1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.364 2006/06/30 13:57:46 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 SC/01 Use a glob alias rather than an array ref in eximstats generated
10 parser. This improves both readability and performance.
12 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
13 Don't display local sender or destination tables in eximstats unless
14 there is data to show.
15 Added average volumes into the eximstats top table text output.
17 SC/03 Collect data on the number of addresses (recipients) as well
18 as the number of messages in eximstats.
20 TF/01 Correct an error in the documentation for the redirect router. Exim
21 does (usually) call initgroups() when daemonizing.
23 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
24 with consistent privilege compared to when running as a daemon.
26 TF/03 Note in the spec that $authenticated_id is not set for local
27 submissions from trusted users.
29 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
30 Thanks to Dean Brooks <dean@iglou.com> for the patch.
32 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
33 by adding some example configuration directives to the default
34 configuration file. A little bit of work is required to uncomment the
35 directives and define how usernames and passwords are checked, but
36 there is now a framework to start from.
38 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
39 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
40 without this. I don't know how relevant this is to other LDAP libraries.
42 PH/02 Add the verb name to the "unknown ACL verb" error.
44 PH/03 Magnus Holmgren's patch for filter_prepend_home.
46 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
48 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
49 directory not expanded when it should be if an expanded home directory
50 was set for the address (which is overridden by the transport).
52 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
55 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
56 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
57 because it is too late at that time, and has no effect.
59 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
60 security issue with \' (bugzilla #107). I could not use the
61 PQescapeStringConn() function, because it needs a PGconn value as one of
68 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
69 other effects) broke the use of negated acl sub-conditions.
71 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
74 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
75 "Deny" causes Exim to reject the incoming connection with a 554 error.
76 Unfortunately, if there is a major crisis, such as a disk failure,
77 tcp-wrappers gives "deny", whereas what one would like would be some
78 kind of temporary error. A kludge has been added to help with this.
79 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
80 554 error is used if errno is still zero or contains ENOENT (which occurs
81 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
84 PH/03 Add -lutil to the default FreeBSD LIBS setting.
86 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
87 errors. Otherwise a message that provokes a temporary error (when other
88 messages do not) can cause a whole host to time out.
90 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
91 addresses were routed directly to files or pipes from a redirect router.
92 File deliveries just didn't batch; pipe deliveries might have suffered
95 PH/06 A failure to get a lock for a hints database would erroneously always say
96 "Failed to get write lock", even when it was really a read lock.
98 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
99 of 0600. This has been changed to use the value of the lockfile_mode
100 option (which defaults to 0600).
102 PH/08 Applied small patch from the Sieve maintainer.
104 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
105 folder from quota calculations, a direct delivery into this folder messed
106 up the contents of the maildirsize file. This was because the regex was
107 used only to exclude .Trash (or whatever) when the size of the mailbox
108 was calculated. There was no check that a delivery was happening into an
109 excluded directory. This bug has been fixed by ignoring all quota
110 processing for deliveries into excluded directories.
112 PH/10 Added the maildirfolder_create_regex option to appendfile.
118 PH/01 The code for finding all the local interface addresses on a FreeBSD
119 system running IPv6 was broken. This may well have applied to all BSD
120 systems, as well as to others that have similar system calls. The broken
121 code found IPv4 interfaces correctly, but gave incorrect values for the
122 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
123 that it would not match correctly against @[] and not recognize the IPv6
126 PH/02 The ipliteral router was not recognizing addresses of the form user@
127 [ipv6:....] because it didn't know about the "ipv6:" prefix.
129 PH/03 Added disable_ipv6.
131 PH/04 Changed $reply_address to use the raw form of the headers instead of the
132 decoded form, because it is most often used to construct To: headers
133 lines in autoreplies, and the decoded form may well be syntactically
134 invalid. However, $reply_address has leading white space removed, and all
135 newlines turned into spaces so that the autoreply transport does not
138 PH/05 If group was specified without a user on a router, and no group or user
139 was specified on a transport, the group from the router was ignored.
141 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
142 for visible compile-time settings that can be used to change these
143 numbers, for those that want even more. Backwards compatibility with old
144 spool files has been maintained. However, going back to a previous Exim
145 release will lost any variables that are in spool files.
147 PH/07 Two small changes when running in the test harness: increase delay when
148 passing a TCP/IP connection to a new process, in case the original
149 process has to generate a bounce, and remove special handling of
150 127.0.0.2 (sic), which is no longer necessary.
152 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
153 be the same on different OS.
155 PH/09 Moved a debug statement in filter processing to avoid a race problem when
158 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
159 whether --show-vars was specified or not
161 JJ/02 exipick: Added support for new ACL variable spool format introduced
164 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
165 syntactically invalid From: or Reply-to: line, and a filter used this to
166 generate an autoreply, and therefore failed to obtain an address for the
167 autoreply, Exim could try to deliver to a non-existent relative file
168 name, causing unrelated and misleading errors. What now happens is that
169 it logs this as a hard delivery error, but does not attempt to create a
172 PH/11 The exinext utility has a -C option for testing purposes, but although
173 the given file was scanned by exinext itself; it wasn't being passed on
174 when Exim was called.
176 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
177 an end-of-file indication when reading a command response.
179 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
180 compiled. In many other places in Exim, IPv6 addresses are always
181 recognized, so I have changed this. It also means that IPv4 domain
182 literals of the form [IPV4:n.n.n.n] are now always recognized.
184 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
185 used if the router is not running as root, for example, when verifying at
186 ACL time, or when using -bh. The debugging output from this situation was
187 non-existent - all you got was a failure to exec. I have made two
190 (a) Failures to set uid/gid, the current directory, or a process leader
191 in a subprocess such as that created by queryprogram now generate
192 suitable debugging ouput when -d is set.
194 (b) The queryprogram router detects when it is not running as root,
195 outputs suitable debugging information if -d is set, and then runs
196 the subprocess without attempting to change uid/gid.
198 PH/15 Minor change to Makefile for building test_host (undocumented testing
201 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
202 additional section of a DNS packet that returns MX or SRV records.
203 Instead, it always explicitly searches for A/AAAA records. This avoids
204 major problems that occur when a DNS server includes only records of one
205 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
206 fixed another bug: if SRV records were looked up and the corresponding
207 address records were *not* found in the additional section, the port
208 values from the SRV records were lost.
210 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
211 using the correct key (the original address) when searching the retry
212 rules in order to find which one to use for generating the retry hint.
214 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
215 adding the default one. Similarly, if it contains a Reply-To: header, the
216 errors_reply_to option, if set, is not used.
218 PH/19 When calculating a retry time, Exim used to measure the "time since
219 failure" by looking at the "first failed" field in the retry record. Now
220 it does not use this if it is later than than the arrival time of the
221 message. Instead it uses the arrival time. This makes for better
222 behaviour in cases where some deliveries succeed, thus re-setting the
223 "first failed" field. An example is a quota failure for a huge message
224 when small messages continue to be delivered. Without this change, the
225 "time since failure" will always be short, possible causing more frequent
226 delivery attempts for the huge message than are intended.
227 [Note: This change was subsequently modified - see PH/04 for 4.62.]
229 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
230 $1, $2, $3) because the numerical variables can be reset during some
231 expansion items (e.g. "match"), thereby losing the authentication data.
233 PH/21 Make -bV show the size of off_t variables so that the test suite can
234 decide whether to run tests for quotas > 2G.
236 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
237 mailbox_size, and mailbox_filecount in the appendfile transport. If a
238 filecount value is greater than 2G or if a quota value is greater than 2G
239 on a system where the size of off_t is not greater than 4, a panic error
242 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
243 never match. The debug and -bh output now contains an explicit error
244 message indicating a malformed IPv4 address or mask.
246 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
247 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
250 PH/25 Do not write to syslog when running in the test harness. The only
251 occasion when this arises is a failure to open the main or panic logs
252 (for which there is an explicit test).
254 PH/26 Added the /no_tell option to "control=freeze".
256 PH/27 If a host name lookup failed very early in a connection, for example, if
257 the IP address matched host_lookup and the reverse lookup yielded a name
258 that did not have a forward lookup, an error message of the form "no IP
259 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
260 could be logged. Now it outputs the IP address instead of "NULL".
262 PH/28 An enabling patch from MH: add new function child_open_exim2() which
263 allows the sender and the authenticated sender to be set when
264 submitting a message from within Exim. Since child_open_exim() is
265 documented for local_scan(), the new function should be too.
267 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
268 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
269 results in an empty string is now treated as unset.
271 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
273 PH/31 Added sender_verify_fail logging option.
275 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
276 needed by RFC 822 but not by RFC 2822 was commented out. I have now
277 tidied the source and removed it altogether.
279 PH/33 When a queue run was abandoned because the load average was too high, a
280 log line was always written; now it is written only if the queue_run log
281 selector is set. In addition, the log line for abandonment now contains
282 information about the queue run such as the pid. This is always present
283 in "start" and "stop" lines but was omitted from the "abandon" line.
285 PH/34 Omit spaces between a header name and the colon in the error message that
286 is given when verify = headers_syntax fails (if there are lots of them,
287 the message gets confusing).
289 PH/35 Change the default for dns_check_names_pattern to allow slashes within
290 names, as there are now some PTR records that contain slashes. This check
291 is only to protect against broken name servers that fall over on strange
292 characters, so the fact that it applies to all lookups doesn't matter.
294 PH/36 Now that the new test suite is complete, we can remove some of the
295 special code in Exim that was needed for the old test suite. For example,
296 sorting DNS records because real resolvers return them in an arbitrary
297 order. The new test suite's fake resolver always returns records in the
300 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
301 bounces) except when queue_only is set, to avoid logging races between
302 the different processes.
304 PH/38 Panic-die if .include specifies a non-absolute path.
306 PH/39 A tweak to the "H" retry rule from its user.
308 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
309 a label. They prevented compilation on older perls.
311 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
312 a warning to be raised on newish perls.
314 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
315 on queue. Changes to match documented behaviour of showing count of
316 messages matching specified criteria.
318 PH/40 Changed the default ident timeout from 30s to 5s.
320 PH/41 Added support for the use of login_cap features, on those BSD systems
321 that have them, for controlling the resources used by pipe deliveries.
323 PH/42 The content-scanning code uses fopen() to create files in which to put
324 message data. Previously it was not paying any attention to the mode of
325 the files. Exim runs with umask(0) because the rest of the code creates
326 files with open(), and sets the required mode explicitly. Thus, these
327 files were ending up world-writeable. This was not a big issue, because,
328 being within the spool directory, they were not world-accessible. I have
329 created a function called modefopen, which takes an additional mode
330 argument. It sets umask(777), creates the file, chmods it to the required
331 mode, then resets the umask. All the relevant calls to fopen() in the
332 content scanning code have been changed to use this function.
334 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
335 to 24 hours. This avoids potential overflow problems when processing G
336 and H retry rules. I suspect nobody ever tinkers with this value.
338 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
340 PH/45 When the plaintext authenticator is running as a client, the server's
341 challenges are checked to ensure they are valid base64 strings. By
342 default, the authentication attempt is cancelled if an invalid string is
343 received. Setting client_ignore_invalid_base64 true ignores these errors.
344 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
345 they are received. Thus, the responses can be made to depend on the
346 challenges. If an invalid string is ignored, an empty string is placed in
349 PH/46 Messages that are created by the autoreply transport now contains a
350 References: header, in accordance with RFCs 2822 and 3834.
352 PH/47 Added authenticated_sender_force to the smtp transport.
354 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
356 PH/49 Installed latest patch from the Sieve maintainer.
358 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
359 was also set, the appendfile transport was unnecessarily scanning a
360 directory of message files (e.g. for maildir delivery) to find the count
361 of files (along with the size), even though it did not need this
362 information. It now does the scan only if it needs to find either the
363 size of the count of files.
365 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
367 PH/52 Two bugs concerned with error handling when the smtp transport is
370 (i) Exim was not creating retry information for temporary errors given
371 for individual recipients after the DATA command when the smtp transport
372 was used in LMTP mode. This meant that they could be retried too
373 frequently, and not timed out correctly.
375 (ii) Exim was setting the flag that allows error details to be returned
376 for LMTP errors on RCPT commands, but not for LMTP errors for individual
377 recipients that were returned after the DATA command.
379 PH/53 This is related to PH/52, but is more general: for any failing address,
380 when detailed error information was permitted to be returned to the
381 sender, but the error was temporary, then after the final timeout, only
382 "retry timeout exceeded" was returned. Now it returns the full error as
383 well as "retry timeout exceeded".
385 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
386 do this, and (what is worse) MTAs that accept it.
388 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
389 will now be deprecated.
391 PH/56 New os.c-cygwin from the Cygwin maintainer.
393 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
394 formats (previously only available in exim formats via -bpr, -bpru,
395 and -bpra. Now also available in native and exiqgrep formats)
397 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
398 with very large, slow to parse queues
400 JJ/08 exipick: added ! as generic prefix to negate any criteria format
402 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
404 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
405 responses to authentication challenges, though it was showing the
406 challenges; (ii) I've removed the CR characters from the debug output for
409 PH/58 Allow for the insertion of a newline as well as a space when a string
410 is turned into more than one encoded-word during RFC 2047 encoding. The
411 Sieve code now uses this.
413 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
414 data_4xx, lost_connection, tls_required.
416 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
417 message was being sent as an SMTP response.
419 PH/61 Add -l and -k options to exicyclog.
421 PH/62 When verifying, if an address was redirected to one new address, so that
422 verification continued, and the new address failed or deferred after
423 having set something in $address_data, the value of $address_data was not
424 passed back to the ACL. This was different to the case when no
425 redirection occurred. The value is now passed back in both cases.
427 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
428 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
429 use login_cap.h, so on its own it isn't the distinguishing feature. The
430 new name refers directly to the setclassresources() function.
432 PH/65 Added configuration files for NetBSD3.
434 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
436 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
437 is preferred over IPv4.
439 PH/68 The bounce_return_message and bounce_return_body options were not being
440 honoured for bounces generated during the reception of non-SMTP messages.
441 In particular, this applied to messages rejected by the ACL. This bug has
442 been fixed. However, if bounce_return_message is true and bounce_return_
443 body is false, the headers that are returned for a non-SMTP message
444 include only those that have been read before the error was detected.
445 (In the case of an ACL rejection, they have all been read.)
447 PH/69 The HTML version of the specification is now built in a directory called
448 spec_html instead of spec.html, because the latter looks like a path with
449 a MIME-type, and this confuses some software.
451 PH/70 Catch two compiler warnings in sieve.c.
453 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
454 function verify_get_ident() calls ip_connect() to connect a socket, but
455 if the "connect()" function timed out, ip_connect() used to close the
456 socket. However, verify_get_ident() also closes the socket later, and in
457 between Exim writes to the log, which may get opened at this point. When
458 the socket was closed in ip_connect(), the log could get the same file
459 descriptor number as the socket. This naturally causes chaos. The fix is
460 not to close the socket in ip_connect(); the socket should be closed by
461 the function that creates it. There was only one place in the code where
462 this was missing, in the iplookup router, which I don't think anybody now
463 uses, but I've fixed it anyway.
465 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
466 well as to direct DNS lookups. Otherwise the handling of names in host
467 lists is inconsistent and therefore confusing.
473 PH/01 Two changes to the default runtime configuration:
475 (1) Move the checks for relay_from_hosts and authenticated clients from
476 after to before the (commented out) DNS black list checks.
478 (2) Add control=submission to the relay_from_hosts and authenticated
479 clients checks, on the grounds that messages accepted by these
480 statements are most likely to be submissions.
482 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
484 (1) Generate an error if the third argument for the ${prvs expansion is
487 (2) Treat a missing third argument of ${prvscheck as if it were an empty
490 (3) Reset the variables that are obtained from the first argument of
491 ${prvscheck and used in the second argument before leaving the code,
492 because their memory is reclaimed, so using them afterwards may do
495 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
496 one (it's much easier than Tom thought :-).
498 (5) Because of (4), we can now allow for the use of $prvscheck_result
499 inside the third argument.
501 PH/03 For some reason, the default setting of PATH when running a command from
502 a pipe transport was just "/usr/bin". I have changed it to
505 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
506 anything to be listed in the output from -bV.
508 PH/05 When a filter generated an autoreply, the entire To: header line was
509 quoted in the delivery log line, like this:
511 => >A.N.Other <ano@some.domain> <original@ddress> ...
513 This has been changed so that it extracts the operative address. There
514 may be more than one such address. If so, they are comma-separated, like
517 => >ano@some.domain,ona@other.domain <original@ddress> ...
519 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
520 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
521 not being looked up in the rDNS to get a host name, Exim was showing the
522 IP address twice in Received: lines, even though the IP addresses were
523 identical. For example:
525 Received: from [1.2.3.4] (helo=[1.2.3.4])
527 However, if the real host name was known, it was omitting the HELO data
528 if it matched the actual IP address. This has been tidied up so that it
529 doesn't show the same IP address twice.
531 PH/07 When both +timestamp and +memory debugging was on, the value given by
532 $tod_xxx expansions could be wrong, because the tod_stamp() function was
533 called by the debug printing, thereby overwriting the timestamp buffer.
534 Debugging no longer uses the tod_stamp() function when +timestamp is set.
536 PH/08 When the original message was included in an autoreply transport, it
537 always said "this is a copy of the message, including all the headers",
538 even if body_only or headers_only was set. It now gives an appropriate
541 PH/09 Applied a patch from the Sieve maintainer which:
543 o fixes some comments
544 o adds the (disabled) notify extension core
545 o adds some debug output for the result of if/elsif tests
546 o points to the current vacation draft in the documentation
547 and documents the missing references header update
551 o fixes a bug in processing the envelope test (when testing
552 multiple envelope elements, the last element determinted the
555 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
556 Electronic Mail") by including:
558 Auto-submitted: auto-generated
560 in the messages that it generates (bounce messages and others, such as
561 warnings). In the case of bounce messages for non-SMTP mesages, there was
562 also a typo: it was using "Auto_submitted" (underscore instead of
563 hyphen). Since every message generated by Exim is necessarily in response
564 to another message, thes have all been changed to:
566 Auto-Submitted: auto-replied
568 in accordance with these statements in the RFC:
570 The auto-replied keyword:
572 - SHOULD be used on messages sent in direct response to another
573 message by an automatic process,
575 - MUST NOT be used on manually-generated messages,
577 - MAY be used on Delivery Status Notifications (DSNs) and Message
578 Disposition Notifications (MDNs),
580 - MUST NOT be used on messages generated by automatic or periodic
581 processes, except for messages which are automatic responses to
584 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
585 to the default Received: header definition.
587 PH/12 Added log selector acl_warn_skipped (default on).
589 PH/13 After a successful wildlsearch lookup, discard the values of numeric
590 variables because (a) they are in the wrong storage pool and (b) even if
591 they were copied, it wouldn't work properly because of the caching.
593 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
594 checking when decoding. Apparently there are clients that generate
595 overlong encoded strings. Why am I not surprised?
597 PH/15 If the first argument of "${if match_address" was not empty, but did not
598 contain an "@" character, Exim crashed. Now it writes a panic log message
599 and treats the condition as false.
601 PH/16 In autoreply, treat an empty string for "once" the same as unset.
603 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
604 extension "envelope-auth". The code is finished and in agreement with
605 other implementations, but there is no documentation so far and in fact,
606 nobody wrote the draft yet. This extension is currently #undef'ed, thus
607 not changing the active code.
609 Print executed "if" and "elsif" statements when debugging is used. This
610 helps a great deal to understand what a filter does.
612 Document more things not specified clearly in RFC3028. I had all this
613 sorted out, when out of a sudden new issues came to my mind. Oops."
615 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
618 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
619 canonical form (as documented). However, after a host name lookup from
620 the IP address, check_host() was doing a simple string comparison with
621 addresses acquired from the DNS when checking that the found name did
622 have the original IP as one of its addresses. Since any found IPv6
623 addresses are likely to be in abbreviated form, the comparison could
624 fail. Luckily, there already exists a function for doing the comparison
625 by converting both addresses to binary, so now that is used instead of
628 PH/20 There was another similar case to PH/19, when a complete host name was
629 given in a host list; looking up its IP address could give an abbreviated
630 form, whereas the current host's name might or might not be abbreviated.
631 The same fix has been applied.
637 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
638 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
641 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
642 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
644 PH/03 Typo: missing ".o" in src/pcre/Makefile.
646 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
647 header line, restrict the check to what is listed in RFCs 2369 and 2929.
648 Also, for "Auto-Submitted", treat anything other than "no" as
649 non-personal, in accordance with RFC 3834. (Previously it treated
650 anything starting "auto-" as non-personal.)
652 TF/01 The control=submission/name=... option had a problem with syntax
653 errors if the name included a slash character. The /name= option
654 now slurps the rest of the string, so it can include any characters
655 but it must come last in the list of options (after /sender_retain
658 PH/05 Some modifications to the interface to the fake nameserver for the new
666 TK/01 Added the "success_on_redirect" address verification option. See
667 NewStuff for rationale and an example.
669 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
671 PH/02 Patch to exigrep to allow it to work on syslog lines.
673 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
674 fread() to skip over the body file's header line, because in Cygwin the
675 header line is locked and is inaccessible.
677 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
678 co-exist for some time) to make it clear that it is the Exim ID that is
679 referenced, not the Message-ID: header line.
681 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
682 string_format() function, because snprintf() does not exist on all
685 PH/06 The use of forbid_filter_existstest now also locks out the use of the
686 ${stat: expansion item.
688 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
689 protocol synchronization error", to keep the pedants happy.
691 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
692 well as for IRIX systems, when gcc is being used. See the host.c source
695 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
697 PH/10 Named domain lists were not working if used in a queue_smtp_domains
700 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
701 transport and to the smtp transport in LMTP mode.
703 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
705 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
706 run a filter in a subprocess. This could lead to confusion in subsequent
707 lookups in the parent process. There should also be a search_tidyup() at
708 the end of the subprocess.
710 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
711 only if the host matched helo_try_verify_hosts, which caused the
712 verification to occur when the EHLO/HELO command was issued. The ACL just
713 tested the remembered result. Now, if a previous verification attempt has
714 not happened, "verify = helo" does it there and then.
716 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
718 TK/03 Fix log output including CR from clamd.
720 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
721 exist provoked a memory error which could cause a segfault.
723 PH/15 Installed PCRE 6.2
725 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
727 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
728 of the problem. Specifically, suggested +O2 rather than +O1 for the
731 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
733 PH/20 If a delivery was routed to a non-standard port by means of an SRV
734 record, the port was not correctly logged when the outgoing_port log
735 selector was set (it logged the transort's default port).
737 PH/21 Added support for host-specific ports to manualroute, queryprogram,
738 fallback_hosts, and "hosts" in the smtp transport.
740 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
741 host errors such as "Connection refused".
743 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
744 authentication with radiusclient 0.4.9:
746 - Error returned from rc_read_config was caught wrongly
747 - Username/password not passed on to radius server due to wrong length.
749 The presumption is that some radiusclient API changes for 4.51/PH/17
750 were not taken care of correctly. The code is still untested by me (my
751 Linux distribution still has 0.3.2 of radiusclient), but it was
752 contributed by a Radius user.
754 PH/24 When doing a callout, the value of $domain wasn't set correctly when
755 expanding the "port" option of the smtp transport.
757 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
758 while reading a MIME header. Thanks to Tom Hughes for a patch.
760 PH/24 Include config.h inside local_scan.h so that configuration settings are
763 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
764 that in an ACL for RCPT (for example), you can examine exactly what was
767 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
768 commands, but it was not correctly comparing the address with the actual
769 client host address. Thus, it would show the EHLO address in Received:
770 header lines when this was not necessary.
772 PH/27 Added the % operator to ${eval:}.
774 PH/28 Exim tries to create and chdir to its spool directory when it starts;
775 it should be ignoring failures (because with -C, for example, it has lost
776 privilege). It wasn't ignoring creation failures other than "already
779 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
782 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
783 because an input line was too long, either on its own, or by virtue of
784 too many continuations, the temporary file was not being removed, and the
785 return code was incorrect.
787 PH/31 Missing "BOOL" in function definition in filtertest.c.
789 PH/32 Applied Sieve patches from the maintainer.
791 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
793 PH/33 Added "verify = not_blind".
795 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
796 Local/Makefile (with some defaults set). These are used in built scripts
797 such as exicyclog, but they have never been used in the exim_install
798 script (though there are many overriding facilities there). I have
799 arranged that the exim_install script now takes note of these two
802 PH/35 Installed configuration files for Dragonfly.
804 PH/36 When a locally submitted message by a trusted user did not contain a
805 From: header, and the sender address was obtained from -f or from an SMTP
806 MAIL command, and the trusted user did not use -F to supply a sender
807 name, $originator_name was incorrectly used when constructing a From:
808 header. Furthermore, $originator_name was used for submission mode
809 messages from external hosts without From: headers in a similar way,
810 which is clearly wrong.
812 PH/37 Added control=suppress_local_fixups.
814 PH/38 When log_selector = +received_sender was set, and the addition of the
815 sender made the log line's construction buffer exactly full, or one byte
816 less than full, an overflow happened when the terminating "\n" was
819 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
820 when the result of a list match is failure because a DNS lookup failed.
822 PH/40 RM_COMMAND is now used in the building process.
824 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
825 the "build-* directories that it finds.
827 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
828 address in a domain literal was a prefix of an interface address.
830 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
831 when verifying a sender address, unless rewrite_headers is false.
833 PH/44 Wrote a long comment about why errors_to addresses are verified as
834 recipients, not senders.
836 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
837 the ratelimit ACL was added.
839 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
841 PH/46 Added extra information about PostgreSQL errors to the error string.
843 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
844 avoiding the need to install special zones in a real server. This is
845 backwards compatible; if it can't find the fake resolver, it drops back.
846 Thus, both old and new test suites can be run.
848 TF/02 Added util/ratelimit.pl
850 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
851 clock is set back in time.
853 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
854 Candler <B.Candler@pobox.com>.
856 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
857 OK for addresses that are the result of redirections.
859 PH/48 A number of further additions for the benefit of the new test suite,
860 including a fake gethostbyname() that interfaces to the fake DNS resolver
863 TF/06 The fix for widen_domains has also been applied to qualify_single and
864 search_parents which are the other dnslookup options that can cause
867 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
870 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
872 TF/07 Exim produced the error message "an SRV record indicated no SMTP
873 service" if it encountered an MX record with an empty target hostname.
874 The message is now "an MX or SRV record indicated no SMTP service".
876 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
877 if the DNS of the sending site is misconfigured. This is quite a
878 common situation. This change restores the behaviour of treating a
879 helo verification defer as a failure.
881 PH/51 If self=fail was set on a router, the bounce message did not include the
882 actual error message.
888 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
890 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
891 command itself ended in error, the underlying message about the transport
892 filter timeout was being overwritten with the pipe command error. Now the
893 underlying error message should be appended to the second error message.
895 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
897 PH/02 Reduce the amount of output that "make" produces by default. Full output
898 can still be requested.
900 PH/03 The warning log line about a condition test deferring for a "warn" verb
901 was being output only once per connection, rather than after each
902 occurrence (because it was using the same function as for successful
903 "warn" verbs). This seems wrong, so I have changed it.
905 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
906 it should not have, which might have caused a crash in the right
907 circumstances, but probably never did.
909 PH/04 Installed a modified version of Tony Finch's patch to make submission
910 mode fix the return path as well as the Sender: header line, and to
911 add a /name= option so that you can make the user's friendly name appear
914 TF/03 Added the control = fakedefer ACL modifier.
916 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
917 Mark Lowes for thorough testing.
919 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
921 TK/03 Merged latest SRS patch from Miles Wilton.
923 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
924 with the definition in sysexits.h (which is #included earlier).
925 Fortunately, Exim does not actually use EX_OK. The code used to try to
926 preserve the sysexits.h value, by assumimg that macro definitions were
927 scanned for macro replacements. I have been disabused of this notion,
928 so now the code just undefines EX_OK before #including unistd.h.
930 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
931 in the smtp transport. When a block could not be written in a single
932 write() function, the timeout was being re-applied to each part-write.
933 This seems wrong - if the receiver was accepting one byte at a time it
934 would take for ever. The timeout is now adjusted when this happens. It
935 doesn't have to be particularly precise.
937 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
938 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
940 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
941 without a domain if the check to <postmaster@domain> fails.
943 SC/01 Eximstats: added -xls and the ability to specify output files
944 (patch written by Frank Heydlauf).
946 SC/02 Eximstats: use FileHandles for outputing results.
948 SC/03 Eximstats: allow any combination of xls, txt, and html output.
950 SC/04 Eximstats: fixed display of large numbers with -nvr option
952 SC/05 Eximstats: fixed merging of reports with empty tables.
954 SC/06 Eximstats: added the -include_original_destination flag
956 SC/07 Eximstats: removed tabs and trailing whitespace.
958 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
960 TK/06 MBOX spool code: Add real "From " MBOX separator line
961 so the .eml file is really in mbox format (even though
962 most programs do not really care). Patch from Alex Miller.
964 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
965 The latter is generated from $received_to and is only set if the
966 message has one envelope recipient. SA can use these headers,
967 obviously out-of-the-box. Patch from Alex Miller.
969 PH/08 The ${def test on a variable was returning false if the variable's
970 value was "0", contrary to what the specification has always said!
971 The result should be true unless the variable is empty.
973 PH/09 The syntax error of a character other than { following "${if
974 def:variable_name" (after optional whitespace) was not being diagnosed.
975 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
976 accidental colon was present, for example, could give incorrect results.
978 PH/10 Tidied the code in a number of places where the st_size field of a stat()
979 result is used (not including appendfile, where other changes are about
982 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
983 This involved changing a lot of size variables from int to off_t. It
984 should work with maildirs and everything.
986 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
987 spamd dying while we are connected to it.
989 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
992 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
993 function for BDB 4.3.
995 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
997 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
998 the sources for PCRE have been re-arranged and more files are now
1001 PH/15 The code I had for printing potentially long long variables in PH/11
1002 above was not the best (it lost precision). The length of off_t variables
1003 is now inspected at build time, and an appropriate printing format (%ld
1004 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1005 to be "long long int" or "long int". This is needed for the internal
1006 formatting function string_vformat().
1008 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1009 the configuration file to be ":syslog", then the script "guesses" where
1010 the logs files are, rather than using the compiled in default. In our
1011 case the guess is not the same as the compiled default, so the script
1012 suddenly stopped working when I started to use syslog. The patch checks
1013 to see if log_file_path is "". If so, it attempts to read it from exim
1014 with no configuration file to get the compiled in version, before it
1015 falls back to the previous guessing code."
1017 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1018 implementing BATV in an Exim configuration. See NewStuff for the gory
1021 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1022 Makefile that are specific to HP-UX.
1024 PH/18 If the "use_postmaster" option was set for a recipient callout together
1025 with the "random" option, the postmaster address was used as the MAIL
1026 FROM address for the random test, but not for the subsequent recipient
1027 test. It is now used for both.
1029 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1030 patch removes a few documentation additions to RFC 3028, because the
1031 latest draft now contains them. It adds the new en;ascii-case comparator
1032 and a new error check for 8bit text in MIME parts. Comparator and
1033 require names are now matched exactly. I enabled the subaddress
1034 extension, but it is not well tested yet (read: it works for me)."
1036 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1037 rework some of the code of TK/09 above to avoid the hardwired use of
1038 "%lld" and "long long". Replaced the call to snprintf() with a call to
1041 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1042 records point to non-existent hosts", "retry timeout exceeded", and
1043 "retry time not reached for any host after a long failure period".
1045 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1046 experimental DomainKeys support:
1048 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1049 (2) On an error such as an illegally used "control", the wrong name for
1050 the control was given.
1052 These problems did NOT occur unless DomainKeys support was compiled.
1054 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1056 PH/24 Added ${if match_ip condition.
1058 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1059 databases so that it will be absolutely obvious if a crash occurs in the
1060 DB library. This is a regular occurrence (often caused by mis-matched
1063 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1064 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1065 picked up on a user's system that detects such things. There doesn't seem
1066 to be a gcc warning option for this - only an attribute that has to be
1067 put on the function's prototype. It seems that in Fedora Core 4 they have
1068 set this on a number of new functions. No doubt there will be more in due
1071 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1072 specify a transport. However, if an address that was verified by such a
1073 router was the subject of a callout, Exim crashed because it tried to
1074 read the rcpt_include_affixes from the non-existent transport. Now it
1075 just assumes that the setting of that option is false. This bug was
1076 introduced by 4.51/PH/31.
1078 PH/28 Changed -d+all to exclude +memory, because that information is very
1079 rarely of interest, but it makes the output a lot bigger. People tend to
1080 do -d+all out of habit.
1082 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1083 code in os-type was giving problems when libc.so lives in lib64, like on
1086 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1087 aren't the modern standard, and it seems that some systems' include files
1088 don't always have them. Exim was already checking for some of the newer
1089 ones like T_AAAA, and defining it itself. I've added checks for all the
1090 record types that Exim uses.
1092 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1093 not automatically generating a new one, as it is supposed to. This
1094 prevented TLS from working. If the file did exist, but contained invalid
1095 data, a new version was generated, as expected. It was only the case of a
1096 non-existent file that was broken.
1098 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1099 with a change in libdomainkeys > 0.64.
1101 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1102 from DNS. If the selector record carries the flag, it now has
1103 precedence over the domain-wide flag.
1105 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1107 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1108 the use of a port name that isn't defined in /etc/services) occurred, the
1109 message was deferred as in a normal delivery, and thus remained on the
1110 spool, instead of being failed because of the mua_wrapper setting. This
1111 is now fixed, and I tidied up some of the mua_wrapper messages at the
1114 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1115 the messages in a hash of arrays rather than using individual hashes.
1116 This is a bit cleaner and results in dramatic memory savings, albeit
1117 at a slight CPU cost.
1119 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1120 as requested by Marc Sherman.
1122 SC/10 Eximstats: added histograms for user specified patterns as requested
1125 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1127 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1128 fopen() in the content-scanning modules that did not already have it.
1134 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1135 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1137 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1139 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1141 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1142 to test Sieve filters that use "vacation".
1144 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1145 that changes the way the GnuTLS parameters are stored in the cache file.
1146 The new format can be generated externally. For backward compatibility,
1147 if the data in the cache doesn't make sense, Exim assumes it has read an
1148 old-format file, and it generates new data and writes a new file. This
1149 means that you can't go back to an older release without removing the
1152 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1153 work if there are any delivery delays because "one_time" forces the
1154 parent to be marked "delivered", so its unseen clone is never tried
1155 again. For this reason, Exim now forbids the simultaneous setting of
1158 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1159 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1160 entry if you want to know the details. The fix, however, neglected to
1161 consider the case when local delivery batching is involved. The test for
1162 "previously delivered" was not happening when checking to see if an
1163 address could be batched with a previous (undelivered) one; under
1164 certain circumstances this could lead to multiple deliveries to the same
1167 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1168 in its include files, and this causes problems building Exim.
1170 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1171 header_syntax) but Exim was just ignoring anything given after a slash.
1172 In particular, this caused confusion with an attempt to use "verify =
1173 reverse_host_lookup/defer_ok". An error is now given when options are
1174 supplied for verify items that do not have them. (Maybe reverse_host_
1175 lookup should have a defer_ok option, but that's a different point.)
1177 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1178 defined by RFC 821) to 2048, because there were problems with some AUTH
1179 commands, and RFC 1869 says the size should be increased for extended
1180 SMTP commands that take arguments.
1182 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1185 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1186 "unknown" error; now it says that the functionality isn't in the binary.
1188 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1189 an address' error message when a string expansion fails (syntax or
1190 whatever). Otherwise the password may appear in the log. Following change
1191 PH/42 below, there is no longer a chance of it appearing in a bounce
1194 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1196 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1197 of its addresses was ever tried. (Bugzilla bug #2.)
1199 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1200 the result incorrectly in the debug output. (It correctly added a newline
1201 to what was transported.)
1203 TF/01 Added $received_time.
1205 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1206 commented out examples of how to interface to a virus scanner and to
1207 SpamAssassin. Also added commented examples of av_scanner and
1208 spamd_address settings.
1210 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1211 and controls are allowed in which ACLs. There were a couple of minor
1212 errors. Some of the entries in the conditions table (which is a table of
1213 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1214 negation of where the condition IS allowed.
1216 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1218 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1219 header file does not have a version number, so I've had to invent a new
1220 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1221 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1222 radiusclient), but it was contributed by a Radius user.
1224 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1225 files or directories, for OpenSSL.
1227 PH/19 When an Exim process that is running as root has to create an Exim log
1228 file, it does so in a subprocess that runs as exim:exim so as to get the
1229 ownership right at creation (otherwise, other Exim processes might see
1230 the file with the wrong ownership). There was no test for failure of this
1231 fork() call, which would lead to the process getting stuck as it waited
1232 for a non-existent subprocess. Forks do occasionally fail when resources
1233 run out. I reviewed all the other calls to fork(); they all seem to check
1236 PH/20 When checking for unexpected SMTP input at connect time (before writing
1237 the banner), Exim was not dealing correctly with a non-positive return
1238 from the read() function. If the client had disconnected by this time,
1239 the result was a log entry for a synchronization error with an empty
1240 string after "input=" when read() returned zero. If read() returned -1
1241 (an event I could not check), uninitialized data bytes were printed.
1242 There were reports of junk text (parts of files, etc) appearing after
1245 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1247 PH/22 Added support for macro redefinition, and (re)definition in between
1248 driver and ACL definitions.
1250 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1251 forgetting to use the resulting value; it was using the unexpanded value.
1253 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1254 hadn't been configured. The fix is from Juergen Kreileder, who
1255 understands it better than I do:
1257 "Here's what I see happening with three configured cyrus_sasl
1258 authenticators configured (plain, login, cram-md5):
1260 On startup auth_cyrus_sasl_init() gets called for each of these.
1261 This means three calls to sasl_listmech() without a specified mech_list.
1262 => SASL tests which mechs of all available mechs actually work
1263 => three warnings about OTP not working
1264 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1266 With the patch, sasl_listmech() also gets called three times. But now
1267 SASL's mech_list option is set to the server_mech specified in the the
1268 authenticator. Or in other words, the answer from sasl_listmech()
1269 gets limited to just the mech you're testing for (which is different
1271 => the return list contains just 'plain' or 'login', 'cram-md5' or
1272 nothing depending on the value of ob->server_mech.
1274 I've just tested the patch: Authentication still works fine,
1275 unavailable mechs specified in the exim configuration are still
1276 caught, and the auth.log warnings about OTP are gone."
1278 PH/25 When debugging is enabled, the contents of the command line are added
1279 to the debugging output, even when log_selector=+arguments is not
1282 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1283 answer is "GNU", and only if the return is "GNU/something" is the answer
1286 PH/27 $acl_verify_message is now set immediately after the failure of a
1287 verification in an ACL, and so is available in subsequent modifiers. In
1288 particular, the message can be preserved by coding like this:
1290 warn !verify = sender
1291 set acl_m0 = $acl_verify_message
1293 Previously, $acl_verify_message was set only while expanding "message"
1294 and "log_message" when a very denied access.
1296 PH/28 Modified OS/os.c-Linux with
1298 -#ifndef OS_LOAD_AVERAGE
1299 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1301 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1302 nomenclature these days.)
1304 PH/29 Installed patch from the Sieve maintainer that adds the options
1305 sieve_useraddress and sieve_subaddress to the redirect router.
1307 PH/30 In these circumstances:
1308 . Two addresses routed to the same list of hosts;
1309 . First host does not offer TLS;
1310 . First host accepts first address;
1311 . First host gives temporary error to second address;
1312 . Second host offers TLS and a TLS session is established;
1313 . Second host accepts second address.
1314 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1315 and peerdn, if requested) that were in fact used only for the second
1318 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1319 attention to any local part prefix or suffix that was matched by the
1320 router that accepted the address. It now behaves in the same way as it
1321 does for delivery: the affixes are removed from the local part unless
1322 rcpt_include_affixes is set on the transport.
1324 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1325 timeout during the DATA phase of an incoming message.
1327 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1328 applied a patch sanctioned by the Sieve maintainer.
1330 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1331 the uid or gid is negative. A case of a negative gid caused this to be
1332 noticed. The fix allows for either to be negative.
1334 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1335 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1338 PH/36 The change PH/12 above was broken. Fixed it.
1340 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1341 the grounds that routing the same address twice would always produce the
1342 same answer. This might have been true once, but it is certainly no
1343 longer true now. Routing a child address may depend on the previous
1344 routing that produced that child. Some complicated redirection strategies
1345 went wrong when messages had multiple recipients, and made Exim's
1346 behaviour dependent on the order in which the addresses were given.
1348 I have moved the duplicate checking until after the routing is complete.
1349 Exim scans the addresses that are assigned to local and remote
1350 transports, and removes any duplicates. This means that more work will be
1351 done, as duplicates will always all be routed, but duplicates are
1352 presumably rare, so I don't expect this is of any significance.
1354 For deliveries to pipes, files, and autoreplies, the duplicate checking
1355 still happens during the routing process, since they are not going to be
1358 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1359 It corrects a timeout issue with spamd. This is Ian's comment: "The
1360 background is that sometimes spamd either never reads data from a
1361 connection it has accepted, or it never writes response data. The exiscan
1362 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1363 blindly assumes that writes won't block so it may never time out."
1365 PH/39 Allow G after quota size as well as K and M.
1367 PH/40 The value set for $authenticated_id in an authenticator may not contain
1368 binary zeroes or newlines because the value is written to log lines and
1369 to spool files. There was no check on this. Now the value is run through
1370 the string_printing() function so that such characters are converted to
1371 printable escape sequences.
1373 PH/41 $message_linecount is a new variable that contains the total number of
1374 lines in the message. Compare $body_linecount, which is the count for the
1377 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1378 bounce and delay warning messages, except in certain special cases, which
1381 (a) An SMTP error message from a remote host;
1382 (b) A message specified in a :fail: redirection;
1383 (c) A message specified in a "fail" command in a system filter;
1384 (d) A message specified in a FAIL return from the queryprogram router;
1385 (e) A message specified by the cannot_route_message router option.
1387 In these cases only, Exim does include the error details in bounce and
1388 warning messages. There are also a few cases where bland messages such
1389 as "unrouteable address" or "local delivery error" are given.
1391 PH/43 $value is now also set for the "else" part of a ${run expansion.
1393 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1394 being worked on, but at least Exim now implements the latest version to
1397 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1398 process to complete was treated as a delivery failure, a timeout while
1399 writing the message to the pipe was logged, but erroneously treated as a
1400 successful delivery. Such timeouts include transport filter timeouts. For
1401 consistency with the overall process timeout, these timeouts are now
1402 treated as errors, giving rise to delivery failures by default. However,
1403 there is now a new Boolean option for the pipe transport called
1404 timeout_defer, which, if set TRUE, converts the failures into defers for
1405 both kinds of timeout. A transport filter timeout is now identified in
1408 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1409 systems where "make" and "gmake" are different, calling "gmake" at top
1410 level broke things. I've arranged for the value of $(MAKE) to be passed
1411 from the Makefile to this script so that it can call the same version of
1415 A note about Exim versions 4.44 and 4.50
1416 ----------------------------------------
1418 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1419 changes of various kinds. As a consequence, a big documentation update was
1420 needed. This delayed the release for rather longer than seemed good, especially
1421 in the light of a couple of (minor) security issues. Therefore, the changes
1422 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1423 release. So 4.44 and 4.50 are in effect two different branches that both start
1426 I have left the 4.50 change log unchanged; it contains all the changes since
1427 4.43. The change log for 4.44 is below; many of its items are identical to
1428 those for 4.50. This seems to be the most sensible way to preserve the
1429 historical information.
1435 1. Minor wording change to the doc/README.SIEVE file.
1437 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1438 computation of the current number of files was incorrect.
1440 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1441 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1444 4. Give more explanation in the error message when the command for a transport
1445 filter fails to execute.
1447 5. There are several places where Exim runs a non-Exim command in a
1448 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1449 was being done only for the command run by the queryprogram router. It is
1450 now done for all such subprocesses. The other cases are: ${run, transport
1451 filters, and the commands run by the lmtp and pipe transports.
1453 6. Added CONFIGURE_GROUP build-time option.
1455 7. Some older OS have a limit of 256 on the maximum number of file
1456 descriptors. Exim was using setrlimit() to set 1000 as a large value
1457 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1458 systems. I've change it so that if it can't get 1000, it tries for 256.
1460 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1461 was an oversight, and furthermore, ever since the addition of extra
1462 controls (e.g. 4.43/32), the checks on when to allow different forms of
1463 "control" were broken. There should now be diagnostics for all cases when a
1464 control that does not make sense is encountered.
1466 9. Added the /retain_sender option to "control=submission".
1468 10. $recipients is now available in the predata ACL (oversight).
1470 11. Tidy the search cache before the fork to do a delivery from a message
1471 received from the command line. Otherwise the child will trigger a lookup
1472 failure and thereby defer the delivery if it tries to use (for example) a
1473 cached ldap connection that the parent has called unbind on.
1475 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1476 of $address_data from the recipient verification was clobbered by the
1477 sender verification.
1479 13. The value of address_data from a sender verification is now available in
1480 $sender_address_data in subsequent conditions in the ACL statement.
1482 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1484 15. Added a new option "connect=<time>" to callout options, to set a different
1487 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1488 was its contents. (It was OK if the option was not defined at all.)
1490 17. A "Completed" log line is now written for messages that are removed from
1491 the spool by the -Mrm option.
1493 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1494 information about exactly what failed.
1496 19. Added -dd to debug only the daemon process.
1498 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1499 handles timeouts, both on the server side and network timeouts. Renamed the
1500 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1502 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1503 It is now set to "smtps".
1505 22. $host_address is now set to the target address during the checking of
1506 ignore_target_hosts.
1508 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1509 being passed; this would have caused $sender_host_name to have been used if
1510 matching the list had actually called for a host name (not very likely,
1511 since this list is usually IP addresses). A host name is now passed as
1514 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1515 code that specifies a non-restarting handler (typically sigaction() in
1516 modern systems) in an attempt to fix a rare and obscure crash bug.
1518 25. Narrowed the window for a race in the daemon that could cause it to ignore
1519 SIGCHLD signals. This is not a major problem, because they are used only to
1520 wake it up if nothing else does.
1522 26. A malformed maildirsize file could cause Exim to calculate negative values
1523 for the mailbox size or file count. Odd effects could occur as a result.
1524 The maildirsize information is now recalculated if the size or filecount
1527 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1528 support for a long time. Removed HAVE_SYS_VFS_H.
1530 28. Installed the latest version of exipick from John Jetmore.
1532 29. In an address list, if the pattern was not a regular expression, an empty
1533 subject address (from a bounce message) matched only if the pattern was an
1534 empty string. Non-empty patterns were not even tested. This was the wrong
1535 because it is perfectly reasonable to use an empty address as part of a
1536 database query. An empty address is now tested by patterns that are
1537 lookups. However, all the other forms of pattern expect the subject to
1538 contain a local part and a domain, and therefore, for them, an empty
1539 address still always fails if the pattern is not itself empty.
1541 30. Exim went into a mad DNS loop when attempting to do a callout where the
1542 host was specified on an smtp transport, and looking it up yielded more
1543 than one IP address.
1545 31. Re-factored the code for checking spool and log partition space into a
1546 function that finds that data and another that does the check. The former
1547 is then used to implement four new variables: $spool_space, $log_space,
1548 $spool_inodes, and $log_inodes.
1550 32. The RFC2047 encoding function was originally intended for short strings
1551 such as real names; it was not keeping to the 75-character limit for
1552 encoded words that the RFC imposes. It now respects the limit, and
1553 generates multiple encoded words if necessary. To be on the safe side, I
1554 have increased the buffer size for the ${rfc2047: expansion operator from
1557 33. It is now permitted to omit both strings after an "if" condition; if the
1558 condition is true, the result is "true". As before, when the second string
1559 is omitted, a false condition yields an empty string. This makes it less
1560 cumbersome to write custom ACL and router conditions.
1562 34. Failure to deliver a bounce message always caused it to be frozen, even if
1563 there was an errors_to setting on the router. The errors_to setting is now
1566 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1567 canonical form (fully expanded) before being placed in
1568 $sender_host_address.
1570 36. The table in the code that translates DNS record types into text (T_A to
1571 "A" for instance) was missing entries for NS and CNAME. It is just possible
1572 that this could have caused confusion if both these types were looked up
1573 for the same domain, because the text type is used as part of Exim's
1574 per-process caching. But the chance of anyone hitting this buglet seems
1577 37. The dnsdb lookup has been extended in a number of ways.
1579 (1) There is a new type, "zns", which walks up the domain tree until it
1580 finds some nameserver records. It should be used with care.
1582 (2) There is a new type, "mxh", which is like "mx" except that it returns
1583 just the host names, not the priorities.
1585 (3) It is now possible to give a list of domains (or IP addresses) to be
1586 looked up. The behaviour when one of the lookups defers can be
1587 controlled by a keyword.
1589 (4) It is now possible to specify the separator character for use when
1590 multiple records are returned.
1592 38. The dnslists ACL condition has been extended: it is now possible to supply
1593 a list of IP addresses and/or domains to be looked up in a particular DNS
1596 39. Added log_selector=+queue_time_overall.
1598 40. When running the queue in the test harness, wait just a tad after forking a
1599 delivery process, to get repeatability of debugging output.
1601 41. Include certificate and key file names in error message when GnuTLS fails
1602 to set them up, because the GnuTLS error message doesn't include the name
1603 of the failing file when there is a problem reading it.
1605 42. Allow both -bf and -bF in the same test run.
1607 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1609 44. The "Exiscan patch" is now merged into the mainline Exim source.
1611 45. Sometimes the final signoff response after QUIT could fail to get
1612 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1613 < 0 before doing a fflush(). This bug looks as though it goes back to the
1614 introduction of TLS in release 3.20, but "sometimes" must have been rare
1615 because the tests only now provoked it.
1617 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1618 (this can affect the format of dates).
1620 47. exim_tidydb, when checking for the continued existence of a message for
1621 which it has found a message-specific retry record, was not finding
1622 messages that were in split spool directories. Consequently, it was
1623 deleting retry records that should have stayed in existence.
1625 48. Steve fixed some bugs in eximstats.
1627 49. The SPA authentication driver was not abandoning authentication and moving
1628 on to the next authenticator when an expansion was forced to fail,
1629 contradicting the general specification for all authenticators. Instead it
1630 was generating a temporary error. It now behaves as specified.
1632 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1633 (the order specifies the preference for clients). The order is now AES256,
1634 AES128, 3DES, ARCFOUR128.
1636 51. Small patch to Sieve code - explicitly set From: when generating an
1639 52. Exim crashed if a remote delivery caused a very long error message to be
1640 recorded - for instance if somebody sent an entire SpamAssassin report back
1641 as a large number of 550 error lines. This bug was coincidentally fixed by
1642 increasing the size of one of Exim's internal buffers (big_buffer) that
1643 happened as part of the Exiscan merge. However, to be on the safe side, I
1644 have made the code more robust (and fixed the comments that describe what
1647 53. Now that there can be additional text after "Completed" in log lines (if
1648 the queue_time_overall log selector is set), a one-byte patch to exigrep
1649 was needed to allow it to recognize "Completed" as not the last thing in
1652 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1653 patch that reportedly fixes this has been added. I am not expert enough to
1654 create a test for it. This is what the patch creator wrote:
1656 "I found a little strange behaviour of ldap code when working with
1657 Windows 2003 AD Domain, where users was placed in more than one
1658 Organization Units. When I tried to give exim partial DN, the exit code
1659 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1660 But simultaneously result of request was absolutely normal ldap result,
1661 so I produce this patch..."
1663 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1664 REFERENCE, so I have modified the code to exclude the patch when that macro
1667 55. Some experimental protocols are using DNS PTR records for new purposes. The
1668 keys for these records are domain names, not reversed IP addresses. The
1669 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1670 leaves it alone. Component reversal etc. now happens only for IP addresses.
1673 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1675 57. Double the size of the debug message buffer (to 2048) so that more of very
1676 long debug lines gets shown.
1678 58. The exicyclog utility now does better if the number of log files to keep
1679 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1681 59. Two changes related to the smtp_active_hostname option:
1683 (1) $smtp_active_hostname is now available as a variable.
1684 (2) The default for smtp_banner uses $smtp_active_hostname instead
1685 of $primary_hostname.
1687 60. The host_aton() function is supposed to be passed a string that is known
1688 to be a valid IP address. However, in the case of IPv6 addresses, it was
1689 not checking this. This is a hostage to fortune. Exim now panics and dies
1690 if the condition is not met. A case was found where this could be provoked
1691 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1692 components; fortuitously, this particular loophole had already been fixed
1693 by change 4.50/55 above.
1695 If there are any other similar loopholes, the new check in host_aton()
1696 itself should stop them being exploited. The report I received stated that
1697 data on the command line could provoke the exploit when Exim was running as
1698 exim, but did not say which command line option was involved. All I could
1699 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1700 running as the user.
1703 61. There was a buffer overflow vulnerability in the SPA authentication code
1704 (which came originally from the Samba project). I have added a test to the
1705 spa_base64_to_bits() function which I hope fixes it.
1708 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1709 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1711 63. The daemon start-up calls getloadavg() while still root for those OS that
1712 need the first call to be done as root, but it missed one case: when
1713 deliver_queue_load_max is set with deliver_drop_privilege. This is
1714 necessary for the benefit of the queue runner, because there is no re-exec
1715 when deliver_drop_privilege is set.
1717 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1718 This has been fixed.
1720 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1721 was in use, was not putting the data itself into the right store pool;
1722 consequently, it could be overwritten for a subsequent message in the same
1723 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1726 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1728 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1729 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1730 the function were treating the return as a boolean value, which happened to
1731 work because 0=false and not-0=true, but is not correct code.
1733 68. The host_aton() function was not handling scoped IPv6 addresses (those
1734 with, for example, "%eth0" on the end) correctly.
1736 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1737 negated items (that is, ~something) in unsigned ints. Some compilers
1738 apparently mutter when there is no cast.
1740 70. If an address verification called from an ACL failed, and did not produce a
1741 user-specific message (i.e. there was only a "system" message), nothing was
1742 put in $acl_verify_message. In this situation, it now puts the system
1745 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1746 session; change 4.31/43 added the unwanted input to the log line - except
1747 that it did not do this in the start of session case. It now does.
1749 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1750 This is wrong and can cause the other end to generate a synchronization
1751 error if it is another Exim or anything else that does the synchronization
1752 check. A QUIT command is no longer sent after a timeout.
1754 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1755 during host lookups.
1757 74. The defer_ok option of callout verification was not working if it was used
1758 when verifying addresses in header lines, that is, for this case:
1760 verify = header_sender/callout=defer_ok
1762 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1763 those file descriptors could be used for SMTP connections. If anything
1764 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1765 could be sent to the SMTP client, causing chaos. The daemon now opens
1766 stdin, stdout, and stderr to /dev/null when it puts itself into the
1769 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1770 log by default. The user can override this with suitable Perl magic.
1772 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1773 the log message when discard triggers, was not working for the DATA ACL or
1774 for the non-SMTP ACL.
1776 78. Error message wording change in sieve.c.
1778 79. If smtp_accept_max_per_host was set, the number of connections could be
1779 restricted to fewer than expected, because the daemon was trying to set up
1780 a new connection before checking whether the processes handling previous
1781 connections had finished. The check for completed processes is now done
1782 earlier. On busy systems, this bug wouldn't be noticed because something
1783 else would have woken the daemon, and it would have reaped the completed
1786 80. If a message was submitted locally by a user whose login name contained one
1787 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1788 It caused a spool format error. I have fixed the spool reading code. A
1789 related problem was that the "from" clause in the Received: line became
1790 illegal because of the space(s). It is now covered by ${quote_local_part.
1792 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1795 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1797 83. Patch from Sieve maintainer to fix unterminated string problem in
1798 "vacation" handling.
1800 84. Some minor changes to the Linux configuration files to help with other
1801 OS variants using glibc.
1803 85. One more patch for Sieve to update vacation handling to latest spec.
1806 ----------------------------------------------------
1807 See the note above about the 4.44 and 4.50 releases.
1808 ----------------------------------------------------
1814 1. Change 4.43/35 introduced a bug that caused file counts to be
1815 incorrectly computed when quota_filecount was set in an appendfile
1818 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1819 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1822 3. Give more explanation in the error message when the command for a transport
1823 filter fails to execute.
1825 4. There are several places where Exim runs a non-Exim command in a
1826 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1827 was being done only for the command run by the queryprogram router. It is
1828 now done for all such subprocesses. The other cases are: ${run, transport
1829 filters, and the commands run by the lmtp and pipe transports.
1831 5. Some older OS have a limit of 256 on the maximum number of file
1832 descriptors. Exim was using setrlimit() to set 1000 as a large value
1833 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1834 systems. I've change it so that if it can't get 1000, it tries for 256.
1836 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1837 was an oversight, and furthermore, ever since the addition of extra
1838 controls (e.g. 4.43/32), the checks on when to allow different forms of
1839 "control" were broken. There should now be diagnostics for all cases when a
1840 control that does not make sense is encountered.
1842 7. $recipients is now available in the predata ACL (oversight).
1844 8. Tidy the search cache before the fork to do a delivery from a message
1845 received from the command line. Otherwise the child will trigger a lookup
1846 failure and thereby defer the delivery if it tries to use (for example) a
1847 cached ldap connection that the parent has called unbind on.
1849 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1850 of $address_data from the recipient verification was clobbered by the
1851 sender verification.
1853 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1854 was its contents. (It was OK if the option was not defined at all.)
1856 11. A "Completed" log line is now written for messages that are removed from
1857 the spool by the -Mrm option.
1859 12. $host_address is now set to the target address during the checking of
1860 ignore_target_hosts.
1862 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1863 being passed; this would have caused $sender_host_name to have been used if
1864 matching the list had actually called for a host name (not very likely,
1865 since this list is usually IP addresses). A host name is now passed as
1868 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1869 code that specifies a non-restarting handler (typically sigaction() in
1870 modern systems) in an attempt to fix a rare and obscure crash bug.
1872 15. Narrowed the window for a race in the daemon that could cause it to ignore
1873 SIGCHLD signals. This is not a major problem, because they are used only to
1874 wake it up if nothing else does.
1876 16. A malformed maildirsize file could cause Exim to calculate negative values
1877 for the mailbox size or file count. Odd effects could occur as a result.
1878 The maildirsize information is now recalculated if the size or filecount
1881 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1882 support for a long time. Removed HAVE_SYS_VFS_H.
1884 18. Updated exipick to current release from John Jetmore.
1886 19. Allow an empty sender to be matched against a lookup in an address list.
1887 Previously the only cases considered were a regular expression, or an
1890 20. Exim went into a mad DNS lookup loop when doing a callout where the
1891 host was specified on the transport, if the DNS lookup yielded more than
1894 21. The RFC2047 encoding function was originally intended for short strings
1895 such as real names; it was not keeping to the 75-character limit for
1896 encoded words that the RFC imposes. It now respects the limit, and
1897 generates multiple encoded words if necessary. To be on the safe side, I
1898 have increased the buffer size for the ${rfc2047: expansion operator from
1901 22. Failure to deliver a bounce message always caused it to be frozen, even if
1902 there was an errors_to setting on the router. The errors_to setting is now
1905 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1906 canonical form (fully expanded) before being placed in
1907 $sender_host_address.
1909 24. Updated eximstats to version 1.33
1911 25. Include certificate and key file names in error message when GnuTLS fails
1912 to set them up, because the GnuTLS error message doesn't include the name
1913 of the failing file when there is a problem reading it.
1915 26. Expand error message when OpenSSL has problems setting up cert/key files.
1918 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1919 (this can affect the format of dates).
1921 28. exim_tidydb, when checking for the continued existence of a message for
1922 which it has found a message-specific retry record, was not finding
1923 messages that were in split spool directories. Consequently, it was
1924 deleting retry records that should have stayed in existence.
1926 29. eximstats updated to version 1.35
1927 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1928 1.35 - bugfix such that pie charts by volume are generated correctly
1930 30. The SPA authentication driver was not abandoning authentication and moving
1931 on to the next authenticator when an expansion was forced to fail,
1932 contradicting the general specification for all authenticators. Instead it
1933 was generating a temporary error. It now behaves as specified.
1935 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1936 (the order specifies the preference for clients). The order is now AES256,
1937 AES128, 3DES, ARCFOUR128.
1939 31. Small patch to Sieve code - explicitly set From: when generating an
1942 32. Exim crashed if a remote delivery caused a very long error message to be
1943 recorded - for instance if somebody sent an entire SpamAssassin report back
1944 as a large number of 550 error lines. This bug was coincidentally fixed by
1945 increasing the size of one of Exim's internal buffers (big_buffer) that
1946 happened as part of the Exiscan merge. However, to be on the safe side, I
1947 have made the code more robust (and fixed the comments that describe what
1950 33. Some experimental protocols are using DNS PTR records for new purposes. The
1951 keys for these records are domain names, not reversed IP addresses. The
1952 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1953 leaves it alone. Component reversal etc. now happens only for IP addresses.
1956 34. The host_aton() function is supposed to be passed a string that is known
1957 to be a valid IP address. However, in the case of IPv6 addresses, it was
1958 not checking this. This is a hostage to fortune. Exim now panics and dies
1959 if the condition is not met. A case was found where this could be provoked
1960 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1961 components; fortuitously, this particular loophole had already been fixed
1962 by change 4.50/55 or 4.44/33 above.
1964 If there are any other similar loopholes, the new check in host_aton()
1965 itself should stop them being exploited. The report I received stated that
1966 data on the command line could provoke the exploit when Exim was running as
1967 exim, but did not say which command line option was involved. All I could
1968 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1969 running as the user.
1972 35. There was a buffer overflow vulnerability in the SPA authentication code
1973 (which came originally from the Samba project). I have added a test to the
1974 spa_base64_to_bits() function which I hope fixes it.
1977 36. The daemon start-up calls getloadavg() while still root for those OS that
1978 need the first call to be done as root, but it missed one case: when
1979 deliver_queue_load_max is set with deliver_drop_privilege. This is
1980 necessary for the benefit of the queue runner, because there is no re-exec
1981 when deliver_drop_privilege is set.
1983 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1984 was in use, was not putting the data itself into the right store pool;
1985 consequently, it could be overwritten for a subsequent message in the same
1986 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1989 38. Sometimes the final signoff response after QUIT could fail to get
1990 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1991 < 0 before doing a fflush(). This bug looks as though it goes back to the
1992 introduction of TLS in release 3.20, but "sometimes" must have been rare
1993 because the tests only now provoked it.
1999 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2000 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2001 Now it returns an integer. A number of calls were still expecting a T/F
2002 return. Fortuitously, in all cases, the tests worked in OK situations,
2003 which is the norm. However, things would have gone wrong on any write
2004 failures on the smtp file descriptor. This function is used when sending
2005 messages over SMTP and also when doing verify callouts.
2007 2. When Exim is called to do synchronous delivery of a locally submitted
2008 message (the -odf or -odi options), it no longer closes stderr before doing
2011 3. Implemented the mua_wrapper option.
2013 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2015 5. Implemented the functions header_remove(), header_testname(),
2016 header_add_at_position(), and receive_remove_recipient(), and exported them
2019 6. If an ACL "warn" statement specified the addition of headers, Exim already
2020 inserted X-ACL-Warn: at the start if there was no header name. However, it
2021 was not making this test for the second and subsequent header lines if
2022 there were newlines in the string. This meant that an invalid header could
2023 be inserted if Exim was badly configured.
2025 7. Allow an ACL "warn" statement to add header lines at the start or after all
2026 the Received: headers, as well as at the end.
2028 8. Added the rcpt_4xx retry error code.
2030 9. Added postmaster_mailfrom=xxx to callout verification option.
2032 10. Added mailfrom=xxxx to the callout verification option, for verify=
2035 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2036 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2038 12. Inserted some casts to stop certain compilers warning when using pointer
2039 differences as field lengths or precisions in printf-type calls (mostly
2040 affecting debugging statements).
2042 13. Added optional readline() support for -be (dynamically loaded).
2044 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2045 same clock tick as a message's arrival, so that its received time was the
2046 same as the "first fail" time on the retry record, and that message
2047 remained on the queue past the ultimate address timeout, every queue runner
2048 would try a delivery (because it was past the ultimate address timeout) but
2049 after another failure, the ultimate address timeout, which should have then
2050 bounced the address, did not kick in. This was a "< instead of <=" error;
2051 in most cases the first failure would have been in the next clock tick
2052 after the received time, and all would be well.
2054 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2055 being recognized when the domain list was tested by the match_domain
2056 condition in an expansion string.
2058 16. Added the ${str2b64: operator.
2060 17. Exim was always calling setrlimit() to set a large limit for the number of
2061 processes, without checking whether the existing limit was already
2062 adequate. (It did check for the limit on file descriptors.) Furthermore,
2063 errors from getrlimit() and setrlimit() were being ignored. Now they are
2064 logged to the main and panic logs, but Exim does carry on, to try to do its
2065 job under whatever limits there are.
2067 18. Imported PCRE 5.0.
2069 19. Trivial typo in log message " temporarily refused connection" (the leading
2072 20. If the log selector return_path_on_delivery was set and an address was
2073 redirected to /dev/null, the delivery process crashed because it assumed
2074 that a return path would always be set for a "successful" delivery. In this
2075 case, the whole delivery is bypassed as an optimization, and therefore no
2078 21. Internal re-arrangement: the function for sending a challenge and reading
2079 a response while authentication was assuming a zero-terminated challenge
2080 string. It's now changed to take a pointer and a length, to allow for
2081 binary data in such strings.
2083 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2085 23. Exim was not respecting finduser_retries when seeking the login of the
2086 uid under which it was called; it was always trying 10 times. (The default
2087 setting of finduser_retries is zero.) Also, it was sleeping after the final
2088 failure, which is pointless.
2090 24. Implemented tls_on_connect_ports.
2092 25. Implemented acl_smtp_predata.
2094 26. If the domain in control=submission is set empty, Exim assumes that the
2095 authenticated id is a complete email address when it generates From: or
2096 Sender: header lines.
2098 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2099 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2100 chown and chgrp in /bin and hostname in /usr/bin.
2102 28. Exim was keeping the "process log" file open after each use, just as it
2103 does for the main log. This opens the possibility of it remaining open for
2104 long periods when the USR1 signal hits a daemon. Occasional processlog
2105 errors were reported, that could have been caused by this. Anyway, it seems
2106 much more sensible not to leave this file open at all, so that is what now
2109 29. The long-running daemon process does not normally write to the log once it
2110 has entered its main loop, and it closes the log before doing so. This is
2111 so that log files can straightforwardly be renamed and moved. However,
2112 there are a couple of unusual error situations where the daemon does write
2113 log entries, and I had neglected to close the log afterwards.
2115 30. The text of an SMTP error response that was received during a remote
2116 delivery was being truncated at 512 bytes. This is too short for some of
2117 the long messages that one sometimes sees. I've increased the limit to
2120 31. It is now possible to make retry rules that apply only when a message has a
2121 specific sender, in particular, an empty sender.
2123 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2124 it possible to be selective about when SMTP synchronization is enforced.
2126 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2128 32. Implemented hosts_connection_nolog.
2130 33. Added an ACL for QUIT.
2132 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2135 35. Added mailbox_size and mailbox_filecount to appendfile.
2137 36. Added control = no_multiline_responses to ACLs.
2139 37. There was a bug in the logic of the code that waits for the clock to tick
2140 in the case where the clock went backwards by a substantial amount such
2141 that the microsecond fraction of "now" was more than the microsecond
2142 fraction of "then" (but the whole seconds number was less).
2144 38. Added support for the libradius Radius client library this is found on
2145 FreeBSD (previously only the radiusclient library was supported).
2151 1. When certain lookups returned multiple values in the form name=value, the
2152 quoting of the values was not always being done properly. Specifically:
2153 (a) If the value started with a double quote, but contained no whitespace,
2155 (b) If the value contained whitespace other than a space character (i.e.
2156 tabs or newlines or carriage returns) it was not quoted.
2157 This fix has been applied to the mysql and pgsql lookups by writing a
2158 separate quoting function and calling it from the lookup code. The fix
2159 should probably also be applied to nisplus, ibase and oracle lookups, but
2160 since I cannot test any of those, I have not disturbed their existing code.
2162 2. A hit in the callout cache for a specific address caused a log line with no
2163 reason for rejecting RCPT. Now it says "Previous (cached) callout
2164 verification failure".
2166 3. There was an off-by-one bug in the queryprogram router. An over-long
2167 return line was truncated at 256 instead of 255 characters, thereby
2168 overflowing its buffer with the terminating zero. As well as fixing this, I
2169 have increased the buffer size to 1024 (and made a note to document this).
2171 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2172 when Exim is waiting for an SMTP response from a remote server, Exim
2173 restarts its select() call on the socket, thereby resetting its timeout.
2174 This is not a problem when such interrupts are rare. Somebody set up a cron
2175 job to run exiwhat every 2 minutes, which is less than the normal select()
2176 timeout (5 or 10 minutes). This meant that the select() timeout never
2177 kicked in because it was always reset. I have fixed this by comparing the
2178 time when an interrupt arrives with the time at the start of the first call
2179 to select(). If more time than the timeout has elapsed, the interrupt is
2180 treated as a timeout.
2182 5. Some internal re-factoring in preparation for the addition of Sieve
2183 extensions (by MH). In particular, the "personal" test is moved to a
2184 separate function, and given an option for scanning Cc: and Bcc: (which is
2185 not set for Exim filters).
2187 6. When Exim created an email address using the login of the caller as the
2188 local part (e.g. when creating a From: or Sender: header line), it was not
2189 quoting the local part when it contained special characters such as @.
2191 7. Installed new OpenBSD configuration files.
2193 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2194 try to make them clearer.
2196 9. Callout options, other than the timeout value, were being ignored when
2197 verifying sender addresses in header lines. For example, when using
2199 verify = header_sender/callout=no_cache
2201 the cache was (incorrectly) being used.
2203 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2204 only the Cygwin environment.
2206 11. When return_path_on_delivery was set as a log selector, if different remote
2207 addresses in the same message used different return paths and parallel
2208 remote delivery occurred, the wrong values would sometimes be logged.
2209 (Whenever a remote delivery process finished, the return path value from
2210 the most recently started remote delivery process was logged.)
2212 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2213 lines when AUTH and/or TLS are in use. This is the "received protocol"
2214 field. Exim used to use "asmtp" for authenticated SMTP, without any
2215 indication (in the protocol name) for TLS use. Now it follows the RFC and
2216 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2217 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2218 names appear in log lines as well as in Received: header lines.
2220 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2221 extensions, and comparison tests, and to fix some bugs.
2223 14. Changes to the "personal" filter test:
2225 (1) The test was buggy in that it was just doing the equivalent of
2226 "contains" tests on header lines. For example, if a user's address was
2227 anne@some.where, the "personal" test would incorrectly be true for
2229 To: susanne@some.where
2231 This test is now done by extracting each address from the header in turn,
2232 and checking the entire address. Other tests that are part of "personal"
2233 are now done using regular expressions (for example, to check local parts
2234 of addresses in From: header lines).
2236 (2) The list of non-personal local parts in From: addresses has been
2237 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2238 taken from the Sieve specification recommendations.
2240 (3) If the message contains any header line starting with "List-" it is
2241 treated as non-personal.
2243 (4) The test for "circular" in the Subject: header line has been removed
2244 because it now seems ill-conceived.
2246 15. Minor typos in src/EDITME comments corrected.
2248 16. Installed latest exipick from John Jetmore.
2250 17. If headers_add on a router specified a text string that was too long for
2251 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2252 of string_sprintf() is now avoided.
2254 18. $message_body_size was not set (it was always zero) when running the DATA
2255 ACL and the local_scan() function.
2257 19. For the "mail" command in an Exim filter, no default was being set for
2258 the once_repeat time, causing a random time value to be used if "once" was
2259 specified. (If the value happened to be <= 0, no repeat happened.) The
2260 default is now 0s, meaning "never repeat". The "vacation" command was OK
2261 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2262 (I found it when inspecting the code).
2264 20. There is now an overall timeout for performing a callout verification. It
2265 defaults to 4 times the callout timeout, which applies to individual SMTP
2266 commands during the callout. The overall timeout applies when there is more
2267 than one host that can be tried. The timeout is checked before trying the
2268 next host. This prevents very long delays if there are a large number of
2269 hosts and all are timing out (e.g. when the network connections are timing
2270 out). The value of the overall timeout can be changed by specifying an
2271 additional sub-option for "callout", called "maxwait". For example:
2273 verify = sender/callout=5s,maxwait=20s
2275 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2276 to the end before writing, but this should make it even safer).
2278 22. Exim was forgetting that it had advertised PIPELINING for the second and
2279 subsequent messages on an SMTP connection. It was also not resetting its
2280 memory on STARTTLS and an internal HELO.
2282 23. When Exim logs an SMTP synchronization error within a session, it now
2283 records whether PIPELINING has been advertised or not.
2285 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2286 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2287 rather than long int.
2289 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2291 26. Added the never_mail option to autoreply.
2297 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2298 crash if the getsockname() call failed; this can happen if a connection is
2299 closed very soon after it is established. The problem was simply in the
2300 order in which certain operations were done, causing Exim to try to write
2301 to the SMTP stream before it had set up the file descriptor. The bug has
2302 been fixed by making things happen in the correct order.
2308 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2309 before the connection was closed, thus losing the rejection response.
2311 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2312 some early Solaris releases, but causes trouble in current releases where
2313 socklen_t is defined.
2315 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2318 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2321 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2322 that releasing the top of it at the end releases what was used for sub-
2323 expansions (unless the block got too big). However, discard this block if
2324 the first thing is a variable or header, so that we can use its block when
2325 it is dynamic (useful for very large $message_headers, for example).
2327 6. Lookups now cache *every* query, not just the most recent. A new, separate
2328 store pool is used for this. It can be recovered when all lookup caches are
2329 flushed. Lookups now release memory at the end of their result strings.
2330 This has involved some general refactoring of the lookup sources.
2332 7. Some code has been added to the store_xxx() functions to reduce the amount
2333 of flapping under certain conditions.
2335 8. log_incoming_interface used to affect only the <= reception log lines. Now
2336 it causes the local interface and port to be added to several more SMTP log
2337 lines, for example "SMTP connection from", and rejection lines.
2339 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2341 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2343 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2344 could be overwritten at the end of the current message (or the start of a
2345 new message if it was set in a HELO ACL). The value is now preserved for
2346 the duration of the SMTP connection.
2348 12. If a transport had a headers_rewrite setting, and a matching header line
2349 contained an unqualified address, that address was qualified, even if it
2350 did not match any rewriting rules. The underlying bug was that the values
2351 of the flags that permit the existence of unqualified sender and recipient
2352 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2353 non-local messages, and by -bnq for local messages) were not being
2354 preserved with the message after it was received.
2356 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2357 "next input=" as part of the text comprising the host identity instead of
2358 the correct text. The code was using the same buffer for two different
2359 strings. However, depending on which order the printing function evaluated
2360 its arguments, the bug did not always show up. Under Linux, for example, my
2361 test suite worked just fine.
2363 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2364 doesn't work with some older versions of Perl. It has been changed to "my",
2365 which in any case is probably the better facility to use.
2367 15. A really picky compiler found some instances of statements for creating
2368 error messages that either had too many or two few arguments for the format
2371 16. The size of the buffer for calls to the DNS resolver has been increased
2372 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2373 for addresses that have a lot of PTR records. This alleviates a problem; it
2374 does not fully solve it.
2376 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2377 buffer now truncates the list and logs the incident, which is the same
2378 action as happens when Exim is looking up a host name and its aliases.
2379 Previously in this situation something unpredictable would happen;
2380 sometimes it was "internal error: store_reset failed".
2382 18. If a server dropped the connection unexpectedly when an Exim client was
2383 using GnuTLS and trying to read a response, the client delivery process
2384 crashed while trying to generate an error log message.
2386 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2387 string, for example:
2389 warn message = H1: something\nH2: something
2391 the text was added as a single header line from Exim's point of view
2392 though it ended up OK in the delivered message. However, searching for the
2393 second and subsequent header lines using $h_h2: did not work. This has been
2394 fixed. Similarly, if a system filter added multiple headers in this way,
2395 the routers could not see them.
2397 20. Expanded the error message when iplsearch is called with an invalid key to
2398 suggest using net-iplsearch in a host list.
2400 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2401 ACLs are no longer actually imposed (and a message to that effect is
2404 22. If a "gecos" field in a passwd entry contained escaped characters, in
2405 particular, if it contained a \" sequence, Exim got it wrong when building
2406 a From: or a Sender: header from that name. A second bug also caused
2407 incorrect handling when an unquoted " was present following a character
2408 that needed quoting.
2410 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2411 was not being matched caselessly.
2413 24. Arranged for all hyphens in the exim.8 source to be escaped with
2416 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2417 properly. Recipient callout cache records were still being keyed to include
2418 the sender, even when use_sender was set false. This led to far more
2419 callouts that were necessary. The sender is no longer included in the key
2420 when use_sender is false.
2422 26. Added "control = submission" modifier to ACLs.
2424 27. Added the ${base62d: operator to decode base 62 numbers.
2426 28. dnsdb lookups can now access SRV records.
2428 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2429 the configuration file.
2431 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2432 (-v) mode. This makes the output for a verbose queue run more intelligible.
2434 31. Added a use_postmaster feature to recipient callouts.
2436 32. Added the $body_zerocount variable, containing the number of binary zero
2437 bytes in the message body.
2439 33. The time of last modification of the "new" subdirectory is now used as the
2440 "mailbox time last read" when there is a quota error for a maildir
2443 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2445 35. Added +ignore_unknown as a special item in host lists.
2447 36. Code for decoding IPv6 addresses in host lists is now included, even if
2448 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2449 address was recognized as an IP address, but was then not correctly decoded
2450 into binary, causing unexpected and incorrect effects when compared with
2457 1. Very minor rewording of debugging text in manualroute to say "list of
2458 hosts" instead of "hostlist".
2460 2. If verify=header_syntax was set, and a header line with an unqualified
2461 address (no domain) and a large number of spaces between the end of the
2462 name and the colon was received, the reception process suffered a buffer
2463 overflow, and (when I tested it) crashed. This was caused by some obsolete
2464 code that should have been removed. The fix is to remove it!
2466 3. When running in the test harness, delay a bit after writing a bounce
2467 message to get a bit more predictability in the log output.
2469 4. Added a call to search_tidyup() just before forking a reception process. In
2470 theory, someone could use a lookup in the expansion of smtp_accept_max_
2471 per_host which, without the tidyup, could leave open a database connection.
2473 5. Added the variables $recipient_data and $sender_data which get set from a
2474 lookup success in an ACL "recipients" or "senders" condition, or a router
2475 "senders" option, similar to $domain_data and $local_part_data.
2477 6. Moved the writing of debug_print from before to after the "senders" test
2480 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2481 problems for message scanning, either using a data ACL, or using
2482 local_scan() because the Received: header was not generated till after they
2483 were called (in order to set the time as the time of reception completion).
2484 I have revised the way this works. The header is now generated after the
2485 body is received, but before the ACL or local_scan() are called. After they
2486 are run, the timestamp in the header is updated.
2492 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2493 before starting a queue runner without re-exec. This happened only when
2494 deliver_drop_privilege was set or when the Exim user was set to root. The
2495 effect of the bug was that timeouts during subsequent deliveries caused
2496 crashes instead of being properly handled. The handler is now left at its
2497 default (and expected) setting.
2499 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2500 message, again when deliver_drop_privilege is set or Exim is run as root.
2501 The bug described in (1) was not present in this case, but the tidying up
2502 of the other signals was missing. I have made the two cases consistent.
2504 3. The ignore_target_hosts setting on a manualroute router was being ignored
2505 for hosts that were looked up using the /MX notation.
2507 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2510 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2511 operated on the sender address. After changing the $sender_address to <>
2512 for the sender address verify, Exim was re-instated it as the original
2513 (before rewriting) address, but remembering that it had rewritten it, so it
2514 wasn't rewriting it again. This bug also had the effect of breaking the
2515 sender address verification caching when the sender address was rewritten.
2517 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2518 This has been changed so that if the ip literal address matches
2519 ignore_target_hosts, the router declines.
2521 7. Added expansion conditions match_domain, match_address, and match_local_
2522 part (NOT match_host).
2524 8. The placeholder for the Received: header didn't have a length field set.
2526 9. Added code to Exim itself and to exim_lock to test for a specific race
2527 condition that could lead to file corruption when using MBX delivery. The
2528 issue is with the lockfile that is created in /tmp. If this file is removed
2529 after a process has opened it but before that process has acquired a lock,
2530 there is the potential for a second process to recreate the file and also
2531 acquire a lock. This could lead to two Exim processes writing to the file
2532 at the same time. The added code performs the same test as UW imapd; it
2533 checks after acquiring the lock that its file descriptor still refers to
2534 the same named file.
2536 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2537 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2540 11. Added the smtp_active_hostname option. If used, this will typically be made
2541 to depend on the incoming interface address. Because $interface_address is
2542 not set up until the daemon has forked a reception process, error responses
2543 that can happen earlier (such as "too many connections") no longer contain
2546 12. If an expansion in a condition on a "warn" statement fails because a lookup
2547 defers, the "warn" statement is abandoned, and the next ACL statement is
2548 processed. Previously this caused the whole ACL to be aborted.
2550 13. Added the iplsearch lookup type.
2552 14. Added ident_timeout as a log selector.
2554 15. Added tls_certificate_verified as a log selector.
2556 16. Added a global option tls_require_ciphers (compare the smtp transport
2557 option of the same name). This controls incoming TLS connections.
2559 17. I finally figured out how to make tls_require_ciphers do a similar thing
2560 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2561 before starting the TLS session.
2563 18. Tabs are now shown as \t in -bP output.
2565 19. If the log selector return_path_on_delivery was set, Exim crashed when
2566 bouncing a message because it had too many Received: header lines.
2568 20. If two routers both had headers_remove settings, and the first one included
2569 a superfluous trailing colon, the final name in the first list and the
2570 first name in the second list were incorrectly joined into one item (with a
2571 colon in the middle).
2577 1. Added -C and -D options to the exinext utility, mainly to make it easier
2578 to include in the automated testing, but these could be helpful when
2579 multiple configurations are in use.
2581 2. The exinext utility was not formatting the output nicely when there was
2582 an alternate port involved in the retry record key, nor when there was a
2583 message id as well (for retries that were specific to a specific message
2584 and a specific host). It was also confused by IPv6 addresses, because of
2585 the additional colons they contain. I have fixed the IPv4 problem, and
2586 patched it up to do a reasonable job for IPv6.
2588 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2589 delivery, the log line now contains "pipelined" if PIPELINING was used.
2591 4. An SMTP transport process used to panic and die if the bind() call to set
2592 an explicit outgoing interface failed. This has been changed; it is now
2593 treated in the same way as a connect() failure.
2595 5. A reference to $sender_host_name in the part of a conditional expansion
2596 that was being skipped was still causing a DNS lookup. This no longer
2599 6. The def: expansion condition was not recognizing references to header lines
2600 that used bh_ and bheader_.
2602 7. Added the _cache feature to named lists.
2604 8. The code for checking quota_filecount in the appendfile transport was
2605 allowing one more file than it should have been.
2607 9. For compatibility with Sendmail, the command line option
2615 and sets the incoming protocol and host name (for trusted callers). The
2616 host name and its colon can be omitted when only the protocol is to be set.
2617 Note the Exim already has two private options, -pd and -ps, that refer to
2618 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2619 "s", but I don't think that's a major issue.
2621 10. A number of refactoring changes to the code, none of which should affect
2624 (a) The number of logging options was getting close to filling up the
2625 32-bit word that was used as a bit map. I have split them into two classes:
2626 those that are passed in the argument to log_write(), and those that are
2627 only ever tested independently outside of that function. These are now in
2628 separate 32-bit words, so there is plenty of room for expansion again.
2629 There is no change in the user interface or the logging behaviour.
2631 (b) When building, for example, log lines, the code previously used a
2632 macro that called string_cat() twice, in order to add two strings. This is
2633 not really sufficiently general. Furthermore, there was one instance where
2634 it was actually wrong because one of the argument was used twice, and in
2635 one call a function was used. (As it happened, calling the function twice
2636 did not affect the overall behaviour.) The macro has been replaced by a
2637 function that can join an arbitrary number of extra strings onto a growing
2640 (c) The code for expansion conditions now uses a table and a binary chop
2641 instead of a serial search (which was left over from when there were very
2642 few conditions). Also, it now recognizes conditions like "pam" even when
2643 the relevant support is not compiled in: a suitably worded error message is
2644 given if an attempt is made to use such a condition.
2646 11. Added ${time_interval:xxxxx}.
2648 12. A bug was causing one of the ddress fields not to be passed back correctly
2649 from remote delivery subprocesses. The field in question was not being
2650 subsequently used, so this caused to problems in practice.
2652 13. Added new log selectors queue_time and deliver_time.
2654 14. Might have fixed a bug in maildirsizefile handling that threw up
2655 "unexpected character" debug warnings, and recalculated the data
2656 unnecessarily. In any case, I expanded the warning message to give more
2659 15. Added the message "Restricted characters in address" to the statements in
2660 the default ACL that block characters like @ and % in local parts.
2662 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2663 Three changes have been made:
2665 (a) There was a serious bug; a negative response to MAIL caused the whole
2666 recipient domain to be cached as invalid, thereby blocking all messages
2667 to all local parts at the same domain, from all senders. This bug has
2668 been fixed. The domain is no longer cached after a negative response to
2669 MAIL if the sender used is not empty.
2671 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2674 (c) A new callout option, "use_sender" has been added for people who want
2675 the modified behaviour.
2681 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2684 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2685 indeed breaks things for older releases.
2687 3. Added additional logging to the case where there is a problem reading data
2688 from a filter that is running in a subprocess using a pipe, in order to
2689 try to track down a specific problem.
2691 4. Testing facility fudge: when running in the test harness and attempting
2692 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2693 getting "No route to host". Convert this to a timeout.
2695 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2698 6. Some OS don't have socklen_t but use size_t instead. This affects the
2699 fifth argument of getsockopt() amongst other things. This is now
2700 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2701 can be set for individual OS. I have set it for SunOS5, OSF1, and
2702 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2703 some earlier ones do not.
2705 7. Change 4.30/15 was not doing the test caselessly.
2707 8. The standard form for an IPv6 address literal was being rejected by address
2708 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2709 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2710 this, as well as the form without the "IPv6" on the front (but only when
2711 address literals are enabled, of course).
2713 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2715 10. Exim crashed if a message with an empty sender address specified by -f
2716 encountered a router with an errors_to setting. This could be provoked only
2717 by a command such as
2721 where an empty string was supplied; "<>" did not hit this bug.
2723 11. Installed PCRE release 4.5.
2725 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2726 remained set. It is now erased.
2728 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2729 times from message ids (which are base 36 rather than the normal 62).
2731 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2732 were being counted as actual protocol errors, and logged if the log
2733 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2734 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2735 and DATA following a set of rejected RCPTs do not count as protocol errors.
2736 In other words, Exim assumes they were pipelined, though this may not
2737 actually be the case. Of course, in all cases the client gets an
2738 appropriate error code.
2740 15. If a lookup fails in an ACL condition, a message about the failure may
2741 be available; it is used if testing the ACL cannot continue, because most
2742 such messages specify what the cause of the deferral is. However, some
2743 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2744 that caused an old message to be retained and used if a later statement
2745 caused a defer, replacing the real cause of the deferral.
2747 16. If an IP address had so many PTR records that the DNS lookup buffer
2748 was not large enough to hold them, Exim could crash while trying to process
2749 the truncated data. It now detects and logs this case.
2751 17. Further to 4.21/58, another change has been made: if (and only if) the
2752 first line of a message (the first header line) ends with CRLF, a bare LF
2753 in a subsequent header line has a space inserted after it, so as not to
2754 terminate the header.
2756 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2757 unnecessarily, used atoi() instead of strtol(), and didn't check the
2758 termination when getting file sizes from file names by regex.
2760 19. Completely re-implemented the support for maildirsize files, in the light
2761 of a number of problems with the previous contributed implementation
2762 (4.30/29). In particular:
2764 . If the quota is zero, the maildirsize file is maintained, but no quota is
2767 . If the maildir directory does not exist, it is created before any attempt
2768 to write a maildirsize file.
2770 . The quota value in the file is just a cache; if the quota is changed in
2771 the transport, the new value overrides.
2773 . A regular expression is available for excluding directories from the
2776 20. The autoreply transport checks the characters in options that define the
2777 message's headers; it allows continued headers, but it was checking with
2778 isspace() after an embedded newline instead of explicitly looking for a
2781 21. If all the "regular" hosts to which an address was routed had passed their
2782 expiry times, and had not reached their retry times, the address was
2783 bounced, even if fallback hosts were defined. Now Exim should go on to try
2786 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2787 equivalent code in the SMTP transport. Some hosts send humungous responses
2788 to HELO/EHLO, more than 1024 it seems.
2790 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2791 gives compiler warnings in some environments. I've now done it "properly",
2794 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2795 (because of problems with the built-in one) was declared to return uschar *
2796 instead of char *, causing compiler failure.
2798 25. Fixed a file descriptor leak when processing alias/forward files.
2800 26. Fixed a minor format string issue in dbfn.c.
2802 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2804 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2805 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2808 29. When a sender address is verified, it is cached, to save repeating the test
2809 when there is more than one recipient in a message. However, when the
2810 verification involves a callout, it is possible for different callout
2811 options to be set for different recipients. It is too complicated to keep
2812 track of this in the cache, so now Exim always runs a verification when a
2813 callout is required, relying on the callout cache for the optimization.
2814 The overhead is duplication of the address routing, but this should not be
2817 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2818 to be verified with callout=postmaster, and the main callout worked but the
2819 postmaster check failed, the verification correctly failed. However, if a
2820 subsequent RCPT command asked for sender verification *without* the
2821 postmaster check, incorrect caching caused this verification also to fail,
2824 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2825 it was not caching the DNS options (qualify_single, search_parents) that
2826 were used when the lookup failed. A subsequent lookup with different
2827 options therefore always gave the same answer, though there were cases
2828 where it should not have. (Example: a "domains = !$mx_any" option on a
2829 dnslookup router: the "domains" option is always processed without any
2830 widening, but the router might have qualify_single set.) Now Exim uses the
2831 cached value only when the same options are set.
2833 32. Added John Jetmore's "exipick" utility to the distribution.
2835 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2836 than a timeout (e.g. a certificate is required, and is not provided), an
2837 Exim server now closes the connection immediately. Previously it waited for
2838 the client to close - but if the client is SSL, it seems that they each
2839 wait for each other, leading to a delay before one of them times out.
2841 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2842 maintained 0.8.x compatibility because I don't think many are using it, and
2843 it is clearly obsolete.
2845 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2848 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2849 client certificate was expired. A simple patch fixes this, though I don't
2850 understand the full logic of why the verify callback is called multiple
2853 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2854 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2855 which causes problems with some clients (such as the Certicom SSL Plus
2856 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2857 disables the coutermeasure allowing Eudora to connect."
2859 38. Exim was not checking that a write() to a log file succeeded. This could
2860 lead to Bad Things if a log got too big, in particular if it hit a file
2861 size limit. Exim now panics and dies if it cannot write to a log file, just
2862 as it does if it cannot open a log file.
2864 39. Modified OS/Makefile-Linux so that it now contains
2866 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2868 The two -D definitions ensure that Exim is compiled with large file
2869 support, which makes it possible to handle log files that are bigger than
2872 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2873 instance) a domain was checked against a named list that involved a lookup,
2874 causing $domain_data to be set, then another domain was checked against the
2875 same list, then the first domain was re-checked, the value of $domain_data
2876 after the final check could be wrong. In particular, if the second check
2877 failed, it could be set empty. This bug probably also applied to
2880 41. The strip_trailing_dot option was not being applied to the address given
2881 with the -f command-line option.
2883 42. The code for reading a message's header from the spool was incrementing
2884 $received_count, but never initializing it. This meant that the value was
2885 incorrect (doubled) while delivering a message in the same process in which
2886 it was received. In the most common configuration of Exim, this never
2887 happens - a fresh exec is done - but it can happen when
2888 deliver_drop_privilege is set.
2890 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2891 it now includes up to 150 characters of the unexpected data in the log
2894 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2895 and building data strings. The size of both of these buffers was 10 000
2896 bytes - far larger than anybody would *ever* want, thought I. Needless to
2897 say, somebody hit the limit. I have increased the maximum line length to
2898 20 000 and the maximum data length of concatenated lines to 100 000. I have
2899 also fixed two bugs, because there was no checking on these buffers. Tsk,
2900 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2901 buffer is too small.
2903 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2904 lsearch lookups. Now it does.
2906 46. When parsing a route_list item in a manualroute router, a fixed-length
2907 buffer was used for the list of hosts. I made this 1024 bytes long,
2908 thinking that nobody would ever have a list of hosts that long. Wrong.
2909 Somebody had a whole pile of complicated expansion conditions, and the
2910 string was silently truncated, leading to an expansion error. It turns out
2911 that it is easier to change to an unlimited length (owing to other changes
2912 that have happened since this code was originally written) than to build
2913 structure for giving a limitation error. The length of the item that
2914 expands into the list of hosts is now unlimited.
2916 47. The lsearch lookup could not handle data where the length of text line was
2917 more than 4095 characters. Such lines were truncated, leading to shortened
2918 data being returned. It should now handle lines of any length.
2920 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2921 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2924 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2925 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2926 no longer happens in comment lines. A list of replacements is now placed
2927 at the head of all of the source files, except those whose only change is
2928 to replace PERL_COMMAND in the very first #! line.
2930 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2931 messages on the queue, with a bottom-up merge sort, using code contributed
2932 by Michael Haardt. This should make operations like -bp somewhat faster on
2933 large queues. It won't affect queue runners, except when queue_run_in_order
2936 51. Installed eximstats 1.31 in the distribution.
2938 52. Added support for SRV lookups to the dnslookup router.
2940 53. If an ACL referred to $message_body or $message_body_end, the value was not
2941 reset for any messages that followed in the same SMTP session.
2943 54. The store-handling optimization for building very long strings was not
2944 differentiating between the different store pools. I don't think this
2945 actually made any difference in practice, but I've tidied it.
2947 55. While running the routers to verify a sender address, $sender_address
2948 was still set to the sender address. This is wrong, because when routing to
2949 send a bounce to the sender, it would be empty. Therefore, I have changed
2950 it so that, while verifying a sender address, $sender_address is set to <>.
2951 (There is no change to what happens when verifying a recipient address.)
2953 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2954 target A or AAAA records (if not already returned) without resetting the
2955 qualify_single or search_parents options of the DNS resolver. These are
2956 inappropriate in this case because the targets of MX and SRV records must
2957 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2958 target that, when qualified, matched something in the local domain. These
2959 two options are now turned off when doing these lookups.
2961 57. It seems that at least some releases of Reiserfs (which does not have the
2962 concept of a fixed number of inodes) returns zero and not -1 for the
2963 number of available inodes. This interacted badly with check_spool_inodes,
2964 which assumed that -1 was the "no such thing" setting. What I have done is
2965 to check that the total number of inodes is greater than zero before doing
2966 the test of how many are available.
2968 58. When a "warn" ACL statement has a log_message modifier, the message is
2969 remembered, and not repeated. This is to avoid a lot of repetition when a
2970 message has many recipients that cause the same warning to be written.
2971 Howewer, Exim was preserving the list of already written lines for an
2972 entire SMTP session, which doesn't seem right. The memory is now reset if a
2973 new message is started.
2975 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2976 debugging output unless log_rewrite was also set.
2978 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2979 of (int)(handle) when we know that handle contains (void *)(-1).
2981 61. The Exim daemon panic-logs an error return when it closes the incoming
2982 connection. However "connection reset by peer" seems to be common, and
2983 isn't really an error worthy of noting specially, so that particular error
2986 62. When Exim is trying to find all the local interfaces, it used to panic and
2987 die if the ioctl to get the interface flags failed. However, it seems that
2988 on at least one OS (Solaris 9) it is possible to have an interface that is
2989 included in the list of interfaces, but for which you get a failure error
2990 for this call. This happens when the interface is not "plumbed" into a
2991 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2992 failure of the "get flags" call assumes that the interface is down.
2994 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2995 makes life easier for people who are doing arithmetic on fields extracted
2996 from dates, where you often get leading zeros that should not be
2997 interpreted as octal.
2999 64. Added qualify_domain to the redirect router, to override the global
3002 65. If a pathologically long header line contained very many addresses (the
3003 report of this problem mentioned 10 000) and each of them was rewritten,
3004 Exim could use up a very large amount of memory. (It kept on making new
3005 copies of the header line as it rewrote, and never released the old ones.)
3006 At the expense of a bit more processing, the header rewriting function has
3007 been changed so that it no longer eats memory in this way.
3009 66. The generation of the Received: header has been moved from the time that a
3010 message starts to be received, to the time that it finishes. The timestamp
3011 in the Received: header should now be very close to that of the <= log
3012 line. There are two side-effects of this change:
3014 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3015 logged header lines no longer include the local Received: line, because
3016 it has not yet been created. The same applies to a copy of the message
3017 that is returned to a non-SMTP sender when a message is rejected.
3019 (b) When a filter file is tested using -bf, no additional Received: header
3020 is added to the test message. After some thought, I decided that this
3023 This change does not affect the value of $received_for. It is still set
3024 after address rewriting, but before local_scan() is called.
3026 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3028 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3029 gave an unhelpful panic error message, and a defer error. I have managed to
3030 change this behaviour so that it now rejects any supplied certificate,
3031 which seems right, as the list of acceptable certificates is empty.
3033 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3034 gave an unhelpful defer error. I have not managed to make this reject any
3035 supplied certificates, but the error message it gives is "no certificate
3036 supplied", which is not helpful.
3038 70. exigrep's output now also includes lines that are not associated with any
3039 message, but which match the given pattern. Implemented by a patch from
3040 Martin Sluka, which also tidied up the Perl a bit.
3042 71. Recipient callout verification, like sender verification, was using <> in
3043 the MAIL FROM command. This isn't really the right thing, since the actual
3044 sender may affect whether the remote host accepts the recipient or not. I
3045 have changed it to use the actual sender in the callout; this means that
3046 the cache record is now keyed on a recipient/sender pair, not just the
3047 recipient address. There doesn't seem to be a real danger of callout loops,
3048 since a callout by the remote host to check the sender would use <>.
3049 [SEE ABOVE: changed after hitting problems.]
3051 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3052 temporary errors. However, in the case of such a code being given after
3053 the end of a data transmission (i.e. after ".") Exim was failing to write
3054 a retry record for the message. (Yes, there was some broken host that was
3055 actually sending 8xx at this point.)
3057 73. An unknown lookup type in a host list could cause Exim to panic-die when
3058 the list was checked. (An example that provoked this was putting <; in the
3059 middle of a list instead of at the start.) If this happened during a DATA
3060 ACL check, a -D file could be left lying around. This kind of configuration
3061 error no longer causes Exim to die; instead it causes a defer errror. The
3062 incident is still logged to the main and panic logs.
3064 74. Buglet left over from Exim 3 conversion. The message "too many messages
3065 in one connection" was written to the rejectlog but not the mainlog, except
3066 when address rewriting (yes!) was being logged.
3068 75. Added write_rejectlog option.
3070 76. When a system filter was run not as root (that is, when system_filter_user
3071 was set), the values of the $n variables were not being returned to the
3072 main process; thus, they were not subsequently available in the $sn
3075 77. Added +return_path_on_delivery log selector.
3077 78. A connection timeout was being treated differently from recipients deferred
3078 when testing hosts_max_try with a message that was older than the host's
3079 retry timeout. (The host should not be counted, thus allowing all hosts to
3080 be tried at least once before bouncing.) This may have been the cause of an
3081 occasionally reported bug whereby a message would remain on the queue
3082 longer than the retry timeout, but would be bounced if a delivery was
3083 forced. I say "may" because I never totally pinned down the problem;
3084 setting up timeout/retry tests is difficult. See also the next item.
3086 79. The ultimate address timeout was not being applied to errors that involved
3087 a combination of host plus message (for example, a timeout on a MAIL
3088 command). When an address resolved to a number of possible hosts, and they
3089 were not all tried for each delivery (e.g. because of hosts_max_try), a
3090 message could remain on the queue longer than the retry timeout.
3092 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3095 81. Fixed an obscure SMTP outgoing bug which required at least the following
3096 conditions: (a) there was another message waiting for the same server;
3097 (b) the server returned 5xx to all RCPT commands in the first message so
3098 that the message was not completed; (c) the server dropped the connection
3099 or gave a negative response to the RSET that Exim sends to abort the
3100 transaction. The observed case was a dropped connection after DATA that had
3101 been sent in pipelining mode. That is, the server had advertised PIPELINING
3102 but was not implementing it correctly. The effect of the bug was incorrect
3103 behaviour, such as trying another host, and this could lead to a crash.
3109 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3110 and daemon.c were passed as pointers to ints; they should have been
3111 pointers to socklen_t variables (which are typically unsigned ints).
3113 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3116 3. Fixed a really odd bug that affected only the testing scheme; patching a
3117 certain fixed string in the binary changed the value of another string that
3118 happened to be identical to the end of the original first string.
3120 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3121 name", it returns that address as the IP address. On some operating
3122 systems (e.g. Solaris), it also passes back the IP address string as the
3123 "host name". However, on others (e.g. Linux), it passes back an empty
3124 string. Exim wasn't checking for this, and was changing the host name to an
3125 empty string, assuming it had been canonicized.
3127 5. Although rare, it is permitted to have more than one PTR record for a given
3128 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3129 all the names associated with an address, because they do in Solaris.
3130 However, it seems that they do not in Linux for data that comes from the
3131 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3132 I found this out when I moved to a new Linux workstation and tried to run
3133 the Exim test suite.
3135 To get round this problem I have changed the code so that it now does its
3136 own call to the DNS to look up PTR records when searching for a host name.
3137 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3138 addresses that are only in /etc/hosts are still found.
3140 This behaviour is, however, controlled by an option called host_lookup_
3141 order, which defaults to "bydns:byaddr". If people want to use the other
3142 order, or indeed, just use one or the other means of lookup, they can
3143 specify it in this variable.
3145 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3146 some operating systems, this comes back from gethostbyaddr() as an empty
3147 string, and this is what Exim used to test for. However, it seems that in
3148 other systems, "." is yielded. Exim now tests for this case too.
3150 7. The values of check_spool_space and check_log_space are now held internally
3151 as a number of kilobytes instead of an absolute number of bytes. If a
3152 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3153 kilobyte. This means that much larger values can be stored.
3155 8. Exim monitor: an attempt to get the action menu when not actually pointing
3156 at a message produces an empty menu entitled "No message selected". This
3157 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3158 no entries in it ("Shell widget menu has zero width and/or height"). So I
3159 have added a single, blank menu entry in this case.
3161 9. Added ${quote_local_part.
3163 10. MIME decoding is now applied to the contents of Subject: header lines when
3166 11. Now that a reference to $sender_host_address automatically causes a reverse
3167 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3168 host lookup before query-style lookups in lists that might use this
3169 variable. This has therefore been abolished, and the "net-" prefix is no
3170 longer necessary for query-style lookups.
3172 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3173 have been a typo for LFLAGS, so it has been changed.
3175 13. The install script calls Exim with "-C /dev/null" in order to find the
3176 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3177 to be output. Howeve, since Exim outputs its version number before the
3178 error, it didn't break the script. It just looked ugly. I fixed this by
3179 always allowing "-C /dev/null" if the caller is root.
3181 14. Ignore overlarge ACL variable number when reading spool file - insurance
3182 against a later release with more variables having written the file.
3184 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3185 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3186 this, as well as the form without the "IPv6" on the front.
3188 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3189 OS/Makefile-Darwin file.
3191 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3192 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3194 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3195 variable. However, some OS use ints for inodes. I've added cast to long int
3196 to get rid of the compiler warning.
3198 19. I had forgotten to lock out "/../" in configuration file names when
3199 ALT_CONFIG_PREFIX was set.
3201 20. Routers used for verification do not need to specify transports. However,
3202 if such a router generated a host list, and callout was configured, Exim
3203 crashed, because it could not find a port number from the (non-existent)
3204 transport. It now assumes port 25 in this circumstance.
3206 21. Added the -t option to exigrep.
3208 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3209 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3210 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3211 them is set, LOOKUP_LSEARCH is forced.
3213 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3215 24. Added sender and host information to the "rejected by local_scan()" log
3216 line; previously there was no indication of these.
3218 25. Added .include_if_exists.
3220 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3221 Linux. It turns out that not all file systems support this. Apparently some
3222 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3223 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3224 not supported on the file descriptor, is now ignored when Exim is trying to
3225 sync a directory. This applies only to Linux.
3227 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3229 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3230 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3233 29. Added support for maildirsize files from supplied patch (modified a bit).
3235 30. The use of :fail: followed by an empty string could lead Exim to respond to
3236 sender verification failures with (e.g.):
3238 550 Verification failed for <xxx>
3239 550 Sender verify failed
3241 where the first response line was missing the '-' that indicates it is not
3242 the final line of the response.
3244 31. The loop for finding the name of the user that called Exim had a hardwired
3245 limit of 10; it now uses the value of finduser_retries, which is used for
3246 all other user lookups.
3248 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3251 33. Exim was neglecting to zero errno before one call of strtol() when
3252 expanding a string and expecting an integer value. On some systems this
3253 resulted in spurious "integer overflow" errors. Also, it was casting the
3254 result into an int without checking.
3256 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3257 did not work. The code looks as if it has *never* worked, though it appears
3258 to have been documented since at least releast 1.62. I have made it work.
3260 35. The "timeout_DNS" error in retry rules, also documented since at least
3261 1.62, also never worked. As it isn't clear exactly what this means, and
3262 clearly it isn't a major issue, I have abolished the feature by treating it
3263 as "timeout", and writing a warning to the main and panic logs.
3265 36. The display of retry rules for -brt wasn't always showing the error code
3268 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3269 timeout_connect_A, timeout_connect_MX.
3271 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3272 to the empty sender.
3274 39. The daemon was not analysing the content of -oX till after it had closed
3275 stderr and disconnected from the controlling terminal. This meant that any
3276 syntax errors were only noted on the panic log, and the return code from
3277 the command was 0. By re-arranging the code a little, I've made the
3278 decoding happen first, so such errors now appear on stderr, and the return
3279 code is 1. However, the actual setting up of the sockets still happens in
3280 the disconnected process, so errors there are still only recorded on the
3283 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3284 connections (as happens on some IP stacks) was logged at start up time as
3285 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3286 it from "IPv6 and IPv4", which means that two separate sockets are being
3289 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3290 says whether AF_INET or AF_INET6 was passed as an argument.
3292 42. Exiwhat output was messed up when time zones were included in log
3295 43. Exiwhat now gives more information about the daemon's listening ports,
3296 and whether -tls-on-connect was used.
3298 44. The "port" option of the smtp transport is now expanded.
3300 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3301 silently ignored. Now an error message is written to the main and panic
3304 46. There's a new ACL modifier called "logwrite" which writes to a log file
3305 as soon as it is encountered.
3307 47. Added $local_user_uid and $local_user_gid at routing time.
3309 48. Exim crashed when trying to verify a sender address that was being
3312 49. Exim was recognizing only a space character after ".include". It now also
3313 recognizes a tab character.
3315 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3316 extracting the relevant information from the specification. The man page no
3317 longer contains scrambled data for the -d option, and I've added a section
3318 at the front about calling Exim under different names.
3320 51. Added "extra_headers" argument to the "mail" command in filter files.
3322 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3325 53. Installed eximstats 1.29.
3327 54. Added transport_filter_timeout as a generic transport option.
3329 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3330 Cc: header lines. This was required by RFC 822, but it not required by RFC
3333 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3334 incoming messages that did not have them. Now it does so only if the
3335 message originates locally, that is, if there is no associated remote host
3336 address. When Resent- header lines are present, this applies to the Resent-
3337 lines rather than the non-Resent- lines.
3339 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3340 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3342 58. Messages for configuration errors now include the name of the main
3343 configuration file - useful now that there may be more than one file in a
3344 list (.included file names were always shown).
3346 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3347 for those rare installations that do not start the daemon as root or run it
3348 setuid root. I've cut out the call to initgroups() if the daemon is not
3351 60. The Exim user and group can now be bound into the binary as text strings
3352 that are looked up at the start of Exim's processing.
3354 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3356 62. Added $mailstore_basename variable.
3358 63. Installed patch to sieve.c from Michael Haardt.
3360 64. When Exim failed to open the panic log after failing to open the main log,
3361 the original message it was trying to log was written to stderr and debug
3362 output, but if they were not available (the usual case in production), it
3363 was lost. Now it is written to syslog before the two lines that record the
3364 failures to open the logs.
3366 65. Users' Exim filters run in subprocesses under the user's uid. It is
3367 possible for a "deliver" command or an alias in a "personal" command to
3368 provoke an address rewrite. If logging of address rewriting is configured,
3369 this fails because the process is not running as root or exim. There may be
3370 a better way of dealing with this, but for the moment (because 4.30 needs
3371 to be released), I have disabled address rewrite logging when running a
3372 filter in a non-root, non-exim process.
3378 1. The buildconfig auxiliary program wasn't quoting the value set for
3379 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3380 not defined. This bug was masked in 4.22 by the effect that was fixed in
3383 2. Some messages that were rejected after a message id was allocated were
3384 shown as "incomplete" by exigrep. It no longer does this for messages that
3385 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3387 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3388 have allow_domain_literals set, the ID did not get logged in the <= line.
3389 Domain literals are now always recognized in Message-ID: header lines.
3391 4. The first argument for a ${extract expansion item is the key name or field
3392 number. Leading and trailing spaces in this item were not being ignored,
3393 causing some misleading effects.
3395 5. When deliver_drop_privilege was set, single queue runner processes started
3396 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3397 same command in the process it spins off) were not dropping privilege.
3399 6. When the daemon running as "exim" started a queue runner, it always
3400 re-executed Exim in the spun-off process. This is a waste of effort when
3401 deliver_drop_privilege is set. The new process now just calls the
3402 queue-runner function directly.
3408 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3411 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3412 ignored. Though the use of -oP was forcing the writing of a pid file, it
3413 was always written to the default place.
3415 3. If the message "no IP address found for host xxxx" is generated during
3416 incoming verification, it is now followed by identification of the incoming
3417 connection (so you can more easily find what provoked it).
3419 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3421 5. Added some features to "harden" Exim a bit more against certain attacks:
3423 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3424 be put in Local/Makefile. This is like the never_users runtime option,
3425 but it cannot be overridden. The default setting is "root".
3427 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3428 prefix string with which any file named in a -C command line option
3431 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3432 is retained for -C and -D only if the caller of Exim is root. Without
3433 it, the exim user may also use -C and -D and retain privilege.
3435 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3436 command line option is disabled.
3438 6. Macro names set by the -D option must start with an upper case letter, just
3439 like macro names defined in the configuration file.
3441 7. Added "dereference=" facility to LDAP.
3443 8. Two instances of the typo "uknown" in the source files are fixed.
3445 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3446 the Configure-Makefile script screwed up while processing it.
3448 10. Incorporated PCRE 4.4.
3450 11. The SMTP synchronization check was not operating right at the start of an
3451 SMTP session. For example, it could not catch a HELO sent before the client
3452 waited for the greeting. There is now a check for outstanding input at the
3453 point when the greeting is written. Because of the duplex, asynchronous
3454 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3455 way, but not yet received, when the check is performed.
3457 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3458 on TCP/IP sockets, because this apparently causes some broken clients to
3461 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3462 unchanged) from the Cygwin maintainer.
3464 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3465 was supported instead of testing for mbx. Effectively a typo.
3467 15. The spa authenticator server code was not checking that the input it
3468 received was valid base64.
3470 16. The debug output line for the "set" modifier in ACLs was not showing the
3471 name of the variable that was being set.
3473 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3475 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3476 lookup on its own. Something else was needed to trigger the lookup. For
3477 example, a match in host_lookup or the need for a host name in a host list.
3478 Now, if $sender_host_name is referenced and the host name has not yet been
3479 looked up, a lookup is performed. If the lookup fails, the variable remains
3480 empty, and $host_lookup_failed is set to "1".
3482 19. Added "eqi" as a case-independent comparison operator.
3484 20. The saslauthd authentication condition could segfault if neither service
3485 nor realm was specified.
3487 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3488 error message that was logged was misleading, and incoming SMTP
3489 connections were dropped. The message is now more accurate, and temporary
3490 errors are given to SMTP connections.
3492 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3493 (except RSET) with a 421 error, until QUIT is received. However, it was
3494 failing to send a response to QUIT.
3496 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3497 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3498 ACL is now run after the helo_try_verify_hosts code.
3500 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3501 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3502 case-independent, so other case variants are also recognized). Apparently
3503 some systems use these upper case variants.
3505 25. If more than two messages were waiting for the same host, and a transport
3506 filter was specified for the transport, Exim sent two messages over the
3507 same TCP/IP connection, and then failed with "socket operation on non-
3508 socket" when it tried to send the third.
3510 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3512 27. The extern definition of crypt16() in expand.c was not being excluded when
3513 the OS had its own crypt16() function.
3515 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3516 as a preferred synonym for return_size_limit, both as an option and as an
3519 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3521 30. Changed the default configuration ACL to relax the local part checking rule
3522 for addresses that are not in any local domains. For these addresses,
3523 slashes and pipe symbols are allowed within local parts, but the sequence
3524 /../ is explicitly forbidden.
3526 31. SPA server authentication was not clearing the challenge buffer before
3529 32. log_message in a "warn" ACL statement was writing to the reject log as
3530 well as to the main log, which contradicts the documentation and doesn't
3531 seem right (because no rejection is happening). So I have stopped it.
3533 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3534 However, I am unable to do any testing of this.
3536 34. Fixed an infelicity in the appendfile transport. When checking directories
3537 for a mailbox, to see if any needed to be created, it was accidentally
3538 using path names with one or more superfluous leading slashes; tracing
3539 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3541 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3542 contents are added to the log line that is written for every discarded
3543 recipient. (Previously a log_message setting was ignored.)
3545 36. The ${quote: operator now quotes the string if it is empty.
3547 37. The install script runs exim in order to find its version number. If for
3548 some reason other than non-existence or emptiness, which it checks, it
3549 could not run './exim', it was installing it with an empty version number,
3550 i.e. as "exim-". This error state is now caught, and the installation is
3553 38. An argument was missing from the function that creates an error message
3554 when Exim fails to connect to the socket for saslauthd authentication.
3555 This could cause Exim to crash, or give a corrupted message.
3557 39. Added isip, isip4, and isip6 to ${if conditions.
3559 40. The ACL variables $acl_xx are now saved with the message, and can be
3560 accessed later in routers, transports, and filters.
3562 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3563 strings in the file are not string-expanded.
3565 42. If a MAIL command specified a SIZE value that was too large to fit into an
3566 int variable, the check against message_size_limit failed. Such values are
3567 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3568 day this will have to be increased, but I don't think I want to be around
3569 when emails are that large.
3576 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3577 iconv() is not standard in FreeBSD.
3579 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3580 IPv6 enabled. The observed symptom was a segmentation fault on return from
3581 the function os_common_find_running_interfaces() in src/os.c.
3583 3. In the check_special_case() function in daemon.c I had used "errno" as an
3584 argument name, which causes warnings on some systems. This was basically a
3585 typo, since it was named "eno" in the comments!
3587 4. The code that waits for the clock to tick (at a resolution of some fraction
3588 of a second) so as to ensure message-id uniqueness was always waiting for
3589 at least one whole tick, when it could have waited for less. [This is
3590 almost certainly not relevant at current processor speeds, where it is
3591 unlikely to ever wait at all. But we try to future-proof.]
3593 5. The function that sleeps for a time interval that includes fractions of a
3594 second contained a race. It did not block SIGALRM between setting the
3595 timer, and suspending (a couple of lines later). If the interval was short
3596 and the sigsuspend() was delayed until after it had expired, the suspension
3597 never ended. On busy systems this could lead to processes getting stuck for
3600 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3601 process, before it forks any delivery processes. The open lookup caching
3602 mechanism meant that the open file or database connection was passed into
3603 the delivery process. The problem was that delivery processes always tidy
3604 up cached lookup data. This could cause a problem for the next delivery
3605 process started by the queue runner, because the external queue runner
3606 process does not know about the closure. So the next delivery process
3607 still has data in the lookup cache. In the case of a file lookup, there was
3608 no problem because closing a file descriptor in a subprocess doesn't affect
3609 the parent. However, if the lookup was caching a connection to a database,
3610 the connection was closed, and the second delivery process was likely to
3611 see errors such as "PGSQL: query failed: server closed the connection
3612 unexpectedly". The problem has been fixed by closing all cached lookups
3613 in a queue runner before running a delivery process.
3615 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3616 seem to have the "const" qualifier which it has on other OS. I've
3619 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3620 *of the same type* (client or server) with the same public name that an
3621 error should be diagnosed.
3623 9. When Exim looked up a host name for an IP address, but failed to find the
3624 original IP address when looking up the host name (a safety check), it
3625 output the message "<ip address> does not match any IP for NULL", which was
3626 confusing, to say the least. The bug was that the host name should have
3627 appeared instead of "NULL".
3629 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3630 user that is built into the binary, and the -C or -D options is used, root
3631 privilege is dropped before the configuration file is read. In addition,
3632 logging is switched to stderr instead of the normal log files. If the
3633 configuration then re-defines the Exim user, the unprivileged environment
3634 is probably not what is expected, so Exim logs a panic warning message (but
3637 However, if deliver_drop_privilege is set, the unprivileged state may well
3638 be exactly what is intended, so the warning has been cut out in that case,
3639 and Exim is allowed to try to write to its normal log files.
3645 1. smtp_return_error_details was not giving details for temporary sender
3646 or receiver verification errors.
3648 2. Diagnose a configuration error if two authenticators have the same public
3651 3. Exim used not to create the message log file for a message until the first
3652 delivery attempt. This could be confusing when incoming messages were held
3653 for policy or load reasons. The message log file is now created at the time
3654 the message is received, and an initial "Received" line is written to it.
3656 4. The automatically generated man page for command line options had a minor
3657 bug that caused no ill effects; however, a more serious problem was that
3658 the procedure for building the man page automatically didn't always
3659 operate. Consequently, release 4.20 contains an out-of-date version. This
3660 shouldn't happen again.
3662 5. When building Exim with embedded Perl support, the script that builds the
3663 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3664 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3666 6. The freeze_tell option was not being used for messages that were frozen on
3667 arrival, either by an ACL or by local_scan().
3669 7. Added the smtp_incomplete_transaction log selector.
3671 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3672 was accepting AUTH without a new EHLO.
3674 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3675 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3678 10. Logging of TCP/IP connections (when configured) now happens in the main
3679 daemon process instead of the child process, so that the TCP/IP connection
3680 count is more accurate (but it can never be perfect).
3682 11. The use of "drop" in a nested ACL was not being handled correctly in the
3683 outer ACL. Now, if condition failure induced by the nested "drop" causes
3684 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3685 or "require"), the connection is dropped.
3687 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3688 that yield "discard" can now be used with an "accept" or a "discard" verb,
3689 but an error is generated for any others (because I can't see a useful way
3690 to define what should happen).
3692 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3693 are now processed in the same way as lines in the Exim configuration file.
3694 In particular, continuation lines are supported.
3696 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3698 15. Added -ti meaning -t -i.
3700 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3701 domains, and warn by logging if others are found.
3703 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3704 returned from GIFCONF when Exim is trying to find the list of interfaces on
3705 a host. The code in os.c has been modified to copy each ifreq to an aligned
3706 structure in all cases.
3708 Also, in some cases, the returned ifreq's were being copied to a 'struct
3709 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3710 means the last couple of bytes of an IPv6 address could be chopped if the
3711 ifreq contained only a normal sockaddr (14 bytes storage).
3713 18. Named domain lists were not supported in the hosts_treat_as_local option.
3714 An entry such as +xxxx was not recognized, and was treated as a literal
3717 19. Ensure that header lines added by a DATA ACL are included in the reject log
3718 if the ACL subsequently rejects the message.
3720 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3721 MD5 (which is deprecated).
3723 21. When testing a filter file using -bf, Exim was writing a message when it
3724 took the sender from a "From " line in the message, but it was not doing so
3725 when it took $return_path from a Return-Path: header line. It now does.
3727 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3728 with a valid header line field name (a series of printing characters
3729 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3731 23. Changed "disc" in the source to "disk" to conform to the documentation and
3732 the book and for uniformity.
3734 24. Ignore Sendmail's -Ooption=value command line item.
3736 25. When execve() failed while trying to run a command in a pipe transport,
3737 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3738 could be confused with a return value of 69 from the command itself. This
3739 has been changed to 127, the value the shell returns if it is asked to run
3740 a non-existent command. The wording for the related log line suggests a
3741 non-existent command as the problem.
3743 26. If received_header_text expands to an empty string, do not add a Received:
3744 header line to the message. (Well, it adds a token one on the spool, but
3745 marks it "old" so that it doesn't get used or transmitted.)
3747 27. Installed eximstats 1.28 (addition of -nt option).
3749 28. There was no check for failure on the call to getsockname() in the daemon
3750 code. This can fail if there is a shortage of resources on the system, with
3751 ENOMEM, for example. A temporary error is now given on failure.
3753 29. Contrary to the C standard, it seems that in some environments, the
3754 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3755 program. Exim now does this explicitly; it affects the formatting of
3756 timestamps using strftime().
3758 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3759 complaints. I've now initialized all the variables, to avoid this.
3761 32. Header lines added by a system filter were not being "seen" during
3762 transport-time rewrites.
3764 33. The info_callback() function passed to OpenSSL is set up with type void
3765 (*)(SSL *, int, int), as described somewhere. However, when calling the
3766 function (actually a macro) that sets it up, the type void(*)() is
3767 expected. I've put in a cast to prevent warnings from picky compilers.
3769 34. If a DNS black list lookup found a CNAME record, but there were no A
3770 records associated with the domain it pointed at, Exim crashed.
3772 35. If a DNS black list lookup returned more than one A record, Exim ignored
3773 all but the first. It now scans all returned addresses if a particular IP
3774 value is being sought. In this situation, the contents of the
3775 $dnslist_value variable are a list of all the addresses, separated by a
3778 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3779 to accept a host name and all its aliases if the forward lookup for any of
3780 them yielded the IP address of the incoming connection. Now it accepts only
3781 those names whose forward lookup yields the correct IP address. Any other
3782 names are discarded. This closes a loophole whereby a rogue DNS
3783 administrator could create reverse DNS records to break through a
3784 wildcarded host restriction in an ACL.
3786 37. If a user filter or a system filter that ran in a subprocess used any of
3787 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3788 the wrong values were passed to the pipe command ($thisaddress had the
3789 value of $0, $0 had the value of $1, etc). This bug was introduced by
3790 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3792 38. Improved the line breaking for long SMTP error messages from ACLs.
3793 Previously, if there was no break point between 40 and 75 characters, Exim
3794 left the rest of the message alone. Two changes have been made: (a) I've
3795 reduced the minimum length to 35 characters; (b) if it can't find a break
3796 point between 35 and 75 characters, it looks ahead and uses the first one
3797 that it finds. This may give the occasional overlong line, but at least the
3798 remaining text gets split now.
3800 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3801 file descriptors might be low, and that setting 1000 would always raise it.
3802 It turns out that in some environments, the limit is already over 1000 and
3803 that lowering it causes trouble. So now Exim takes care not to decrease it.
3805 40. When delivering a message, the value of $return_path is set to $sender_
3806 address at the start of routing (routers may change the value). By an
3807 oversight, this default was not being set up when an address was tested by
3808 -bt or -bv, which affected the outcome if any router or filter referred to
3811 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3812 log only when "message" or "log_message" are set. However, if one of the
3813 conditions was an address verification, or a call to a nested ACL, the
3814 messages generated by the underlying test were being passed through. This
3815 no longer happens. The underlying message is available in $acl_verify_
3816 message for both "message" and "log_message" expansions, so it can be
3817 passed through if needed.
3819 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3820 new expansion $bh_ to give the encoded byte string without charset
3821 translation. Translation happens only if iconv() is available; HAVE_ICONV
3822 indicates this at build time. HEADERS_CHARSET gives the charset to
3823 translate to; headers_charset can change it in the configuration, and
3824 "headers charset" can change it in an individual filter file.
3826 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3827 that creates RFC 2047 encoded "words" labels them as that charset instead
3828 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3829 expansion operator; (ii) when Exim creates a From: line for a local
3830 message; (iii) when a header line is rewritten to include a "phrase" part.
3832 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3833 buggy, causing it to skip the first lines of messages whose message ID
3834 ended in 'D'. This would not have bitten before Exim release 4.14, because
3835 message IDs were unlikely to end in 'D' before then. The effect was to have
3836 incorrect size information for certain domains.
3838 45. #include "config.h" was missing at the start of the crypt16.c module. This
3839 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3842 46. If there was a timeout during a "random" callout check, Exim treated it as
3843 a failure of the random address, and carried on sending RSET and the real
3844 address. If the delay was just some slowness somewhere, the response to the
3845 original RCPT would be taken as a response to RSET and so on, causing
3846 mayhem of various kinds.
3848 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3849 when I implemented it. It didn't allow for the fact that some option values
3850 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3851 the right test for positive values.
3853 48. Domain names in DNS records are case-independent. Exim always looks them up
3854 in lower case. Some resolvers return domain names in exactly the case they
3855 appear in the zone file, that is, they may contain uppercase letters. Not
3856 all resolvers do this - some return always lower case. Exim was treating a
3857 change of case by a resolver as a change of domain, similar to a widening
3858 of a domain abbreviation. This triggered its re-routing code and so it was
3859 trying to route what was effectively the same domain again. This normally
3860 caused routing to fail (because the router wouldn't handle the domain
3861 twice). Now Exim checks for this case specially, and just changes the
3862 casing of the domain that it ultimately uses when it transmits the message
3865 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3868 50. If a filter generated a file delivery with a non-absolute name (possible if
3869 no home directory exists for the router), the forbid_file option was not
3872 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3873 the existing equality matching.
3875 52. Exim was using ints instead of ino_t variables in some places where it was
3876 dealing with inode numbers.
3878 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3879 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3880 called TMPDIR, and if it finds it is different, it changes its value.
3882 54. The smtp_printf() function is now made available to local_scan() so
3883 additional output lines can be written before returning. There is also an
3884 smtp_fflush() function to enable the detection of a dropped connection.
3885 The variables smtp_input and smtp_batched_input are exported to
3888 55. Changed the default runtime configuration: the message "Unknown user"
3889 has been removed from the ACL, and instead placed on the localuser router,
3890 using the cannot_route_message feature. This means that any verification
3891 failures that generate their own messages won't get overridden. Similarly,
3892 the "Unrouteable address" message that was in the ACL for unverifiable
3893 relay addresses has also been removed.
3895 56. Added hosts_avoid_esmtp to the smtp transport.
3897 57. The exicyclog script was not checking for the esoteric option
3898 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3899 will work only if exicyclog is run under the appropriate euid.
3901 58. Following a discussion on the list, the rules by which Exim recognises line
3902 endings on incoming messages have been changed. The -dropcr and drop_cr
3903 options are now no-ops, retained only for backwards compatibility. The
3904 following line terminators are recognized: LF CRLF CR. However, special
3905 processing applies to CR:
3907 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3908 nor a local message in the state where . is a terminator.
3910 (ii) If a bare CR is encountered in a header line, an extra space is added
3911 after the line terminator so as not to end the header. The reasoning
3912 behind this is that bare CRs in header lines are most likely either
3913 to be mistakes, or people trying to play silly games.
3915 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3916 was being incorrectly given as 18 bytes larger than it should have been.
3917 This is a VOB (very old bug).
3919 60. This may never have affected anything current, but just in case it has:
3920 When the local host is found other than at the start of a list of hosts,
3921 the local host, those with the same MX, and any that follow, are discarded.
3922 When the list in question was part of a longer list of hosts, the following
3923 hosts (not currently being processed) were also being discarded. This no
3924 longer happens. I'm not sure if this situation could ever has previously
3927 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3930 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3931 header. This is something that is recommended in a new Internet Draft, and
3932 is something that is documented as being done by Sendmail. There are two
3933 possible values. For messages generated by the autoreply transport, Exim
3936 Auto-Submitted: auto-replied
3938 whereas for all other generated messages (e.g. bounces) it adds
3940 Auto-Submitted: auto-generated
3942 63. The "personal" condition in filters now includes a test for the
3943 Auto-Submitted: header. If it contains the string "auto-" the message it
3944 not considered personal.
3946 64. Added rcpt_include_affixes as a generic transport option.
3948 65. Added queue_only_override (default true).
3950 66. Added the syslog_duplication option.
3952 67. If what should have been the first header line of a message consisted of
3953 a space followed by a colon, Exim was mis-interpreting it as a header line.
3954 It isn't of course - it is syntactically invalid and should therefore be
3955 treated as the start of the message body. The misbehaviour could have
3956 caused a number of strange effects, including loss of data in subsequent
3957 header lines, and spool format errors.
3959 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3960 client host had authenticated. This control can now be exercised by an ACL
3961 for more flexibility.
3963 69. By default, callouts do not happen when testing with -bh. There is now a
3964 variant, -bhc, which does actually run the callout code, including
3965 consulting and updating the callout cache.
3967 70. Added support for saslauthd authentication, courtesy of Alexander
3970 71. If statvfs() failed on the spool or log directories while checking their
3971 size for availability, Exim confusingly gave the error "space shortage".
3972 Furthermore, in debugging mode it crashed with a floating point exception.
3973 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3974 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3975 serious problem, Exim now writes to the main and panic logs when this
3976 happens, with details of the failure. It then refuses to accept the
3977 incoming message, giving the message "spool directory problem" or "log
3978 directory problem" with a 421 code for SMTP messages.
3980 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3981 0, 1, and 2 exist, because some OS complain for execs without them (see
3982 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3983 descriptors. However, the code omitted to check that the open succeeded,
3984 causing mysterious errors if for some reason the permissions on /dev/null
3985 got screwed. Now Exim writes a message to the main and panic logs, and
3986 bombs out if it can't open /dev/null.
3988 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3989 interact so that it is all more flexible. It is supposed to remain
3990 backwards compatible. Also added extra_local_interfaces.
3992 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3993 to bomb out with an assertion failure - to the client this appears as a
3994 connection drop. This problem occurs in the part of the code that was taken
3995 from the Samba project. Fortunately, the assertion is in a very simple
3996 function, so I have fixed this by reproducing the function inline in the
3997 one place where it is called, and arranging for authentication to fail
3998 instead of killing the process with assert().
4000 75. The SPA client code was not working when the server requested OEM rather
4001 than Unicode encoding.
4003 76. Added code to make require_files with a specific uid setting more usable in
4004 the case where statting the file as root fails - usually a non-root-mounted
4005 NFS file system. When this happens and the failure is EACCES, Exim now
4006 forks a subprocess and does the per-uid checking as the relevant uid.
4008 77. Added process_log_path.
4010 78. If log_file_path was not explicitly set, a setting of check_log_space or
4011 check_log_inodes was ignored.
4013 79. If a space check for the spool or log partitions fails, the incident is now
4014 logged. Of course, in the latter case the data may get lost...
4016 80. Added the %p formatting code to string_format() so that it can be used to
4017 print addresses in debug_print(). Adjusted all the address printing in the
4018 debugging in store.c to use %p rather than %d.
4020 81. There was a concern that a line of code in smtp_in.c could overflow a
4021 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4022 initially expressed, the concern was not well-founded, because trailing
4023 spaces are removed early. However, if the trailing spaces were followed by
4024 a NULL, they did not get removed, so the overflow was possible. Two fixes
4027 (a) I re-wrote the offending code in a cleaner fashion.
4028 (b) If an incoming SMTP command contains a NULL character, it is rejected
4031 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4032 runs initgroups(), so that if the Exim user is in any additional groups,
4033 they will be used during message reception.
4039 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob