363 was not being matched caselessly.
2365 24. Arranged for all hyphens in the exim.8 source to be escaped with
2368 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2369 properly. Recipient callout cache records were still being keyed to include
2370 the sender, even when use_sender was set false. This led to far more
2371 callouts that were necessary. The sender is no longer included in the key
2372 when use_sender is false.
2374 26. Added "control = submission" modifier to ACLs.
2376 27. Added the ${base62d: operator to decode base 62 numbers.
2378 28. dnsdb lookups can now access SRV records.
2380 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2381 the configuration file.
2383 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2384 (-v) mode. This makes the output for a verbose queue run more intelligible.
2386 31. Added a use_postmaster feature to recipient callouts.
2388 32. Added the $body_zerocount variable, containing the number of binary zero
2389 bytes in the message body.
2391 33. The time of last modification of the "new" subdirectory is now used as the
2392 "mailbox time last read" when there is a quota error for a maildir
2395 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2397 35. Added +ignore_unknown as a special item in host lists.
2399 36. Code for decoding IPv6 addresses in host lists is now included, even if
2400 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2401 address was recognized as an IP address, but was then not correctly decoded
2402 into binary, causing unexpected and incorrect effects when compared with
2409 1. Very minor rewording of debugging text in manualroute to say "list of
2410 hosts" instead of "hostlist".
2412 2. If verify=header_syntax was set, and a header line with an unqualified
2413 address (no domain) and a large number of spaces between the end of the
2414 name and the colon was received, the reception process suffered a buffer
2415 overflow, and (when I tested it) crashed. This was caused by some obsolete
2416 code that should have been removed. The fix is to remove it!
2418 3. When running in the test harness, delay a bit after writing a bounce
2419 message to get a bit more predictability in the log output.
2421 4. Added a call to search_tidyup() just before forking a reception process. In
2422 theory, someone could use a lookup in the expansion of smtp_accept_max_
2423 per_host which, without the tidyup, could leave open a database connection.
2425 5. Added the variables $recipient_data and $sender_data which get set from a
2426 lookup success in an ACL "recipients" or "senders" condition, or a router
2427 "senders" option, similar to $domain_data and $local_part_data.
2429 6. Moved the writing of debug_print from before to after the "senders" test
2432 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2433 problems for message scanning, either using a data ACL, or using
2434 local_scan() because the Received: header was not generated till after they
2435 were called (in order to set the time as the time of reception completion).
2436 I have revised the way this works. The header is now generated after the
2437 body is received, but before the ACL or local_scan() are called. After they
2438 are run, the timestamp in the header is updated.
2444 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2445 before starting a queue runner without re-exec. This happened only when
2446 deliver_drop_privilege was set or when the Exim user was set to root. The
2447 effect of the bug was that timeouts during subsequent deliveries caused
2448 crashes instead of being properly handled. The handler is now left at its
2449 default (and expected) setting.
2451 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2452 message, again when deliver_drop_privilege is set or Exim is run as root.
2453 The bug described in (1) was not present in this case, but the tidying up
2454 of the other signals was missing. I have made the two cases consistent.
2456 3. The ignore_target_hosts setting on a manualroute router was being ignored
2457 for hosts that were looked up using the /MX notation.
2459 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2462 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2463 operated on the sender address. After changing the $sender_address to <>
2464 for the sender address verify, Exim was re-instated it as the original
2465 (before rewriting) address, but remembering that it had rewritten it, so it
2466 wasn't rewriting it again. This bug also had the effect of breaking the
2467 sender address verification caching when the sender address was rewritten.
2469 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2470 This has been changed so that if the ip literal address matches
2471 ignore_target_hosts, the router declines.
2473 7. Added expansion conditions match_domain, match_address, and match_local_
2474 part (NOT match_host).
2476 8. The placeholder for the Received: header didn't have a length field set.
2478 9. Added code to Exim itself and to exim_lock to test for a specific race
2479 condition that could lead to file corruption when using MBX delivery. The
2480 issue is with the lockfile that is created in /tmp. If this file is removed
2481 after a process has opened it but before that process has acquired a lock,
2482 there is the potential for a second process to recreate the file and also
2483 acquire a lock. This could lead to two Exim processes writing to the file
2484 at the same time. The added code performs the same test as UW imapd; it
2485 checks after acquiring the lock that its file descriptor still refers to
2486 the same named file.
2488 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2489 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2492 11. Added the smtp_active_hostname option. If used, this will typically be made
2493 to depend on the incoming interface address. Because $interface_address is
2494 not set up until the daemon has forked a reception process, error responses
2495 that can happen earlier (such as "too many connections") no longer contain
2498 12. If an expansion in a condition on a "warn" statement fails because a lookup
2499 defers, the "warn" statement is abandoned, and the next ACL statement is
2500 processed. Previously this caused the whole ACL to be aborted.
2502 13. Added the iplsearch lookup type.
2504 14. Added ident_timeout as a log selector.
2506 15. Added tls_certificate_verified as a log selector.
2508 16. Added a global option tls_require_ciphers (compare the smtp transport
2509 option of the same name). This controls incoming TLS connections.
2511 17. I finally figured out how to make tls_require_ciphers do a similar thing
2512 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2513 before starting the TLS session.
2515 18. Tabs are now shown as \t in -bP output.
2517 19. If the log selector return_path_on_delivery was set, Exim crashed when
2518 bouncing a message because it had too many Received: header lines.
2520 20. If two routers both had headers_remove settings, and the first one included
2521 a superfluous trailing colon, the final name in the first list and the
2522 first name in the second list were incorrectly joined into one item (with a
2523 colon in the middle).
2529 1. Added -C and -D options to the exinext utility, mainly to make it easier
2530 to include in the automated testing, but these could be helpful when
2531 multiple configurations are in use.
2533 2. The exinext utility was not formatting the output nicely when there was
2534 an alternate port involved in the retry record key, nor when there was a
2535 message id as well (for retries that were specific to a specific message
2536 and a specific host). It was also confused by IPv6 addresses, because of
2537 the additional colons they contain. I have fixed the IPv4 problem, and
2538 patched it up to do a reasonable job for IPv6.
2540 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2541 delivery, the log line now contains "pipelined" if PIPELINING was used.
2543 4. An SMTP transport process used to panic and die if the bind() call to set
2544 an explicit outgoing interface failed. This has been changed; it is now
2545 treated in the same way as a connect() failure.
2547 5. A reference to $sender_host_name in the part of a conditional expansion
2548 that was being skipped was still causing a DNS lookup. This no longer
2551 6. The def: expansion condition was not recognizing references to header lines
2552 that used bh_ and bheader_.
2554 7. Added the _cache feature to named lists.
2556 8. The code for checking quota_filecount in the appendfile transport was
2557 allowing one more file than it should have been.
2559 9. For compatibility with Sendmail, the command line option
2567 and sets the incoming protocol and host name (for trusted callers). The
2568 host name and its colon can be omitted when only the protocol is to be set.
2569 Note the Exim already has two private options, -pd and -ps, that refer to
2570 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2571 "s", but I don't think that's a major issue.
2573 10. A number of refactoring changes to the code, none of which should affect
2576 (a) The number of logging options was getting close to filling up the
2577 32-bit word that was used as a bit map. I have split them into two classes:
2578 those that are passed in the argument to log_write(), and those that are
2579 only ever tested independently outside of that function. These are now in
2580 separate 32-bit words, so there is plenty of room for expansion again.
2581 There is no change in the user interface or the logging behaviour.
2583 (b) When building, for example, log lines, the code previously used a
2584 macro that called string_cat() twice, in order to add two strings. This is
2585 not really sufficiently general. Furthermore, there was one instance where
2586 it was actually wrong because one of the argument was used twice, and in
2587 one call a function was used. (As it happened, calling the function twice
2588 did not affect the overall behaviour.) The macro has been replaced by a
2589 function that can join an arbitrary number of extra strings onto a growing
2592 (c) The code for expansion conditions now uses a table and a binary chop
2593 instead of a serial search (which was left over from when there were very
2594 few conditions). Also, it now recognizes conditions like "pam" even when
2595 the relevant support is not compiled in: a suitably worded error message is
2596 given if an attempt is made to use such a condition.
2598 11. Added ${time_interval:xxxxx}.
2600 12. A bug was causing one of the ddress fields not to be passed back correctly
2601 from remote delivery subprocesses. The field in question was not being
2602 subsequently used, so this caused to problems in practice.
2604 13. Added new log selectors queue_time and deliver_time.
2606 14. Might have fixed a bug in maildirsizefile handling that threw up
2607 "unexpected character" debug warnings, and recalculated the data
2608 unnecessarily. In any case, I expanded the warning message to give more
2611 15. Added the message "Restricted characters in address" to the statements in
2612 the default ACL that block characters like @ and % in local parts.
2614 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2615 Three changes have been made:
2617 (a) There was a serious bug; a negative response to MAIL caused the whole
2618 recipient domain to be cached as invalid, thereby blocking all messages
2619 to all local parts at the same domain, from all senders. This bug has
2620 been fixed. The domain is no longer cached after a negative response to
2621 MAIL if the sender used is not empty.
2623 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2626 (c) A new callout option, "use_sender" has been added for people who want
2627 the modified behaviour.
2633 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2636 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2637 indeed breaks things for older releases.
2639 3. Added additional logging to the case where there is a problem reading data
2640 from a filter that is running in a subprocess using a pipe, in order to
2641 try to track down a specific problem.
2643 4. Testing facility fudge: when running in the test harness and attempting
2644 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2645 getting "No route to host". Convert this to a timeout.
2647 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2650 6. Some OS don't have socklen_t but use size_t instead. This affects the
2651 fifth argument of getsockopt() amongst other things. This is now
2652 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2653 can be set for individual OS. I have set it for SunOS5, OSF1, and
2654 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2655 some earlier ones do not.
2657 7. Change 4.30/15 was not doing the test caselessly.
2659 8. The standard form for an IPv6 address literal was being rejected by address
2660 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2661 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2662 this, as well as the form without the "IPv6" on the front (but only when
2663 address literals are enabled, of course).
2665 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2667 10. Exim crashed if a message with an empty sender address specified by -f
2668 encountered a router with an errors_to setting. This could be provoked only
2669 by a command such as
2673 where an empty string was supplied; "<>" did not hit this bug.
2675 11. Installed PCRE release 4.5.
2677 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2678 remained set. It is now erased.
2680 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2681 times from message ids (which are base 36 rather than the normal 62).
2683 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2684 were being counted as actual protocol errors, and logged if the log
2685 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2686 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2687 and DATA following a set of rejected RCPTs do not count as protocol errors.
2688 In other words, Exim assumes they were pipelined, though this may not
2689 actually be the case. Of course, in all cases the client gets an
2690 appropriate error code.
2692 15. If a lookup fails in an ACL condition, a message about the failure may
2693 be available; it is used if testing the ACL cannot continue, because most
2694 such messages specify what the cause of the deferral is. However, some
2695 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2696 that caused an old message to be retained and used if a later statement
2697 caused a defer, replacing the real cause of the deferral.
2699 16. If an IP address had so many PTR records that the DNS lookup buffer
2700 was not large enough to hold them, Exim could crash while trying to process
2701 the truncated data. It now detects and logs this case.
2703 17. Further to 4.21/58, another change has been made: if (and only if) the
2704 first line of a message (the first header line) ends with CRLF, a bare LF
2705 in a subsequent header line has a space inserted after it, so as not to
2706 terminate the header.
2708 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2709 unnecessarily, used atoi() instead of strtol(), and didn't check the
2710 termination when getting file sizes from file names by regex.
2712 19. Completely re-implemented the support for maildirsize files, in the light
2713 of a number of problems with the previous contributed implementation
2714 (4.30/29). In particular:
2716 . If the quota is zero, the maildirsize file is maintained, but no quota is
2719 . If the maildir directory does not exist, it is created before any attempt
2720 to write a maildirsize file.
2722 . The quota value in the file is just a cache; if the quota is changed in
2723 the transport, the new value overrides.
2725 . A regular expression is available for excluding directories from the
2728 20. The autoreply transport checks the characters in options that define the
2729 message's headers; it allows continued headers, but it was checking with
2730 isspace() after an embedded newline instead of explicitly looking for a
2733 21. If all the "regular" hosts to which an address was routed had passed their
2734 expiry times, and had not reached their retry times, the address was
2735 bounced, even if fallback hosts were defined. Now Exim should go on to try
2738 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2739 equivalent code in the SMTP transport. Some hosts send humungous responses
2740 to HELO/EHLO, more than 1024 it seems.
2742 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2743 gives compiler warnings in some environments. I've now done it "properly",
2746 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2747 (because of problems with the built-in one) was declared to return uschar *
2748 instead of char *, causing compiler failure.
2750 25. Fixed a file descriptor leak when processing alias/forward files.
2752 26. Fixed a minor format string issue in dbfn.c.
2754 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2756 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2757 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2760 29. When a sender address is verified, it is cached, to save repeating the test
2761 when there is more than one recipient in a message. However, when the
2762 verification involves a callout, it is possible for different callout
2763 options to be set for different recipients. It is too complicated to keep
2764 track of this in the cache, so now Exim always runs a verification when a
2765 callout is required, relying on the callout cache for the optimization.
2766 The overhead is duplication of the address routing, but this should not be
2769 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2770 to be verified with callout=postmaster, and the main callout worked but the
2771 postmaster check failed, the verification correctly failed. However, if a
2772 subsequent RCPT command asked for sender verification *without* the
2773 postmaster check, incorrect caching caused this verification also to fail,
2776 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2777 it was not caching the DNS options (qualify_single, search_parents) that
2778 were used when the lookup failed. A subsequent lookup with different
2779 options therefore always gave the same answer, though there were cases
2780 where it should not have. (Example: a "domains = !$mx_any" option on a
2781 dnslookup router: the "domains" option is always processed without any
2782 widening, but the router might have qualify_single set.) Now Exim uses the
2783 cached value only when the same options are set.
2785 32. Added John Jetmore's "exipick" utility to the distribution.
2787 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2788 than a timeout (e.g. a certificate is required, and is not provided), an
2789 Exim server now closes the connection immediately. Previously it waited for
2790 the client to close - but if the client is SSL, it seems that they each
2791 wait for each other, leading to a delay before one of them times out.
2793 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2794 maintained 0.8.x compatibility because I don't think many are using it, and
2795 it is clearly obsolete.
2797 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2800 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2801 client certificate was expired. A simple patch fixes this, though I don't
2802 understand the full logic of why the verify callback is called multiple
2805 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2806 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2807 which causes problems with some clients (such as the Certicom SSL Plus
2808 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2809 disables the coutermeasure allowing Eudora to connect."
2811 38. Exim was not checking that a write() to a log file succeeded. This could
2812 lead to Bad Things if a log got too big, in particular if it hit a file
2813 size limit. Exim now panics and dies if it cannot write to a log file, just
2814 as it does if it cannot open a log file.
2816 39. Modified OS/Makefile-Linux so that it now contains
2818 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2820 The two -D definitions ensure that Exim is compiled with large file
2821 support, which makes it possible to handle log files that are bigger than
2824 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2825 instance) a domain was checked against a named list that involved a lookup,
2826 causing $domain_data to be set, then another domain was checked against the
2827 same list, then the first domain was re-checked, the value of $domain_data
2828 after the final check could be wrong. In particular, if the second check
2829 failed, it could be set empty. This bug probably also applied to
2832 41. The strip_trailing_dot option was not being applied to the address given
2833 with the -f command-line option.
2835 42. The code for reading a message's header from the spool was incrementing
2836 $received_count, but never initializing it. This meant that the value was
2837 incorrect (doubled) while delivering a message in the same process in which
2838 it was received. In the most common configuration of Exim, this never
2839 happens - a fresh exec is done - but it can happen when
2840 deliver_drop_privilege is set.
2842 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2843 it now includes up to 150 characters of the unexpected data in the log
2846 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2847 and building data strings. The size of both of these buffers was 10 000
2848 bytes - far larger than anybody would *ever* want, thought I. Needless to
2849 say, somebody hit the limit. I have increased the maximum line length to
2850 20 000 and the maximum data length of concatenated lines to 100 000. I have
2851 also fixed two bugs, because there was no checking on these buffers. Tsk,
2852 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2853 buffer is too small.
2855 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2856 lsearch lookups. Now it does.
2858 46. When parsing a route_list item in a manualroute router, a fixed-length
2859 buffer was used for the list of hosts. I made this 1024 bytes long,
2860 thinking that nobody would ever have a list of hosts that long. Wrong.
2861 Somebody had a whole pile of complicated expansion conditions, and the
2862 string was silently truncated, leading to an expansion error. It turns out
2863 that it is easier to change to an unlimited length (owing to other changes
2864 that have happened since this code was originally written) than to build
2865 structure for giving a limitation error. The length of the item that
2866 expands into the list of hosts is now unlimited.
2868 47. The lsearch lookup could not handle data where the length of text line was
2869 more than 4095 characters. Such lines were truncated, leading to shortened
2870 data being returned. It should now handle lines of any length.
2872 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2873 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2876 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2877 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2878 no longer happens in comment lines. A list of replacements is now placed
2879 at the head of all of the source files, except those whose only change is
2880 to replace PERL_COMMAND in the very first #! line.
2882 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2883 messages on the queue, with a bottom-up merge sort, using code contributed
2884 by Michael Haardt. This should make operations like -bp somewhat faster on
2885 large queues. It won't affect queue runners, except when queue_run_in_order
2888 51. Installed eximstats 1.31 in the distribution.
2890 52. Added support for SRV lookups to the dnslookup router.
2892 53. If an ACL referred to $message_body or $message_body_end, the value was not
2893 reset for any messages that followed in the same SMTP session.
2895 54. The store-handling optimization for building very long strings was not
2896 differentiating between the different store pools. I don't think this
2897 actually made any difference in practice, but I've tidied it.
2899 55. While running the routers to verify a sender address, $sender_address
2900 was still set to the sender address. This is wrong, because when routing to
2901 send a bounce to the sender, it would be empty. Therefore, I have changed
2902 it so that, while verifying a sender address, $sender_address is set to <>.
2903 (There is no change to what happens when verifying a recipient address.)
2905 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2906 target A or AAAA records (if not already returned) without resetting the
2907 qualify_single or search_parents options of the DNS resolver. These are
2908 inappropriate in this case because the targets of MX and SRV records must
2909 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2910 target that, when qualified, matched something in the local domain. These
2911 two options are now turned off when doing these lookups.
2913 57. It seems that at least some releases of Reiserfs (which does not have the
2914 concept of a fixed number of inodes) returns zero and not -1 for the
2915 number of available inodes. This interacted badly with check_spool_inodes,
2916 which assumed that -1 was the "no such thing" setting. What I have done is
2917 to check that the total number of inodes is greater than zero before doing
2918 the test of how many are available.
2920 58. When a "warn" ACL statement has a log_message modifier, the message is
2921 remembered, and not repeated. This is to avoid a lot of repetition when a
2922 message has many recipients that cause the same warning to be written.
2923 Howewer, Exim was preserving the list of already written lines for an
2924 entire SMTP session, which doesn't seem right. The memory is now reset if a
2925 new message is started.
2927 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2928 debugging output unless log_rewrite was also set.
2930 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2931 of (int)(handle) when we know that handle contains (void *)(-1).
2933 61. The Exim daemon panic-logs an error return when it closes the incoming
2934 connection. However "connection reset by peer" seems to be common, and
2935 isn't really an error worthy of noting specially, so that particular error
2938 62. When Exim is trying to find all the local interfaces, it used to panic and
2939 die if the ioctl to get the interface flags failed. However, it seems that
2940 on at least one OS (Solaris 9) it is possible to have an interface that is
2941 included in the list of interfaces, but for which you get a failure error
2942 for this call. This happens when the interface is not "plumbed" into a
2943 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2944 failure of the "get flags" call assumes that the interface is down.
2946 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2947 makes life easier for people who are doing arithmetic on fields extracted
2948 from dates, where you often get leading zeros that should not be
2949 interpreted as octal.
2951 64. Added qualify_domain to the redirect router, to override the global
2954 65. If a pathologically long header line contained very many addresses (the
2955 report of this problem mentioned 10 000) and each of them was rewritten,
2956 Exim could use up a very large amount of memory. (It kept on making new
2957 copies of the header line as it rewrote, and never released the old ones.)
2958 At the expense of a bit more processing, the header rewriting function has
2959 been changed so that it no longer eats memory in this way.
2961 66. The generation of the Received: header has been moved from the time that a
2962 message starts to be received, to the time that it finishes. The timestamp
2963 in the Received: header should now be very close to that of the <= log
2964 line. There are two side-effects of this change:
2966 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2967 logged header lines no longer include the local Received: line, because
2968 it has not yet been created. The same applies to a copy of the message
2969 that is returned to a non-SMTP sender when a message is rejected.
2971 (b) When a filter file is tested using -bf, no additional Received: header
2972 is added to the test message. After some thought, I decided that this
2975 This change does not affect the value of $received_for. It is still set
2976 after address rewriting, but before local_scan() is called.
2978 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2980 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2981 gave an unhelpful panic error message, and a defer error. I have managed to
2982 change this behaviour so that it now rejects any supplied certificate,
2983 which seems right, as the list of acceptable certificates is empty.
2985 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2986 gave an unhelpful defer error. I have not managed to make this reject any
2987 supplied certificates, but the error message it gives is "no certificate
2988 supplied", which is not helpful.
2990 70. exigrep's output now also includes lines that are not associated with any
2991 message, but which match the given pattern. Implemented by a patch from
2992 Martin Sluka, which also tidied up the Perl a bit.
2994 71. Recipient callout verification, like sender verification, was using <> in
2995 the MAIL FROM command. This isn't really the right thing, since the actual
2996 sender may affect whether the remote host accepts the recipient or not. I
2997 have changed it to use the actual sender in the callout; this means that
2998 the cache record is now keyed on a recipient/sender pair, not just the
2999 recipient address. There doesn't seem to be a real danger of callout loops,
3000 since a callout by the remote host to check the sender would use <>.
3001 [SEE ABOVE: changed after hitting problems.]
3003 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3004 temporary errors. However, in the case of such a code being given after
3005 the end of a data transmission (i.e. after ".") Exim was failing to write
3006 a retry record for the message. (Yes, there was some broken host that was
3007 actually sending 8xx at this point.)
3009 73. An unknown lookup type in a host list could cause Exim to panic-die when
3010 the list was checked. (An example that provoked this was putting <; in the
3011 middle of a list instead of at the start.) If this happened during a DATA
3012 ACL check, a -D file could be left lying around. This kind of configuration
3013 error no longer causes Exim to die; instead it causes a defer errror. The
3014 incident is still logged to the main and panic logs.
3016 74. Buglet left over from Exim 3 conversion. The message "too many messages
3017 in one connection" was written to the rejectlog but not the mainlog, except
3018 when address rewriting (yes!) was being logged.
3020 75. Added write_rejectlog option.
3022 76. When a system filter was run not as root (that is, when system_filter_user
3023 was set), the values of the $n variables were not being returned to the
3024 main process; thus, they were not subsequently available in the $sn
3027 77. Added +return_path_on_delivery log selector.
3029 78. A connection timeout was being treated differently from recipients deferred
3030 when testing hosts_max_try with a message that was older than the host's
3031 retry timeout. (The host should not be counted, thus allowing all hosts to
3032 be tried at least once before bouncing.) This may have been the cause of an
3033 occasionally reported bug whereby a message would remain on the queue
3034 longer than the retry timeout, but would be bounced if a delivery was
3035 forced. I say "may" because I never totally pinned down the problem;
3036 setting up timeout/retry tests is difficult. See also the next item.
3038 79. The ultimate address timeout was not being applied to errors that involved
3039 a combination of host plus message (for example, a timeout on a MAIL
3040 command). When an address resolved to a number of possible hosts, and they
3041 were not all tried for each delivery (e.g. because of hosts_max_try), a
3042 message could remain on the queue longer than the retry timeout.
3044 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3047 81. Fixed an obscure SMTP outgoing bug which required at least the following
3048 conditions: (a) there was another message waiting for the same server;
3049 (b) the server returned 5xx to all RCPT commands in the first message so
3050 that the message was not completed; (c) the server dropped the connection
3051 or gave a negative response to the RSET that Exim sends to abort the
3052 transaction. The observed case was a dropped connection after DATA that had
3053 been sent in pipelining mode. That is, the server had advertised PIPELINING
3054 but was not implementing it correctly. The effect of the bug was incorrect
3055 behaviour, such as trying another host, and this could lead to a crash.
3061 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3062 and daemon.c were passed as pointers to ints; they should have been
3063 pointers to socklen_t variables (which are typically unsigned ints).
3065 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3068 3. Fixed a really odd bug that affected only the testing scheme; patching a
3069 certain fixed string in the binary changed the value of another string that
3070 happened to be identical to the end of the original first string.
3072 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3073 name", it returns that address as the IP address. On some operating
3074 systems (e.g. Solaris), it also passes back the IP address string as the
3075 "host name". However, on others (e.g. Linux), it passes back an empty
3076 string. Exim wasn't checking for this, and was changing the host name to an
3077 empty string, assuming it had been canonicized.
3079 5. Although rare, it is permitted to have more than one PTR record for a given
3080 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3081 all the names associated with an address, because they do in Solaris.
3082 However, it seems that they do not in Linux for data that comes from the
3083 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3084 I found this out when I moved to a new Linux workstation and tried to run
3085 the Exim test suite.
3087 To get round this problem I have changed the code so that it now does its
3088 own call to the DNS to look up PTR records when searching for a host name.
3089 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3090 addresses that are only in /etc/hosts are still found.
3092 This behaviour is, however, controlled by an option called host_lookup_
3093 order, which defaults to "bydns:byaddr". If people want to use the other
3094 order, or indeed, just use one or the other means of lookup, they can
3095 specify it in this variable.
3097 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3098 some operating systems, this comes back from gethostbyaddr() as an empty
3099 string, and this is what Exim used to test for. However, it seems that in
3100 other systems, "." is yielded. Exim now tests for this case too.
3102 7. The values of check_spool_space and check_log_space are now held internally
3103 as a number of kilobytes instead of an absolute number of bytes. If a
3104 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3105 kilobyte. This means that much larger values can be stored.
3107 8. Exim monitor: an attempt to get the action menu when not actually pointing
3108 at a message produces an empty menu entitled "No message selected". This
3109 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3110 no entries in it ("Shell widget menu has zero width and/or height"). So I
3111 have added a single, blank menu entry in this case.
3113 9. Added ${quote_local_part.
3115 10. MIME decoding is now applied to the contents of Subject: header lines when
3118 11. Now that a reference to $sender_host_address automatically causes a reverse
3119 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3120 host lookup before query-style lookups in lists that might use this
3121 variable. This has therefore been abolished, and the "net-" prefix is no
3122 longer necessary for query-style lookups.
3124 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3125 have been a typo for LFLAGS, so it has been changed.
3127 13. The install script calls Exim with "-C /dev/null" in order to find the
3128 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3129 to be output. Howeve, since Exim outputs its version number before the
3130 error, it didn't break the script. It just looked ugly. I fixed this by
3131 always allowing "-C /dev/null" if the caller is root.
3133 14. Ignore overlarge ACL variable number when reading spool file - insurance
3134 against a later release with more variables having written the file.
3136 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3137 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3138 this, as well as the form without the "IPv6" on the front.
3140 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3141 OS/Makefile-Darwin file.
3143 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3144 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3146 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3147 variable. However, some OS use ints for inodes. I've added cast to long int
3148 to get rid of the compiler warning.
3150 19. I had forgotten to lock out "/../" in configuration file names when
3151 ALT_CONFIG_PREFIX was set.
3153 20. Routers used for verification do not need to specify transports. However,
3154 if such a router generated a host list, and callout was configured, Exim
3155 crashed, because it could not find a port number from the (non-existent)
3156 transport. It now assumes port 25 in this circumstance.
3158 21. Added the -t option to exigrep.
3160 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3161 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3162 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3163 them is set, LOOKUP_LSEARCH is forced.
3165 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3167 24. Added sender and host information to the "rejected by local_scan()" log
3168 line; previously there was no indication of these.
3170 25. Added .include_if_exists.
3172 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3173 Linux. It turns out that not all file systems support this. Apparently some
3174 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3175 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3176 not supported on the file descriptor, is now ignored when Exim is trying to
3177 sync a directory. This applies only to Linux.
3179 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3181 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3182 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3185 29. Added support for maildirsize files from supplied patch (modified a bit).
3187 30. The use of :fail: followed by an empty string could lead Exim to respond to
3188 sender verification failures with (e.g.):
3190 550 Verification failed for <xxx>
3191 550 Sender verify failed
3193 where the first response line was missing the '-' that indicates it is not
3194 the final line of the response.
3196 31. The loop for finding the name of the user that called Exim had a hardwired
3197 limit of 10; it now uses the value of finduser_retries, which is used for
3198 all other user lookups.
3200 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3203 33. Exim was neglecting to zero errno before one call of strtol() when
3204 expanding a string and expecting an integer value. On some systems this
3205 resulted in spurious "integer overflow" errors. Also, it was casting the
3206 result into an int without checking.
3208 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3209 did not work. The code looks as if it has *never* worked, though it appears
3210 to have been documented since at least releast 1.62. I have made it work.
3212 35. The "timeout_DNS" error in retry rules, also documented since at least
3213 1.62, also never worked. As it isn't clear exactly what this means, and
3214 clearly it isn't a major issue, I have abolished the feature by treating it
3215 as "timeout", and writing a warning to the main and panic logs.
3217 36. The display of retry rules for -brt wasn't always showing the error code
3220 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3221 timeout_connect_A, timeout_connect_MX.
3223 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3224 to the empty sender.
3226 39. The daemon was not analysing the content of -oX till after it had closed
3227 stderr and disconnected from the controlling terminal. This meant that any
3228 syntax errors were only noted on the panic log, and the return code from
3229 the command was 0. By re-arranging the code a little, I've made the
3230 decoding happen first, so such errors now appear on stderr, and the return
3231 code is 1. However, the actual setting up of the sockets still happens in
3232 the disconnected process, so errors there are still only recorded on the
3235 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3236 connections (as happens on some IP stacks) was logged at start up time as
3237 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3238 it from "IPv6 and IPv4", which means that two separate sockets are being
3241 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3242 says whether AF_INET or AF_INET6 was passed as an argument.
3244 42. Exiwhat output was messed up when time zones were included in log
3247 43. Exiwhat now gives more information about the daemon's listening ports,
3248 and whether -tls-on-connect was used.
3250 44. The "port" option of the smtp transport is now expanded.
3252 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3253 silently ignored. Now an error message is written to the main and panic
3256 46. There's a new ACL modifier called "logwrite" which writes to a log file
3257 as soon as it is encountered.
3259 47. Added $local_user_uid and $local_user_gid at routing time.
3261 48. Exim crashed when trying to verify a sender address that was being
3264 49. Exim was recognizing only a space character after ".include". It now also
3265 recognizes a tab character.
3267 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3268 extracting the relevant information from the specification. The man page no
3269 longer contains scrambled data for the -d option, and I've added a section
3270 at the front about calling Exim under different names.
3272 51. Added "extra_headers" argument to the "mail" command in filter files.
3274 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3277 53. Installed eximstats 1.29.
3279 54. Added transport_filter_timeout as a generic transport option.
3281 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3282 Cc: header lines. This was required by RFC 822, but it not required by RFC
3285 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3286 incoming messages that did not have them. Now it does so only if the
3287 message originates locally, that is, if there is no associated remote host
3288 address. When Resent- header lines are present, this applies to the Resent-
3289 lines rather than the non-Resent- lines.
3291 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3292 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3294 58. Messages for configuration errors now include the name of the main
3295 configuration file - useful now that there may be more than one file in a
3296 list (.included file names were always shown).
3298 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3299 for those rare installations that do not start the daemon as root or run it
3300 setuid root. I've cut out the call to initgroups() if the daemon is not
3303 60. The Exim user and group can now be bound into the binary as text strings
3304 that are looked up at the start of Exim's processing.
3306 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3308 62. Added $mailstore_basename variable.
3310 63. Installed patch to sieve.c from Michael Haardt.
3312 64. When Exim failed to open the panic log after failing to open the main log,
3313 the original message it was trying to log was written to stderr and debug
3314 output, but if they were not available (the usual case in production), it
3315 was lost. Now it is written to syslog before the two lines that record the
3316 failures to open the logs.
3318 65. Users' Exim filters run in subprocesses under the user's uid. It is
3319 possible for a "deliver" command or an alias in a "personal" command to
3320 provoke an address rewrite. If logging of address rewriting is configured,
3321 this fails because the process is not running as root or exim. There may be
3322 a better way of dealing with this, but for the moment (because 4.30 needs
3323 to be released), I have disabled address rewrite logging when running a
3324 filter in a non-root, non-exim process.
3330 1. The buildconfig auxiliary program wasn't quoting the value set for
3331 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3332 not defined. This bug was masked in 4.22 by the effect that was fixed in
3335 2. Some messages that were rejected after a message id was allocated were
3336 shown as "incomplete" by exigrep. It no longer does this for messages that
3337 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3339 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3340 have allow_domain_literals set, the ID did not get logged in the <= line.
3341 Domain literals are now always recognized in Message-ID: header lines.
3343 4. The first argument for a ${extract expansion item is the key name or field
3344 number. Leading and trailing spaces in this item were not being ignored,
3345 causing some misleading effects.
3347 5. When deliver_drop_privilege was set, single queue runner processes started
3348 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3349 same command in the process it spins off) were not dropping privilege.
3351 6. When the daemon running as "exim" started a queue runner, it always
3352 re-executed Exim in the spun-off process. This is a waste of effort when
3353 deliver_drop_privilege is set. The new process now just calls the
3354 queue-runner function directly.
3360 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3363 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3364 ignored. Though the use of -oP was forcing the writing of a pid file, it
3365 was always written to the default place.
3367 3. If the message "no IP address found for host xxxx" is generated during
3368 incoming verification, it is now followed by identification of the incoming
3369 connection (so you can more easily find what provoked it).
3371 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3373 5. Added some features to "harden" Exim a bit more against certain attacks:
3375 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3376 be put in Local/Makefile. This is like the never_users runtime option,
3377 but it cannot be overridden. The default setting is "root".
3379 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3380 prefix string with which any file named in a -C command line option
3383 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3384 is retained for -C and -D only if the caller of Exim is root. Without
3385 it, the exim user may also use -C and -D and retain privilege.
3387 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3388 command line option is disabled.
3390 6. Macro names set by the -D option must start with an upper case letter, just
3391 like macro names defined in the configuration file.
3393 7. Added "dereference=" facility to LDAP.
3395 8. Two instances of the typo "uknown" in the source files are fixed.
3397 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3398 the Configure-Makefile script screwed up while processing it.
3400 10. Incorporated PCRE 4.4.
3402 11. The SMTP synchronization check was not operating right at the start of an
3403 SMTP session. For example, it could not catch a HELO sent before the client
3404 waited for the greeting. There is now a check for outstanding input at the
3405 point when the greeting is written. Because of the duplex, asynchronous
3406 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3407 way, but not yet received, when the check is performed.
3409 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3410 on TCP/IP sockets, because this apparently causes some broken clients to
3413 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3414 unchanged) from the Cygwin maintainer.
3416 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3417 was supported instead of testing for mbx. Effectively a typo.
3419 15. The spa authenticator server code was not checking that the input it
3420 received was valid base64.
3422 16. The debug output line for the "set" modifier in ACLs was not showing the
3423 name of the variable that was being set.
3425 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3427 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3428 lookup on its own. Something else was needed to trigger the lookup. For
3429 example, a match in host_lookup or the need for a host name in a host list.
3430 Now, if $sender_host_name is referenced and the host name has not yet been
3431 looked up, a lookup is performed. If the lookup fails, the variable remains
3432 empty, and $host_lookup_failed is set to "1".
3434 19. Added "eqi" as a case-independent comparison operator.
3436 20. The saslauthd authentication condition could segfault if neither service
3437 nor realm was specified.
3439 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3440 error message that was logged was misleading, and incoming SMTP
3441 connections were dropped. The message is now more accurate, and temporary
3442 errors are given to SMTP connections.
3444 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3445 (except RSET) with a 421 error, until QUIT is received. However, it was
3446 failing to send a response to QUIT.
3448 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3449 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3450 ACL is now run after the helo_try_verify_hosts code.
3452 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3453 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3454 case-independent, so other case variants are also recognized). Apparently
3455 some systems use these upper case variants.
3457 25. If more than two messages were waiting for the same host, and a transport
3458 filter was specified for the transport, Exim sent two messages over the
3459 same TCP/IP connection, and then failed with "socket operation on non-
3460 socket" when it tried to send the third.
3462 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3464 27. The extern definition of crypt16() in expand.c was not being excluded when
3465 the OS had its own crypt16() function.
3467 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3468 as a preferred synonym for return_size_limit, both as an option and as an
3471 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3473 30. Changed the default configuration ACL to relax the local part checking rule
3474 for addresses that are not in any local domains. For these addresses,
3475 slashes and pipe symbols are allowed within local parts, but the sequence
3476 /../ is explicitly forbidden.
3478 31. SPA server authentication was not clearing the challenge buffer before
3481 32. log_message in a "warn" ACL statement was writing to the reject log as
3482 well as to the main log, which contradicts the documentation and doesn't
3483 seem right (because no rejection is happening). So I have stopped it.
3485 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3486 However, I am unable to do any testing of this.
3488 34. Fixed an infelicity in the appendfile transport. When checking directories
3489 for a mailbox, to see if any needed to be created, it was accidentally
3490 using path names with one or more superfluous leading slashes; tracing
3491 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3493 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3494 contents are added to the log line that is written for every discarded
3495 recipient. (Previously a log_message setting was ignored.)
3497 36. The ${quote: operator now quotes the string if it is empty.
3499 37. The install script runs exim in order to find its version number. If for
3500 some reason other than non-existence or emptiness, which it checks, it
3501 could not run './exim', it was installing it with an empty version number,
3502 i.e. as "exim-". This error state is now caught, and the installation is
3505 38. An argument was missing from the function that creates an error message
3506 when Exim fails to connect to the socket for saslauthd authentication.
3507 This could cause Exim to crash, or give a corrupted message.
3509 39. Added isip, isip4, and isip6 to ${if conditions.
3511 40. The ACL variables $acl_xx are now saved with the message, and can be
3512 accessed later in routers, transports, and filters.
3514 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3515 strings in the file are not string-expanded.
3517 42. If a MAIL command specified a SIZE value that was too large to fit into an
3518 int variable, the check against message_size_limit failed. Such values are
3519 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3520 day this will have to be increased, but I don't think I want to be around
3521 when emails are that large.
3528 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3529 iconv() is not standard in FreeBSD.
3531 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3532 IPv6 enabled. The observed symptom was a segmentation fault on return from
3533 the function os_common_find_running_interfaces() in src/os.c.
3535 3. In the check_special_case() function in daemon.c I had used "errno" as an
3536 argument name, which causes warnings on some systems. This was basically a
3537 typo, since it was named "eno" in the comments!
3539 4. The code that waits for the clock to tick (at a resolution of some fraction
3540 of a second) so as to ensure message-id uniqueness was always waiting for
3541 at least one whole tick, when it could have waited for less. [This is
3542 almost certainly not relevant at current processor speeds, where it is
3543 unlikely to ever wait at all. But we try to future-proof.]
3545 5. The function that sleeps for a time interval that includes fractions of a
3546 second contained a race. It did not block SIGALRM between setting the
3547 timer, and suspending (a couple of lines later). If the interval was short
3548 and the sigsuspend() was delayed until after it had expired, the suspension
3549 never ended. On busy systems this could lead to processes getting stuck for
3552 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3553 process, before it forks any delivery processes. The open lookup caching
3554 mechanism meant that the open file or database connection was passed into
3555 the delivery process. The problem was that delivery processes always tidy
3556 up cached lookup data. This could cause a problem for the next delivery
3557 process started by the queue runner, because the external queue runner
3558 process does not know about the closure. So the next delivery process
3559 still has data in the lookup cache. In the case of a file lookup, there was
3560 no problem because closing a file descriptor in a subprocess doesn't affect
3561 the parent. However, if the lookup was caching a connection to a database,
3562 the connection was closed, and the second delivery process was likely to
3563 see errors such as "PGSQL: query failed: server closed the connection
3564 unexpectedly". The problem has been fixed by closing all cached lookups
3565 in a queue runner before running a delivery process.
3567 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3568 seem to have the "const" qualifier which it has on other OS. I've
3571 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3572 *of the same type* (client or server) with the same public name that an
3573 error should be diagnosed.
3575 9. When Exim looked up a host name for an IP address, but failed to find the
3576 original IP address when looking up the host name (a safety check), it
3577 output the message "<ip address> does not match any IP for NULL", which was
3578 confusing, to say the least. The bug was that the host name should have
3579 appeared instead of "NULL".
3581 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3582 user that is built into the binary, and the -C or -D options is used, root
3583 privilege is dropped before the configuration file is read. In addition,
3584 logging is switched to stderr instead of the normal log files. If the
3585 configuration then re-defines the Exim user, the unprivileged environment
3586 is probably not what is expected, so Exim logs a panic warning message (but
3589 However, if deliver_drop_privilege is set, the unprivileged state may well
3590 be exactly what is intended, so the warning has been cut out in that case,
3591 and Exim is allowed to try to write to its normal log files.
3597 1. smtp_return_error_details was not giving details for temporary sender
3598 or receiver verification errors.
3600 2. Diagnose a configuration error if two authenticators have the same public
3603 3. Exim used not to create the message log file for a message until the first
3604 delivery attempt. This could be confusing when incoming messages were held
3605 for policy or load reasons. The message log file is now created at the time
3606 the message is received, and an initial "Received" line is written to it.
3608 4. The automatically generated man page for command line options had a minor
3609 bug that caused no ill effects; however, a more serious problem was that
3610 the procedure for building the man page automatically didn't always
3611 operate. Consequently, release 4.20 contains an out-of-date version. This
3612 shouldn't happen again.
3614 5. When building Exim with embedded Perl support, the script that builds the
3615 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3616 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3618 6. The freeze_tell option was not being used for messages that were frozen on
3619 arrival, either by an ACL or by local_scan().
3621 7. Added the smtp_incomplete_transaction log selector.
3623 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3624 was accepting AUTH without a new EHLO.
3626 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3627 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3630 10. Logging of TCP/IP connections (when configured) now happens in the main
3631 daemon process instead of the child process, so that the TCP/IP connection
3632 count is more accurate (but it can never be perfect).
3634 11. The use of "drop" in a nested ACL was not being handled correctly in the
3635 outer ACL. Now, if condition failure induced by the nested "drop" causes
3636 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3637 or "require"), the connection is dropped.
3639 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3640 that yield "discard" can now be used with an "accept" or a "discard" verb,
3641 but an error is generated for any others (because I can't see a useful way
3642 to define what should happen).
3644 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3645 are now processed in the same way as lines in the Exim configuration file.
3646 In particular, continuation lines are supported.
3648 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3650 15. Added -ti meaning -t -i.
3652 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3653 domains, and warn by logging if others are found.
3655 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3656 returned from GIFCONF when Exim is trying to find the list of interfaces on
3657 a host. The code in os.c has been modified to copy each ifreq to an aligned
3658 structure in all cases.
3660 Also, in some cases, the returned ifreq's were being copied to a 'struct
3661 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3662 means the last couple of bytes of an IPv6 address could be chopped if the
3663 ifreq contained only a normal sockaddr (14 bytes storage).
3665 18. Named domain lists were not supported in the hosts_treat_as_local option.
3666 An entry such as +xxxx was not recognized, and was treated as a literal
3669 19. Ensure that header lines added by a DATA ACL are included in the reject log
3670 if the ACL subsequently rejects the message.
3672 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3673 MD5 (which is deprecated).
3675 21. When testing a filter file using -bf, Exim was writing a message when it
3676 took the sender from a "From " line in the message, but it was not doing so
3677 when it took $return_path from a Return-Path: header line. It now does.
3679 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3680 with a valid header line field name (a series of printing characters
3681 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3683 23. Changed "disc" in the source to "disk" to conform to the documentation and
3684 the book and for uniformity.
3686 24. Ignore Sendmail's -Ooption=value command line item.
3688 25. When execve() failed while trying to run a command in a pipe transport,
3689 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3690 could be confused with a return value of 69 from the command itself. This
3691 has been changed to 127, the value the shell returns if it is asked to run
3692 a non-existent command. The wording for the related log line suggests a
3693 non-existent command as the problem.
3695 26. If received_header_text expands to an empty string, do not add a Received:
3696 header line to the message. (Well, it adds a token one on the spool, but
3697 marks it "old" so that it doesn't get used or transmitted.)
3699 27. Installed eximstats 1.28 (addition of -nt option).
3701 28. There was no check for failure on the call to getsockname() in the daemon
3702 code. This can fail if there is a shortage of resources on the system, with
3703 ENOMEM, for example. A temporary error is now given on failure.
3705 29. Contrary to the C standard, it seems that in some environments, the
3706 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3707 program. Exim now does this explicitly; it affects the formatting of
3708 timestamps using strftime().
3710 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3711 complaints. I've now initialized all the variables, to avoid this.
3713 32. Header lines added by a system filter were not being "seen" during
3714 transport-time rewrites.
3716 33. The info_callback() function passed to OpenSSL is set up with type void
3717 (*)(SSL *, int, int), as described somewhere. However, when calling the
3718 function (actually a macro) that sets it up, the type void(*)() is
3719 expected. I've put in a cast to prevent warnings from picky compilers.
3721 34. If a DNS black list lookup found a CNAME record, but there were no A
3722 records associated with the domain it pointed at, Exim crashed.
3724 35. If a DNS black list lookup returned more than one A record, Exim ignored
3725 all but the first. It now scans all returned addresses if a particular IP
3726 value is being sought. In this situation, the contents of the
3727 $dnslist_value variable are a list of all the addresses, separated by a
3730 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3731 to accept a host name and all its aliases if the forward lookup for any of
3732 them yielded the IP address of the incoming connection. Now it accepts only
3733 those names whose forward lookup yields the correct IP address. Any other
3734 names are discarded. This closes a loophole whereby a rogue DNS
3735 administrator could create reverse DNS records to break through a
3736 wildcarded host restriction in an ACL.
3738 37. If a user filter or a system filter that ran in a subprocess used any of
3739 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3740 the wrong values were passed to the pipe command ($thisaddress had the
3741 value of $0, $0 had the value of $1, etc). This bug was introduced by
3742 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3744 38. Improved the line breaking for long SMTP error messages from ACLs.
3745 Previously, if there was no break point between 40 and 75 characters, Exim
3746 left the rest of the message alone. Two changes have been made: (a) I've
3747 reduced the minimum length to 35 characters; (b) if it can't find a break
3748 point between 35 and 75 characters, it looks ahead and uses the first one
3749 that it finds. This may give the occasional overlong line, but at least the
3750 remaining text gets split now.
3752 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3753 file descriptors might be low, and that setting 1000 would always raise it.
3754 It turns out that in some environments, the limit is already over 1000 and
3755 that lowering it causes trouble. So now Exim takes care not to decrease it.
3757 40. When delivering a message, the value of $return_path is set to $sender_
3758 address at the start of routing (routers may change the value). By an
3759 oversight, this default was not being set up when an address was tested by
3760 -bt or -bv, which affected the outcome if any router or filter referred to
3763 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3764 log only when "message" or "log_message" are set. However, if one of the
3765 conditions was an address verification, or a call to a nested ACL, the
3766 messages generated by the underlying test were being passed through. This
3767 no longer happens. The underlying message is available in $acl_verify_
3768 message for both "message" and "log_message" expansions, so it can be
3769 passed through if needed.
3771 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3772 new expansion $bh_ to give the encoded byte string without charset
3773 translation. Translation happens only if iconv() is available; HAVE_ICONV
3774 indicates this at build time. HEADERS_CHARSET gives the charset to
3775 translate to; headers_charset can change it in the configuration, and
3776 "headers charset" can change it in an individual filter file.
3778 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3779 that creates RFC 2047 encoded "words" labels them as that charset instead
3780 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3781 expansion operator; (ii) when Exim creates a From: line for a local
3782 message; (iii) when a header line is rewritten to include a "phrase" part.
3784 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3785 buggy, causing it to skip the first lines of messages whose message ID
3786 ended in 'D'. This would not have bitten before Exim release 4.14, because
3787 message IDs were unlikely to end in 'D' before then. The effect was to have
3788 incorrect size information for certain domains.
3790 45. #include "config.h" was missing at the start of the crypt16.c module. This
3791 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3794 46. If there was a timeout during a "random" callout check, Exim treated it as
3795 a failure of the random address, and carried on sending RSET and the real
3796 address. If the delay was just some slowness somewhere, the response to the
3797 original RCPT would be taken as a response to RSET and so on, causing
3798 mayhem of various kinds.
3800 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3801 when I implemented it. It didn't allow for the fact that some option values
3802 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3803 the right test for positive values.
3805 48. Domain names in DNS records are case-independent. Exim always looks them up
3806 in lower case. Some resolvers return domain names in exactly the case they
3807 appear in the zone file, that is, they may contain uppercase letters. Not
3808 all resolvers do this - some return always lower case. Exim was treating a
3809 change of case by a resolver as a change of domain, similar to a widening
3810 of a domain abbreviation. This triggered its re-routing code and so it was
3811 trying to route what was effectively the same domain again. This normally
3812 caused routing to fail (because the router wouldn't handle the domain
3813 twice). Now Exim checks for this case specially, and just changes the
3814 casing of the domain that it ultimately uses when it transmits the message
3817 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3820 50. If a filter generated a file delivery with a non-absolute name (possible if
3821 no home directory exists for the router), the forbid_file option was not
3824 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3825 the existing equality matching.
3827 52. Exim was using ints instead of ino_t variables in some places where it was
3828 dealing with inode numbers.
3830 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3831 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3832 called TMPDIR, and if it finds it is different, it changes its value.
3834 54. The smtp_printf() function is now made available to local_scan() so
3835 additional output lines can be written before returning. There is also an
3836 smtp_fflush() function to enable the detection of a dropped connection.
3837 The variables smtp_input and smtp_batched_input are exported to
3840 55. Changed the default runtime configuration: the message "Unknown user"
3841 has been removed from the ACL, and instead placed on the localuser router,
3842 using the cannot_route_message feature. This means that any verification
3843 failures that generate their own messages won't get overridden. Similarly,
3844 the "Unrouteable address" message that was in the ACL for unverifiable
3845 relay addresses has also been removed.
3847 56. Added hosts_avoid_esmtp to the smtp transport.
3849 57. The exicyclog script was not checking for the esoteric option
3850 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3851 will work only if exicyclog is run under the appropriate euid.
3853 58. Following a discussion on the list, the rules by which Exim recognises line
3854 endings on incoming messages have been changed. The -dropcr and drop_cr
3855 options are now no-ops, retained only for backwards compatibility. The
3856 following line terminators are recognized: LF CRLF CR. However, special
3857 processing applies to CR:
3859 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3860 nor a local message in the state where . is a terminator.
3862 (ii) If a bare CR is encountered in a header line, an extra space is added
3863 after the line terminator so as not to end the header. The reasoning
3864 behind this is that bare CRs in header lines are most likely either
3865 to be mistakes, or people trying to play silly games.
3867 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3868 was being incorrectly given as 18 bytes larger than it should have been.
3869 This is a VOB (very old bug).
3871 60. This may never have affected anything current, but just in case it has:
3872 When the local host is found other than at the start of a list of hosts,
3873 the local host, those with the same MX, and any that follow, are discarded.
3874 When the list in question was part of a longer list of hosts, the following
3875 hosts (not currently being processed) were also being discarded. This no
3876 longer happens. I'm not sure if this situation could ever has previously
3879 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3882 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3883 header. This is something that is recommended in a new Internet Draft, and
3884 is something that is documented as being done by Sendmail. There are two
3885 possible values. For messages generated by the autoreply transport, Exim
3888 Auto-Submitted: auto-replied
3890 whereas for all other generated messages (e.g. bounces) it adds
3892 Auto-Submitted: auto-generated
3894 63. The "personal" condition in filters now includes a test for the
3895 Auto-Submitted: header. If it contains the string "auto-" the message it
3896 not considered personal.
3898 64. Added rcpt_include_affixes as a generic transport option.
3900 65. Added queue_only_override (default true).
3902 66. Added the syslog_duplication option.
3904 67. If what should have been the first header line of a message consisted of
3905 a space followed by a colon, Exim was mis-interpreting it as a header line.
3906 It isn't of course - it is syntactically invalid and should therefore be
3907 treated as the start of the message body. The misbehaviour could have
3908 caused a number of strange effects, including loss of data in subsequent
3909 header lines, and spool format errors.
3911 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3912 client host had authenticated. This control can now be exercised by an ACL
3913 for more flexibility.
3915 69. By default, callouts do not happen when testing with -bh. There is now a
3916 variant, -bhc, which does actually run the callout code, including
3917 consulting and updating the callout cache.
3919 70. Added support for saslauthd authentication, courtesy of Alexander
3922 71. If statvfs() failed on the spool or log directories while checking their
3923 size for availability, Exim confusingly gave the error "space shortage".
3924 Furthermore, in debugging mode it crashed with a floating point exception.
3925 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3926 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3927 serious problem, Exim now writes to the main and panic logs when this
3928 happens, with details of the failure. It then refuses to accept the
3929 incoming message, giving the message "spool directory problem" or "log
3930 directory problem" with a 421 code for SMTP messages.
3932 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3933 0, 1, and 2 exist, because some OS complain for execs without them (see
3934 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3935 descriptors. However, the code omitted to check that the open succeeded,
3936 causing mysterious errors if for some reason the permissions on /dev/null
3937 got screwed. Now Exim writes a message to the main and panic logs, and
3938 bombs out if it can't open /dev/null.
3940 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3941 interact so that it is all more flexible. It is supposed to remain
3942 backwards compatible. Also added extra_local_interfaces.
3944 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3945 to bomb out with an assertion failure - to the client this appears as a
3946 connection drop. This problem occurs in the part of the code that was taken
3947 from the Samba project. Fortunately, the assertion is in a very simple
3948 function, so I have fixed this by reproducing the function inline in the
3949 one place where it is called, and arranging for authentication to fail
3950 instead of killing the process with assert().
3952 75. The SPA client code was not working when the server requested OEM rather
3953 than Unicode encoding.
3955 76. Added code to make require_files with a specific uid setting more usable in
3956 the case where statting the file as root fails - usually a non-root-mounted
3957 NFS file system. When this happens and the failure is EACCES, Exim now
3958 forks a subprocess and does the per-uid checking as the relevant uid.
3960 77. Added process_log_path.
3962 78. If log_file_path was not explicitly set, a setting of check_log_space or
3963 check_log_inodes was ignored.
3965 79. If a space check for the spool or log partitions fails, the incident is now
3966 logged. Of course, in the latter case the data may get lost...
3968 80. Added the %p formatting code to string_format() so that it can be used to
3969 print addresses in debug_print(). Adjusted all the address printing in the
3970 debugging in store.c to use %p rather than %d.
3972 81. There was a concern that a line of code in smtp_in.c could overflow a
3973 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3974 initially expressed, the concern was not well-founded, because trailing
3975 spaces are removed early. However, if the trailing spaces were followed by
3976 a NULL, they did not get removed, so the overflow was possible. Two fixes
3979 (a) I re-wrote the offending code in a cleaner fashion.
3980 (b) If an incoming SMTP command contains a NULL character, it is rejected
3983 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3984 runs initgroups(), so that if the Exim user is in any additional groups,
3985 they will be used during message reception.
3991 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob