1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.367 2006/07/03 15:39:06 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
64 PH/08 When testing addresses using -bt, indicate those final addresses that
65 are duplicates that would not cause an additional delivery. At least one
66 person was confused, thinking that -bt output corresponded to deliveries.
67 (Suppressing duplicates isn't a good idea as you lose the information
68 about possibly different redirections that led to the duplicates.)
70 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
71 systems where poll() doesn't work, in particular OS X.
73 PH/10 Added more information to debugging output for retry time not reached.
79 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
80 other effects) broke the use of negated acl sub-conditions.
82 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
85 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
86 "Deny" causes Exim to reject the incoming connection with a 554 error.
87 Unfortunately, if there is a major crisis, such as a disk failure,
88 tcp-wrappers gives "deny", whereas what one would like would be some
89 kind of temporary error. A kludge has been added to help with this.
90 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
91 554 error is used if errno is still zero or contains ENOENT (which occurs
92 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
95 PH/03 Add -lutil to the default FreeBSD LIBS setting.
97 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
98 errors. Otherwise a message that provokes a temporary error (when other
99 messages do not) can cause a whole host to time out.
101 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
102 addresses were routed directly to files or pipes from a redirect router.
103 File deliveries just didn't batch; pipe deliveries might have suffered
106 PH/06 A failure to get a lock for a hints database would erroneously always say
107 "Failed to get write lock", even when it was really a read lock.
109 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
110 of 0600. This has been changed to use the value of the lockfile_mode
111 option (which defaults to 0600).
113 PH/08 Applied small patch from the Sieve maintainer.
115 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
116 folder from quota calculations, a direct delivery into this folder messed
117 up the contents of the maildirsize file. This was because the regex was
118 used only to exclude .Trash (or whatever) when the size of the mailbox
119 was calculated. There was no check that a delivery was happening into an
120 excluded directory. This bug has been fixed by ignoring all quota
121 processing for deliveries into excluded directories.
123 PH/10 Added the maildirfolder_create_regex option to appendfile.
129 PH/01 The code for finding all the local interface addresses on a FreeBSD
130 system running IPv6 was broken. This may well have applied to all BSD
131 systems, as well as to others that have similar system calls. The broken
132 code found IPv4 interfaces correctly, but gave incorrect values for the
133 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
134 that it would not match correctly against @[] and not recognize the IPv6
137 PH/02 The ipliteral router was not recognizing addresses of the form user@
138 [ipv6:....] because it didn't know about the "ipv6:" prefix.
140 PH/03 Added disable_ipv6.
142 PH/04 Changed $reply_address to use the raw form of the headers instead of the
143 decoded form, because it is most often used to construct To: headers
144 lines in autoreplies, and the decoded form may well be syntactically
145 invalid. However, $reply_address has leading white space removed, and all
146 newlines turned into spaces so that the autoreply transport does not
149 PH/05 If group was specified without a user on a router, and no group or user
150 was specified on a transport, the group from the router was ignored.
152 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
153 for visible compile-time settings that can be used to change these
154 numbers, for those that want even more. Backwards compatibility with old
155 spool files has been maintained. However, going back to a previous Exim
156 release will lost any variables that are in spool files.
158 PH/07 Two small changes when running in the test harness: increase delay when
159 passing a TCP/IP connection to a new process, in case the original
160 process has to generate a bounce, and remove special handling of
161 127.0.0.2 (sic), which is no longer necessary.
163 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
164 be the same on different OS.
166 PH/09 Moved a debug statement in filter processing to avoid a race problem when
169 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
170 whether --show-vars was specified or not
172 JJ/02 exipick: Added support for new ACL variable spool format introduced
175 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
176 syntactically invalid From: or Reply-to: line, and a filter used this to
177 generate an autoreply, and therefore failed to obtain an address for the
178 autoreply, Exim could try to deliver to a non-existent relative file
179 name, causing unrelated and misleading errors. What now happens is that
180 it logs this as a hard delivery error, but does not attempt to create a
183 PH/11 The exinext utility has a -C option for testing purposes, but although
184 the given file was scanned by exinext itself; it wasn't being passed on
185 when Exim was called.
187 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
188 an end-of-file indication when reading a command response.
190 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
191 compiled. In many other places in Exim, IPv6 addresses are always
192 recognized, so I have changed this. It also means that IPv4 domain
193 literals of the form [IPV4:n.n.n.n] are now always recognized.
195 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
196 used if the router is not running as root, for example, when verifying at
197 ACL time, or when using -bh. The debugging output from this situation was
198 non-existent - all you got was a failure to exec. I have made two
201 (a) Failures to set uid/gid, the current directory, or a process leader
202 in a subprocess such as that created by queryprogram now generate
203 suitable debugging ouput when -d is set.
205 (b) The queryprogram router detects when it is not running as root,
206 outputs suitable debugging information if -d is set, and then runs
207 the subprocess without attempting to change uid/gid.
209 PH/15 Minor change to Makefile for building test_host (undocumented testing
212 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
213 additional section of a DNS packet that returns MX or SRV records.
214 Instead, it always explicitly searches for A/AAAA records. This avoids
215 major problems that occur when a DNS server includes only records of one
216 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
217 fixed another bug: if SRV records were looked up and the corresponding
218 address records were *not* found in the additional section, the port
219 values from the SRV records were lost.
221 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
222 using the correct key (the original address) when searching the retry
223 rules in order to find which one to use for generating the retry hint.
225 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
226 adding the default one. Similarly, if it contains a Reply-To: header, the
227 errors_reply_to option, if set, is not used.
229 PH/19 When calculating a retry time, Exim used to measure the "time since
230 failure" by looking at the "first failed" field in the retry record. Now
231 it does not use this if it is later than than the arrival time of the
232 message. Instead it uses the arrival time. This makes for better
233 behaviour in cases where some deliveries succeed, thus re-setting the
234 "first failed" field. An example is a quota failure for a huge message
235 when small messages continue to be delivered. Without this change, the
236 "time since failure" will always be short, possible causing more frequent
237 delivery attempts for the huge message than are intended.
238 [Note: This change was subsequently modified - see PH/04 for 4.62.]
240 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
241 $1, $2, $3) because the numerical variables can be reset during some
242 expansion items (e.g. "match"), thereby losing the authentication data.
244 PH/21 Make -bV show the size of off_t variables so that the test suite can
245 decide whether to run tests for quotas > 2G.
247 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
248 mailbox_size, and mailbox_filecount in the appendfile transport. If a
249 filecount value is greater than 2G or if a quota value is greater than 2G
250 on a system where the size of off_t is not greater than 4, a panic error
253 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
254 never match. The debug and -bh output now contains an explicit error
255 message indicating a malformed IPv4 address or mask.
257 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
258 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
261 PH/25 Do not write to syslog when running in the test harness. The only
262 occasion when this arises is a failure to open the main or panic logs
263 (for which there is an explicit test).
265 PH/26 Added the /no_tell option to "control=freeze".
267 PH/27 If a host name lookup failed very early in a connection, for example, if
268 the IP address matched host_lookup and the reverse lookup yielded a name
269 that did not have a forward lookup, an error message of the form "no IP
270 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
271 could be logged. Now it outputs the IP address instead of "NULL".
273 PH/28 An enabling patch from MH: add new function child_open_exim2() which
274 allows the sender and the authenticated sender to be set when
275 submitting a message from within Exim. Since child_open_exim() is
276 documented for local_scan(), the new function should be too.
278 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
279 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
280 results in an empty string is now treated as unset.
282 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
284 PH/31 Added sender_verify_fail logging option.
286 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
287 needed by RFC 822 but not by RFC 2822 was commented out. I have now
288 tidied the source and removed it altogether.
290 PH/33 When a queue run was abandoned because the load average was too high, a
291 log line was always written; now it is written only if the queue_run log
292 selector is set. In addition, the log line for abandonment now contains
293 information about the queue run such as the pid. This is always present
294 in "start" and "stop" lines but was omitted from the "abandon" line.
296 PH/34 Omit spaces between a header name and the colon in the error message that
297 is given when verify = headers_syntax fails (if there are lots of them,
298 the message gets confusing).
300 PH/35 Change the default for dns_check_names_pattern to allow slashes within
301 names, as there are now some PTR records that contain slashes. This check
302 is only to protect against broken name servers that fall over on strange
303 characters, so the fact that it applies to all lookups doesn't matter.
305 PH/36 Now that the new test suite is complete, we can remove some of the
306 special code in Exim that was needed for the old test suite. For example,
307 sorting DNS records because real resolvers return them in an arbitrary
308 order. The new test suite's fake resolver always returns records in the
311 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
312 bounces) except when queue_only is set, to avoid logging races between
313 the different processes.
315 PH/38 Panic-die if .include specifies a non-absolute path.
317 PH/39 A tweak to the "H" retry rule from its user.
319 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
320 a label. They prevented compilation on older perls.
322 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
323 a warning to be raised on newish perls.
325 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
326 on queue. Changes to match documented behaviour of showing count of
327 messages matching specified criteria.
329 PH/40 Changed the default ident timeout from 30s to 5s.
331 PH/41 Added support for the use of login_cap features, on those BSD systems
332 that have them, for controlling the resources used by pipe deliveries.
334 PH/42 The content-scanning code uses fopen() to create files in which to put
335 message data. Previously it was not paying any attention to the mode of
336 the files. Exim runs with umask(0) because the rest of the code creates
337 files with open(), and sets the required mode explicitly. Thus, these
338 files were ending up world-writeable. This was not a big issue, because,
339 being within the spool directory, they were not world-accessible. I have
340 created a function called modefopen, which takes an additional mode
341 argument. It sets umask(777), creates the file, chmods it to the required
342 mode, then resets the umask. All the relevant calls to fopen() in the
343 content scanning code have been changed to use this function.
345 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
346 to 24 hours. This avoids potential overflow problems when processing G
347 and H retry rules. I suspect nobody ever tinkers with this value.
349 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
351 PH/45 When the plaintext authenticator is running as a client, the server's
352 challenges are checked to ensure they are valid base64 strings. By
353 default, the authentication attempt is cancelled if an invalid string is
354 received. Setting client_ignore_invalid_base64 true ignores these errors.
355 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
356 they are received. Thus, the responses can be made to depend on the
357 challenges. If an invalid string is ignored, an empty string is placed in
360 PH/46 Messages that are created by the autoreply transport now contains a
361 References: header, in accordance with RFCs 2822 and 3834.
363 PH/47 Added authenticated_sender_force to the smtp transport.
365 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
367 PH/49 Installed latest patch from the Sieve maintainer.
369 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
370 was also set, the appendfile transport was unnecessarily scanning a
371 directory of message files (e.g. for maildir delivery) to find the count
372 of files (along with the size), even though it did not need this
373 information. It now does the scan only if it needs to find either the
374 size of the count of files.
376 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
378 PH/52 Two bugs concerned with error handling when the smtp transport is
381 (i) Exim was not creating retry information for temporary errors given
382 for individual recipients after the DATA command when the smtp transport
383 was used in LMTP mode. This meant that they could be retried too
384 frequently, and not timed out correctly.
386 (ii) Exim was setting the flag that allows error details to be returned
387 for LMTP errors on RCPT commands, but not for LMTP errors for individual
388 recipients that were returned after the DATA command.
390 PH/53 This is related to PH/52, but is more general: for any failing address,
391 when detailed error information was permitted to be returned to the
392 sender, but the error was temporary, then after the final timeout, only
393 "retry timeout exceeded" was returned. Now it returns the full error as
394 well as "retry timeout exceeded".
396 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
397 do this, and (what is worse) MTAs that accept it.
399 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
400 will now be deprecated.
402 PH/56 New os.c-cygwin from the Cygwin maintainer.
404 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
405 formats (previously only available in exim formats via -bpr, -bpru,
406 and -bpra. Now also available in native and exiqgrep formats)
408 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
409 with very large, slow to parse queues
411 JJ/08 exipick: added ! as generic prefix to negate any criteria format
413 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
415 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
416 responses to authentication challenges, though it was showing the
417 challenges; (ii) I've removed the CR characters from the debug output for
420 PH/58 Allow for the insertion of a newline as well as a space when a string
421 is turned into more than one encoded-word during RFC 2047 encoding. The
422 Sieve code now uses this.
424 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
425 data_4xx, lost_connection, tls_required.
427 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
428 message was being sent as an SMTP response.
430 PH/61 Add -l and -k options to exicyclog.
432 PH/62 When verifying, if an address was redirected to one new address, so that
433 verification continued, and the new address failed or deferred after
434 having set something in $address_data, the value of $address_data was not
435 passed back to the ACL. This was different to the case when no
436 redirection occurred. The value is now passed back in both cases.
438 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
439 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
440 use login_cap.h, so on its own it isn't the distinguishing feature. The
441 new name refers directly to the setclassresources() function.
443 PH/65 Added configuration files for NetBSD3.
445 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
447 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
448 is preferred over IPv4.
450 PH/68 The bounce_return_message and bounce_return_body options were not being
451 honoured for bounces generated during the reception of non-SMTP messages.
452 In particular, this applied to messages rejected by the ACL. This bug has
453 been fixed. However, if bounce_return_message is true and bounce_return_
454 body is false, the headers that are returned for a non-SMTP message
455 include only those that have been read before the error was detected.
456 (In the case of an ACL rejection, they have all been read.)
458 PH/69 The HTML version of the specification is now built in a directory called
459 spec_html instead of spec.html, because the latter looks like a path with
460 a MIME-type, and this confuses some software.
462 PH/70 Catch two compiler warnings in sieve.c.
464 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
465 function verify_get_ident() calls ip_connect() to connect a socket, but
466 if the "connect()" function timed out, ip_connect() used to close the
467 socket. However, verify_get_ident() also closes the socket later, and in
468 between Exim writes to the log, which may get opened at this point. When
469 the socket was closed in ip_connect(), the log could get the same file
470 descriptor number as the socket. This naturally causes chaos. The fix is
471 not to close the socket in ip_connect(); the socket should be closed by
472 the function that creates it. There was only one place in the code where
473 this was missing, in the iplookup router, which I don't think anybody now
474 uses, but I've fixed it anyway.
476 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
477 well as to direct DNS lookups. Otherwise the handling of names in host
478 lists is inconsistent and therefore confusing.
484 PH/01 Two changes to the default runtime configuration:
486 (1) Move the checks for relay_from_hosts and authenticated clients from
487 after to before the (commented out) DNS black list checks.
489 (2) Add control=submission to the relay_from_hosts and authenticated
490 clients checks, on the grounds that messages accepted by these
491 statements are most likely to be submissions.
493 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
495 (1) Generate an error if the third argument for the ${prvs expansion is
498 (2) Treat a missing third argument of ${prvscheck as if it were an empty
501 (3) Reset the variables that are obtained from the first argument of
502 ${prvscheck and used in the second argument before leaving the code,
503 because their memory is reclaimed, so using them afterwards may do
506 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
507 one (it's much easier than Tom thought :-).
509 (5) Because of (4), we can now allow for the use of $prvscheck_result
510 inside the third argument.
512 PH/03 For some reason, the default setting of PATH when running a command from
513 a pipe transport was just "/usr/bin". I have changed it to
516 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
517 anything to be listed in the output from -bV.
519 PH/05 When a filter generated an autoreply, the entire To: header line was
520 quoted in the delivery log line, like this:
522 => >A.N.Other <ano@some.domain> <original@ddress> ...
524 This has been changed so that it extracts the operative address. There
525 may be more than one such address. If so, they are comma-separated, like
528 => >ano@some.domain,ona@other.domain <original@ddress> ...
530 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
531 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
532 not being looked up in the rDNS to get a host name, Exim was showing the
533 IP address twice in Received: lines, even though the IP addresses were
534 identical. For example:
536 Received: from [1.2.3.4] (helo=[1.2.3.4])
538 However, if the real host name was known, it was omitting the HELO data
539 if it matched the actual IP address. This has been tidied up so that it
540 doesn't show the same IP address twice.
542 PH/07 When both +timestamp and +memory debugging was on, the value given by
543 $tod_xxx expansions could be wrong, because the tod_stamp() function was
544 called by the debug printing, thereby overwriting the timestamp buffer.
545 Debugging no longer uses the tod_stamp() function when +timestamp is set.
547 PH/08 When the original message was included in an autoreply transport, it
548 always said "this is a copy of the message, including all the headers",
549 even if body_only or headers_only was set. It now gives an appropriate
552 PH/09 Applied a patch from the Sieve maintainer which:
554 o fixes some comments
555 o adds the (disabled) notify extension core
556 o adds some debug output for the result of if/elsif tests
557 o points to the current vacation draft in the documentation
558 and documents the missing references header update
562 o fixes a bug in processing the envelope test (when testing
563 multiple envelope elements, the last element determinted the
566 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
567 Electronic Mail") by including:
569 Auto-submitted: auto-generated
571 in the messages that it generates (bounce messages and others, such as
572 warnings). In the case of bounce messages for non-SMTP mesages, there was
573 also a typo: it was using "Auto_submitted" (underscore instead of
574 hyphen). Since every message generated by Exim is necessarily in response
575 to another message, thes have all been changed to:
577 Auto-Submitted: auto-replied
579 in accordance with these statements in the RFC:
581 The auto-replied keyword:
583 - SHOULD be used on messages sent in direct response to another
584 message by an automatic process,
586 - MUST NOT be used on manually-generated messages,
588 - MAY be used on Delivery Status Notifications (DSNs) and Message
589 Disposition Notifications (MDNs),
591 - MUST NOT be used on messages generated by automatic or periodic
592 processes, except for messages which are automatic responses to
595 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
596 to the default Received: header definition.
598 PH/12 Added log selector acl_warn_skipped (default on).
600 PH/13 After a successful wildlsearch lookup, discard the values of numeric
601 variables because (a) they are in the wrong storage pool and (b) even if
602 they were copied, it wouldn't work properly because of the caching.
604 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
605 checking when decoding. Apparently there are clients that generate
606 overlong encoded strings. Why am I not surprised?
608 PH/15 If the first argument of "${if match_address" was not empty, but did not
609 contain an "@" character, Exim crashed. Now it writes a panic log message
610 and treats the condition as false.
612 PH/16 In autoreply, treat an empty string for "once" the same as unset.
614 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
615 extension "envelope-auth". The code is finished and in agreement with
616 other implementations, but there is no documentation so far and in fact,
617 nobody wrote the draft yet. This extension is currently #undef'ed, thus
618 not changing the active code.
620 Print executed "if" and "elsif" statements when debugging is used. This
621 helps a great deal to understand what a filter does.
623 Document more things not specified clearly in RFC3028. I had all this
624 sorted out, when out of a sudden new issues came to my mind. Oops."
626 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
629 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
630 canonical form (as documented). However, after a host name lookup from
631 the IP address, check_host() was doing a simple string comparison with
632 addresses acquired from the DNS when checking that the found name did
633 have the original IP as one of its addresses. Since any found IPv6
634 addresses are likely to be in abbreviated form, the comparison could
635 fail. Luckily, there already exists a function for doing the comparison
636 by converting both addresses to binary, so now that is used instead of
639 PH/20 There was another similar case to PH/19, when a complete host name was
640 given in a host list; looking up its IP address could give an abbreviated
641 form, whereas the current host's name might or might not be abbreviated.
642 The same fix has been applied.
648 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
649 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
652 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
653 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
655 PH/03 Typo: missing ".o" in src/pcre/Makefile.
657 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
658 header line, restrict the check to what is listed in RFCs 2369 and 2929.
659 Also, for "Auto-Submitted", treat anything other than "no" as
660 non-personal, in accordance with RFC 3834. (Previously it treated
661 anything starting "auto-" as non-personal.)
663 TF/01 The control=submission/name=... option had a problem with syntax
664 errors if the name included a slash character. The /name= option
665 now slurps the rest of the string, so it can include any characters
666 but it must come last in the list of options (after /sender_retain
669 PH/05 Some modifications to the interface to the fake nameserver for the new
677 TK/01 Added the "success_on_redirect" address verification option. See
678 NewStuff for rationale and an example.
680 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
682 PH/02 Patch to exigrep to allow it to work on syslog lines.
684 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
685 fread() to skip over the body file's header line, because in Cygwin the
686 header line is locked and is inaccessible.
688 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
689 co-exist for some time) to make it clear that it is the Exim ID that is
690 referenced, not the Message-ID: header line.
692 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
693 string_format() function, because snprintf() does not exist on all
696 PH/06 The use of forbid_filter_existstest now also locks out the use of the
697 ${stat: expansion item.
699 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
700 protocol synchronization error", to keep the pedants happy.
702 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
703 well as for IRIX systems, when gcc is being used. See the host.c source
706 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
708 PH/10 Named domain lists were not working if used in a queue_smtp_domains
711 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
712 transport and to the smtp transport in LMTP mode.
714 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
716 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
717 run a filter in a subprocess. This could lead to confusion in subsequent
718 lookups in the parent process. There should also be a search_tidyup() at
719 the end of the subprocess.
721 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
722 only if the host matched helo_try_verify_hosts, which caused the
723 verification to occur when the EHLO/HELO command was issued. The ACL just
724 tested the remembered result. Now, if a previous verification attempt has
725 not happened, "verify = helo" does it there and then.
727 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
729 TK/03 Fix log output including CR from clamd.
731 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
732 exist provoked a memory error which could cause a segfault.
734 PH/15 Installed PCRE 6.2
736 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
738 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
739 of the problem. Specifically, suggested +O2 rather than +O1 for the
742 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
744 PH/20 If a delivery was routed to a non-standard port by means of an SRV
745 record, the port was not correctly logged when the outgoing_port log
746 selector was set (it logged the transort's default port).
748 PH/21 Added support for host-specific ports to manualroute, queryprogram,
749 fallback_hosts, and "hosts" in the smtp transport.
751 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
752 host errors such as "Connection refused".
754 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
755 authentication with radiusclient 0.4.9:
757 - Error returned from rc_read_config was caught wrongly
758 - Username/password not passed on to radius server due to wrong length.
760 The presumption is that some radiusclient API changes for 4.51/PH/17
761 were not taken care of correctly. The code is still untested by me (my
762 Linux distribution still has 0.3.2 of radiusclient), but it was
763 contributed by a Radius user.
765 PH/24 When doing a callout, the value of $domain wasn't set correctly when
766 expanding the "port" option of the smtp transport.
768 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
769 while reading a MIME header. Thanks to Tom Hughes for a patch.
771 PH/24 Include config.h inside local_scan.h so that configuration settings are
774 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
775 that in an ACL for RCPT (for example), you can examine exactly what was
778 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
779 commands, but it was not correctly comparing the address with the actual
780 client host address. Thus, it would show the EHLO address in Received:
781 header lines when this was not necessary.
783 PH/27 Added the % operator to ${eval:}.
785 PH/28 Exim tries to create and chdir to its spool directory when it starts;
786 it should be ignoring failures (because with -C, for example, it has lost
787 privilege). It wasn't ignoring creation failures other than "already
790 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
793 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
794 because an input line was too long, either on its own, or by virtue of
795 too many continuations, the temporary file was not being removed, and the
796 return code was incorrect.
798 PH/31 Missing "BOOL" in function definition in filtertest.c.
800 PH/32 Applied Sieve patches from the maintainer.
802 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
804 PH/33 Added "verify = not_blind".
806 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
807 Local/Makefile (with some defaults set). These are used in built scripts
808 such as exicyclog, but they have never been used in the exim_install
809 script (though there are many overriding facilities there). I have
810 arranged that the exim_install script now takes note of these two
813 PH/35 Installed configuration files for Dragonfly.
815 PH/36 When a locally submitted message by a trusted user did not contain a
816 From: header, and the sender address was obtained from -f or from an SMTP
817 MAIL command, and the trusted user did not use -F to supply a sender
818 name, $originator_name was incorrectly used when constructing a From:
819 header. Furthermore, $originator_name was used for submission mode
820 messages from external hosts without From: headers in a similar way,
821 which is clearly wrong.
823 PH/37 Added control=suppress_local_fixups.
825 PH/38 When log_selector = +received_sender was set, and the addition of the
826 sender made the log line's construction buffer exactly full, or one byte
827 less than full, an overflow happened when the terminating "\n" was
830 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
831 when the result of a list match is failure because a DNS lookup failed.
833 PH/40 RM_COMMAND is now used in the building process.
835 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
836 the "build-* directories that it finds.
838 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
839 address in a domain literal was a prefix of an interface address.
841 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
842 when verifying a sender address, unless rewrite_headers is false.
844 PH/44 Wrote a long comment about why errors_to addresses are verified as
845 recipients, not senders.
847 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
848 the ratelimit ACL was added.
850 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
852 PH/46 Added extra information about PostgreSQL errors to the error string.
854 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
855 avoiding the need to install special zones in a real server. This is
856 backwards compatible; if it can't find the fake resolver, it drops back.
857 Thus, both old and new test suites can be run.
859 TF/02 Added util/ratelimit.pl
861 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
862 clock is set back in time.
864 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
865 Candler <B.Candler@pobox.com>.
867 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
868 OK for addresses that are the result of redirections.
870 PH/48 A number of further additions for the benefit of the new test suite,
871 including a fake gethostbyname() that interfaces to the fake DNS resolver
874 TF/06 The fix for widen_domains has also been applied to qualify_single and
875 search_parents which are the other dnslookup options that can cause
878 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
881 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
883 TF/07 Exim produced the error message "an SRV record indicated no SMTP
884 service" if it encountered an MX record with an empty target hostname.
885 The message is now "an MX or SRV record indicated no SMTP service".
887 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
888 if the DNS of the sending site is misconfigured. This is quite a
889 common situation. This change restores the behaviour of treating a
890 helo verification defer as a failure.
892 PH/51 If self=fail was set on a router, the bounce message did not include the
893 actual error message.
899 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
901 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
902 command itself ended in error, the underlying message about the transport
903 filter timeout was being overwritten with the pipe command error. Now the
904 underlying error message should be appended to the second error message.
906 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
908 PH/02 Reduce the amount of output that "make" produces by default. Full output
909 can still be requested.
911 PH/03 The warning log line about a condition test deferring for a "warn" verb
912 was being output only once per connection, rather than after each
913 occurrence (because it was using the same function as for successful
914 "warn" verbs). This seems wrong, so I have changed it.
916 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
917 it should not have, which might have caused a crash in the right
918 circumstances, but probably never did.
920 PH/04 Installed a modified version of Tony Finch's patch to make submission
921 mode fix the return path as well as the Sender: header line, and to
922 add a /name= option so that you can make the user's friendly name appear
925 TF/03 Added the control = fakedefer ACL modifier.
927 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
928 Mark Lowes for thorough testing.
930 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
932 TK/03 Merged latest SRS patch from Miles Wilton.
934 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
935 with the definition in sysexits.h (which is #included earlier).
936 Fortunately, Exim does not actually use EX_OK. The code used to try to
937 preserve the sysexits.h value, by assumimg that macro definitions were
938 scanned for macro replacements. I have been disabused of this notion,
939 so now the code just undefines EX_OK before #including unistd.h.
941 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
942 in the smtp transport. When a block could not be written in a single
943 write() function, the timeout was being re-applied to each part-write.
944 This seems wrong - if the receiver was accepting one byte at a time it
945 would take for ever. The timeout is now adjusted when this happens. It
946 doesn't have to be particularly precise.
948 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
949 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
951 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
952 without a domain if the check to <postmaster@domain> fails.
954 SC/01 Eximstats: added -xls and the ability to specify output files
955 (patch written by Frank Heydlauf).
957 SC/02 Eximstats: use FileHandles for outputing results.
959 SC/03 Eximstats: allow any combination of xls, txt, and html output.
961 SC/04 Eximstats: fixed display of large numbers with -nvr option
963 SC/05 Eximstats: fixed merging of reports with empty tables.
965 SC/06 Eximstats: added the -include_original_destination flag
967 SC/07 Eximstats: removed tabs and trailing whitespace.
969 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
971 TK/06 MBOX spool code: Add real "From " MBOX separator line
972 so the .eml file is really in mbox format (even though
973 most programs do not really care). Patch from Alex Miller.
975 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
976 The latter is generated from $received_to and is only set if the
977 message has one envelope recipient. SA can use these headers,
978 obviously out-of-the-box. Patch from Alex Miller.
980 PH/08 The ${def test on a variable was returning false if the variable's
981 value was "0", contrary to what the specification has always said!
982 The result should be true unless the variable is empty.
984 PH/09 The syntax error of a character other than { following "${if
985 def:variable_name" (after optional whitespace) was not being diagnosed.
986 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
987 accidental colon was present, for example, could give incorrect results.
989 PH/10 Tidied the code in a number of places where the st_size field of a stat()
990 result is used (not including appendfile, where other changes are about
993 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
994 This involved changing a lot of size variables from int to off_t. It
995 should work with maildirs and everything.
997 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
998 spamd dying while we are connected to it.
1000 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1003 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1004 function for BDB 4.3.
1006 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1008 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1009 the sources for PCRE have been re-arranged and more files are now
1012 PH/15 The code I had for printing potentially long long variables in PH/11
1013 above was not the best (it lost precision). The length of off_t variables
1014 is now inspected at build time, and an appropriate printing format (%ld
1015 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1016 to be "long long int" or "long int". This is needed for the internal
1017 formatting function string_vformat().
1019 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1020 the configuration file to be ":syslog", then the script "guesses" where
1021 the logs files are, rather than using the compiled in default. In our
1022 case the guess is not the same as the compiled default, so the script
1023 suddenly stopped working when I started to use syslog. The patch checks
1024 to see if log_file_path is "". If so, it attempts to read it from exim
1025 with no configuration file to get the compiled in version, before it
1026 falls back to the previous guessing code."
1028 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1029 implementing BATV in an Exim configuration. See NewStuff for the gory
1032 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1033 Makefile that are specific to HP-UX.
1035 PH/18 If the "use_postmaster" option was set for a recipient callout together
1036 with the "random" option, the postmaster address was used as the MAIL
1037 FROM address for the random test, but not for the subsequent recipient
1038 test. It is now used for both.
1040 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1041 patch removes a few documentation additions to RFC 3028, because the
1042 latest draft now contains them. It adds the new en;ascii-case comparator
1043 and a new error check for 8bit text in MIME parts. Comparator and
1044 require names are now matched exactly. I enabled the subaddress
1045 extension, but it is not well tested yet (read: it works for me)."
1047 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1048 rework some of the code of TK/09 above to avoid the hardwired use of
1049 "%lld" and "long long". Replaced the call to snprintf() with a call to
1052 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1053 records point to non-existent hosts", "retry timeout exceeded", and
1054 "retry time not reached for any host after a long failure period".
1056 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1057 experimental DomainKeys support:
1059 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1060 (2) On an error such as an illegally used "control", the wrong name for
1061 the control was given.
1063 These problems did NOT occur unless DomainKeys support was compiled.
1065 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1067 PH/24 Added ${if match_ip condition.
1069 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1070 databases so that it will be absolutely obvious if a crash occurs in the
1071 DB library. This is a regular occurrence (often caused by mis-matched
1074 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1075 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1076 picked up on a user's system that detects such things. There doesn't seem
1077 to be a gcc warning option for this - only an attribute that has to be
1078 put on the function's prototype. It seems that in Fedora Core 4 they have
1079 set this on a number of new functions. No doubt there will be more in due
1082 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1083 specify a transport. However, if an address that was verified by such a
1084 router was the subject of a callout, Exim crashed because it tried to
1085 read the rcpt_include_affixes from the non-existent transport. Now it
1086 just assumes that the setting of that option is false. This bug was
1087 introduced by 4.51/PH/31.
1089 PH/28 Changed -d+all to exclude +memory, because that information is very
1090 rarely of interest, but it makes the output a lot bigger. People tend to
1091 do -d+all out of habit.
1093 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1094 code in os-type was giving problems when libc.so lives in lib64, like on
1097 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1098 aren't the modern standard, and it seems that some systems' include files
1099 don't always have them. Exim was already checking for some of the newer
1100 ones like T_AAAA, and defining it itself. I've added checks for all the
1101 record types that Exim uses.
1103 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1104 not automatically generating a new one, as it is supposed to. This
1105 prevented TLS from working. If the file did exist, but contained invalid
1106 data, a new version was generated, as expected. It was only the case of a
1107 non-existent file that was broken.
1109 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1110 with a change in libdomainkeys > 0.64.
1112 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1113 from DNS. If the selector record carries the flag, it now has
1114 precedence over the domain-wide flag.
1116 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1118 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1119 the use of a port name that isn't defined in /etc/services) occurred, the
1120 message was deferred as in a normal delivery, and thus remained on the
1121 spool, instead of being failed because of the mua_wrapper setting. This
1122 is now fixed, and I tidied up some of the mua_wrapper messages at the
1125 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1126 the messages in a hash of arrays rather than using individual hashes.
1127 This is a bit cleaner and results in dramatic memory savings, albeit
1128 at a slight CPU cost.
1130 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1131 as requested by Marc Sherman.
1133 SC/10 Eximstats: added histograms for user specified patterns as requested
1136 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1138 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1139 fopen() in the content-scanning modules that did not already have it.
1145 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1146 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1148 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1150 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1152 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1153 to test Sieve filters that use "vacation".
1155 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1156 that changes the way the GnuTLS parameters are stored in the cache file.
1157 The new format can be generated externally. For backward compatibility,
1158 if the data in the cache doesn't make sense, Exim assumes it has read an
1159 old-format file, and it generates new data and writes a new file. This
1160 means that you can't go back to an older release without removing the
1163 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1164 work if there are any delivery delays because "one_time" forces the
1165 parent to be marked "delivered", so its unseen clone is never tried
1166 again. For this reason, Exim now forbids the simultaneous setting of
1169 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1170 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1171 entry if you want to know the details. The fix, however, neglected to
1172 consider the case when local delivery batching is involved. The test for
1173 "previously delivered" was not happening when checking to see if an
1174 address could be batched with a previous (undelivered) one; under
1175 certain circumstances this could lead to multiple deliveries to the same
1178 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1179 in its include files, and this causes problems building Exim.
1181 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1182 header_syntax) but Exim was just ignoring anything given after a slash.
1183 In particular, this caused confusion with an attempt to use "verify =
1184 reverse_host_lookup/defer_ok". An error is now given when options are
1185 supplied for verify items that do not have them. (Maybe reverse_host_
1186 lookup should have a defer_ok option, but that's a different point.)
1188 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1189 defined by RFC 821) to 2048, because there were problems with some AUTH
1190 commands, and RFC 1869 says the size should be increased for extended
1191 SMTP commands that take arguments.
1193 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1196 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1197 "unknown" error; now it says that the functionality isn't in the binary.
1199 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1200 an address' error message when a string expansion fails (syntax or
1201 whatever). Otherwise the password may appear in the log. Following change
1202 PH/42 below, there is no longer a chance of it appearing in a bounce
1205 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1207 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1208 of its addresses was ever tried. (Bugzilla bug #2.)
1210 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1211 the result incorrectly in the debug output. (It correctly added a newline
1212 to what was transported.)
1214 TF/01 Added $received_time.
1216 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1217 commented out examples of how to interface to a virus scanner and to
1218 SpamAssassin. Also added commented examples of av_scanner and
1219 spamd_address settings.
1221 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1222 and controls are allowed in which ACLs. There were a couple of minor
1223 errors. Some of the entries in the conditions table (which is a table of
1224 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1225 negation of where the condition IS allowed.
1227 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1229 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1230 header file does not have a version number, so I've had to invent a new
1231 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1232 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1233 radiusclient), but it was contributed by a Radius user.
1235 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1236 files or directories, for OpenSSL.
1238 PH/19 When an Exim process that is running as root has to create an Exim log
1239 file, it does so in a subprocess that runs as exim:exim so as to get the
1240 ownership right at creation (otherwise, other Exim processes might see
1241 the file with the wrong ownership). There was no test for failure of this
1242 fork() call, which would lead to the process getting stuck as it waited
1243 for a non-existent subprocess. Forks do occasionally fail when resources
1244 run out. I reviewed all the other calls to fork(); they all seem to check
1247 PH/20 When checking for unexpected SMTP input at connect time (before writing
1248 the banner), Exim was not dealing correctly with a non-positive return
1249 from the read() function. If the client had disconnected by this time,
1250 the result was a log entry for a synchronization error with an empty
1251 string after "input=" when read() returned zero. If read() returned -1
1252 (an event I could not check), uninitialized data bytes were printed.
1253 There were reports of junk text (parts of files, etc) appearing after
1256 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1258 PH/22 Added support for macro redefinition, and (re)definition in between
1259 driver and ACL definitions.
1261 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1262 forgetting to use the resulting value; it was using the unexpanded value.
1264 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1265 hadn't been configured. The fix is from Juergen Kreileder, who
1266 understands it better than I do:
1268 "Here's what I see happening with three configured cyrus_sasl
1269 authenticators configured (plain, login, cram-md5):
1271 On startup auth_cyrus_sasl_init() gets called for each of these.
1272 This means three calls to sasl_listmech() without a specified mech_list.
1273 => SASL tests which mechs of all available mechs actually work
1274 => three warnings about OTP not working
1275 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1277 With the patch, sasl_listmech() also gets called three times. But now
1278 SASL's mech_list option is set to the server_mech specified in the the
1279 authenticator. Or in other words, the answer from sasl_listmech()
1280 gets limited to just the mech you're testing for (which is different
1282 => the return list contains just 'plain' or 'login', 'cram-md5' or
1283 nothing depending on the value of ob->server_mech.
1285 I've just tested the patch: Authentication still works fine,
1286 unavailable mechs specified in the exim configuration are still
1287 caught, and the auth.log warnings about OTP are gone."
1289 PH/25 When debugging is enabled, the contents of the command line are added
1290 to the debugging output, even when log_selector=+arguments is not
1293 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1294 answer is "GNU", and only if the return is "GNU/something" is the answer
1297 PH/27 $acl_verify_message is now set immediately after the failure of a
1298 verification in an ACL, and so is available in subsequent modifiers. In
1299 particular, the message can be preserved by coding like this:
1301 warn !verify = sender
1302 set acl_m0 = $acl_verify_message
1304 Previously, $acl_verify_message was set only while expanding "message"
1305 and "log_message" when a very denied access.
1307 PH/28 Modified OS/os.c-Linux with
1309 -#ifndef OS_LOAD_AVERAGE
1310 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1312 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1313 nomenclature these days.)
1315 PH/29 Installed patch from the Sieve maintainer that adds the options
1316 sieve_useraddress and sieve_subaddress to the redirect router.
1318 PH/30 In these circumstances:
1319 . Two addresses routed to the same list of hosts;
1320 . First host does not offer TLS;
1321 . First host accepts first address;
1322 . First host gives temporary error to second address;
1323 . Second host offers TLS and a TLS session is established;
1324 . Second host accepts second address.
1325 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1326 and peerdn, if requested) that were in fact used only for the second
1329 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1330 attention to any local part prefix or suffix that was matched by the
1331 router that accepted the address. It now behaves in the same way as it
1332 does for delivery: the affixes are removed from the local part unless
1333 rcpt_include_affixes is set on the transport.
1335 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1336 timeout during the DATA phase of an incoming message.
1338 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1339 applied a patch sanctioned by the Sieve maintainer.
1341 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1342 the uid or gid is negative. A case of a negative gid caused this to be
1343 noticed. The fix allows for either to be negative.
1345 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1346 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1349 PH/36 The change PH/12 above was broken. Fixed it.
1351 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1352 the grounds that routing the same address twice would always produce the
1353 same answer. This might have been true once, but it is certainly no
1354 longer true now. Routing a child address may depend on the previous
1355 routing that produced that child. Some complicated redirection strategies
1356 went wrong when messages had multiple recipients, and made Exim's
1357 behaviour dependent on the order in which the addresses were given.
1359 I have moved the duplicate checking until after the routing is complete.
1360 Exim scans the addresses that are assigned to local and remote
1361 transports, and removes any duplicates. This means that more work will be
1362 done, as duplicates will always all be routed, but duplicates are
1363 presumably rare, so I don't expect this is of any significance.
1365 For deliveries to pipes, files, and autoreplies, the duplicate checking
1366 still happens during the routing process, since they are not going to be
1369 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1370 It corrects a timeout issue with spamd. This is Ian's comment: "The
1371 background is that sometimes spamd either never reads data from a
1372 connection it has accepted, or it never writes response data. The exiscan
1373 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1374 blindly assumes that writes won't block so it may never time out."
1376 PH/39 Allow G after quota size as well as K and M.
1378 PH/40 The value set for $authenticated_id in an authenticator may not contain
1379 binary zeroes or newlines because the value is written to log lines and
1380 to spool files. There was no check on this. Now the value is run through
1381 the string_printing() function so that such characters are converted to
1382 printable escape sequences.
1384 PH/41 $message_linecount is a new variable that contains the total number of
1385 lines in the message. Compare $body_linecount, which is the count for the
1388 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1389 bounce and delay warning messages, except in certain special cases, which
1392 (a) An SMTP error message from a remote host;
1393 (b) A message specified in a :fail: redirection;
1394 (c) A message specified in a "fail" command in a system filter;
1395 (d) A message specified in a FAIL return from the queryprogram router;
1396 (e) A message specified by the cannot_route_message router option.
1398 In these cases only, Exim does include the error details in bounce and
1399 warning messages. There are also a few cases where bland messages such
1400 as "unrouteable address" or "local delivery error" are given.
1402 PH/43 $value is now also set for the "else" part of a ${run expansion.
1404 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1405 being worked on, but at least Exim now implements the latest version to
1408 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1409 process to complete was treated as a delivery failure, a timeout while
1410 writing the message to the pipe was logged, but erroneously treated as a
1411 successful delivery. Such timeouts include transport filter timeouts. For
1412 consistency with the overall process timeout, these timeouts are now
1413 treated as errors, giving rise to delivery failures by default. However,
1414 there is now a new Boolean option for the pipe transport called
1415 timeout_defer, which, if set TRUE, converts the failures into defers for
1416 both kinds of timeout. A transport filter timeout is now identified in
1419 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1420 systems where "make" and "gmake" are different, calling "gmake" at top
1421 level broke things. I've arranged for the value of $(MAKE) to be passed
1422 from the Makefile to this script so that it can call the same version of
1426 A note about Exim versions 4.44 and 4.50
1427 ----------------------------------------
1429 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1430 changes of various kinds. As a consequence, a big documentation update was
1431 needed. This delayed the release for rather longer than seemed good, especially
1432 in the light of a couple of (minor) security issues. Therefore, the changes
1433 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1434 release. So 4.44 and 4.50 are in effect two different branches that both start
1437 I have left the 4.50 change log unchanged; it contains all the changes since
1438 4.43. The change log for 4.44 is below; many of its items are identical to
1439 those for 4.50. This seems to be the most sensible way to preserve the
1440 historical information.
1446 1. Minor wording change to the doc/README.SIEVE file.
1448 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1449 computation of the current number of files was incorrect.
1451 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1452 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1455 4. Give more explanation in the error message when the command for a transport
1456 filter fails to execute.
1458 5. There are several places where Exim runs a non-Exim command in a
1459 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1460 was being done only for the command run by the queryprogram router. It is
1461 now done for all such subprocesses. The other cases are: ${run, transport
1462 filters, and the commands run by the lmtp and pipe transports.
1464 6. Added CONFIGURE_GROUP build-time option.
1466 7. Some older OS have a limit of 256 on the maximum number of file
1467 descriptors. Exim was using setrlimit() to set 1000 as a large value
1468 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1469 systems. I've change it so that if it can't get 1000, it tries for 256.
1471 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1472 was an oversight, and furthermore, ever since the addition of extra
1473 controls (e.g. 4.43/32), the checks on when to allow different forms of
1474 "control" were broken. There should now be diagnostics for all cases when a
1475 control that does not make sense is encountered.
1477 9. Added the /retain_sender option to "control=submission".
1479 10. $recipients is now available in the predata ACL (oversight).
1481 11. Tidy the search cache before the fork to do a delivery from a message
1482 received from the command line. Otherwise the child will trigger a lookup
1483 failure and thereby defer the delivery if it tries to use (for example) a
1484 cached ldap connection that the parent has called unbind on.
1486 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1487 of $address_data from the recipient verification was clobbered by the
1488 sender verification.
1490 13. The value of address_data from a sender verification is now available in
1491 $sender_address_data in subsequent conditions in the ACL statement.
1493 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1495 15. Added a new option "connect=<time>" to callout options, to set a different
1498 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1499 was its contents. (It was OK if the option was not defined at all.)
1501 17. A "Completed" log line is now written for messages that are removed from
1502 the spool by the -Mrm option.
1504 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1505 information about exactly what failed.
1507 19. Added -dd to debug only the daemon process.
1509 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1510 handles timeouts, both on the server side and network timeouts. Renamed the
1511 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1513 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1514 It is now set to "smtps".
1516 22. $host_address is now set to the target address during the checking of
1517 ignore_target_hosts.
1519 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1520 being passed; this would have caused $sender_host_name to have been used if
1521 matching the list had actually called for a host name (not very likely,
1522 since this list is usually IP addresses). A host name is now passed as
1525 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1526 code that specifies a non-restarting handler (typically sigaction() in
1527 modern systems) in an attempt to fix a rare and obscure crash bug.
1529 25. Narrowed the window for a race in the daemon that could cause it to ignore
1530 SIGCHLD signals. This is not a major problem, because they are used only to
1531 wake it up if nothing else does.
1533 26. A malformed maildirsize file could cause Exim to calculate negative values
1534 for the mailbox size or file count. Odd effects could occur as a result.
1535 The maildirsize information is now recalculated if the size or filecount
1538 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1539 support for a long time. Removed HAVE_SYS_VFS_H.
1541 28. Installed the latest version of exipick from John Jetmore.
1543 29. In an address list, if the pattern was not a regular expression, an empty
1544 subject address (from a bounce message) matched only if the pattern was an
1545 empty string. Non-empty patterns were not even tested. This was the wrong
1546 because it is perfectly reasonable to use an empty address as part of a
1547 database query. An empty address is now tested by patterns that are
1548 lookups. However, all the other forms of pattern expect the subject to
1549 contain a local part and a domain, and therefore, for them, an empty
1550 address still always fails if the pattern is not itself empty.
1552 30. Exim went into a mad DNS loop when attempting to do a callout where the
1553 host was specified on an smtp transport, and looking it up yielded more
1554 than one IP address.
1556 31. Re-factored the code for checking spool and log partition space into a
1557 function that finds that data and another that does the check. The former
1558 is then used to implement four new variables: $spool_space, $log_space,
1559 $spool_inodes, and $log_inodes.
1561 32. The RFC2047 encoding function was originally intended for short strings
1562 such as real names; it was not keeping to the 75-character limit for
1563 encoded words that the RFC imposes. It now respects the limit, and
1564 generates multiple encoded words if necessary. To be on the safe side, I
1565 have increased the buffer size for the ${rfc2047: expansion operator from
1568 33. It is now permitted to omit both strings after an "if" condition; if the
1569 condition is true, the result is "true". As before, when the second string
1570 is omitted, a false condition yields an empty string. This makes it less
1571 cumbersome to write custom ACL and router conditions.
1573 34. Failure to deliver a bounce message always caused it to be frozen, even if
1574 there was an errors_to setting on the router. The errors_to setting is now
1577 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1578 canonical form (fully expanded) before being placed in
1579 $sender_host_address.
1581 36. The table in the code that translates DNS record types into text (T_A to
1582 "A" for instance) was missing entries for NS and CNAME. It is just possible
1583 that this could have caused confusion if both these types were looked up
1584 for the same domain, because the text type is used as part of Exim's
1585 per-process caching. But the chance of anyone hitting this buglet seems
1588 37. The dnsdb lookup has been extended in a number of ways.
1590 (1) There is a new type, "zns", which walks up the domain tree until it
1591 finds some nameserver records. It should be used with care.
1593 (2) There is a new type, "mxh", which is like "mx" except that it returns
1594 just the host names, not the priorities.
1596 (3) It is now possible to give a list of domains (or IP addresses) to be
1597 looked up. The behaviour when one of the lookups defers can be
1598 controlled by a keyword.
1600 (4) It is now possible to specify the separator character for use when
1601 multiple records are returned.
1603 38. The dnslists ACL condition has been extended: it is now possible to supply
1604 a list of IP addresses and/or domains to be looked up in a particular DNS
1607 39. Added log_selector=+queue_time_overall.
1609 40. When running the queue in the test harness, wait just a tad after forking a
1610 delivery process, to get repeatability of debugging output.
1612 41. Include certificate and key file names in error message when GnuTLS fails
1613 to set them up, because the GnuTLS error message doesn't include the name
1614 of the failing file when there is a problem reading it.
1616 42. Allow both -bf and -bF in the same test run.
1618 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1620 44. The "Exiscan patch" is now merged into the mainline Exim source.
1622 45. Sometimes the final signoff response after QUIT could fail to get
1623 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1624 < 0 before doing a fflush(). This bug looks as though it goes back to the
1625 introduction of TLS in release 3.20, but "sometimes" must have been rare
1626 because the tests only now provoked it.
1628 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1629 (this can affect the format of dates).
1631 47. exim_tidydb, when checking for the continued existence of a message for
1632 which it has found a message-specific retry record, was not finding
1633 messages that were in split spool directories. Consequently, it was
1634 deleting retry records that should have stayed in existence.
1636 48. Steve fixed some bugs in eximstats.
1638 49. The SPA authentication driver was not abandoning authentication and moving
1639 on to the next authenticator when an expansion was forced to fail,
1640 contradicting the general specification for all authenticators. Instead it
1641 was generating a temporary error. It now behaves as specified.
1643 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1644 (the order specifies the preference for clients). The order is now AES256,
1645 AES128, 3DES, ARCFOUR128.
1647 51. Small patch to Sieve code - explicitly set From: when generating an
1650 52. Exim crashed if a remote delivery caused a very long error message to be
1651 recorded - for instance if somebody sent an entire SpamAssassin report back
1652 as a large number of 550 error lines. This bug was coincidentally fixed by
1653 increasing the size of one of Exim's internal buffers (big_buffer) that
1654 happened as part of the Exiscan merge. However, to be on the safe side, I
1655 have made the code more robust (and fixed the comments that describe what
1658 53. Now that there can be additional text after "Completed" in log lines (if
1659 the queue_time_overall log selector is set), a one-byte patch to exigrep
1660 was needed to allow it to recognize "Completed" as not the last thing in
1663 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1664 patch that reportedly fixes this has been added. I am not expert enough to
1665 create a test for it. This is what the patch creator wrote:
1667 "I found a little strange behaviour of ldap code when working with
1668 Windows 2003 AD Domain, where users was placed in more than one
1669 Organization Units. When I tried to give exim partial DN, the exit code
1670 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1671 But simultaneously result of request was absolutely normal ldap result,
1672 so I produce this patch..."
1674 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1675 REFERENCE, so I have modified the code to exclude the patch when that macro
1678 55. Some experimental protocols are using DNS PTR records for new purposes. The
1679 keys for these records are domain names, not reversed IP addresses. The
1680 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1681 leaves it alone. Component reversal etc. now happens only for IP addresses.
1684 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1686 57. Double the size of the debug message buffer (to 2048) so that more of very
1687 long debug lines gets shown.
1689 58. The exicyclog utility now does better if the number of log files to keep
1690 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1692 59. Two changes related to the smtp_active_hostname option:
1694 (1) $smtp_active_hostname is now available as a variable.
1695 (2) The default for smtp_banner uses $smtp_active_hostname instead
1696 of $primary_hostname.
1698 60. The host_aton() function is supposed to be passed a string that is known
1699 to be a valid IP address. However, in the case of IPv6 addresses, it was
1700 not checking this. This is a hostage to fortune. Exim now panics and dies
1701 if the condition is not met. A case was found where this could be provoked
1702 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1703 components; fortuitously, this particular loophole had already been fixed
1704 by change 4.50/55 above.
1706 If there are any other similar loopholes, the new check in host_aton()
1707 itself should stop them being exploited. The report I received stated that
1708 data on the command line could provoke the exploit when Exim was running as
1709 exim, but did not say which command line option was involved. All I could
1710 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1711 running as the user.
1714 61. There was a buffer overflow vulnerability in the SPA authentication code
1715 (which came originally from the Samba project). I have added a test to the
1716 spa_base64_to_bits() function which I hope fixes it.
1719 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1720 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1722 63. The daemon start-up calls getloadavg() while still root for those OS that
1723 need the first call to be done as root, but it missed one case: when
1724 deliver_queue_load_max is set with deliver_drop_privilege. This is
1725 necessary for the benefit of the queue runner, because there is no re-exec
1726 when deliver_drop_privilege is set.
1728 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1729 This has been fixed.
1731 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1732 was in use, was not putting the data itself into the right store pool;
1733 consequently, it could be overwritten for a subsequent message in the same
1734 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1737 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1739 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1740 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1741 the function were treating the return as a boolean value, which happened to
1742 work because 0=false and not-0=true, but is not correct code.
1744 68. The host_aton() function was not handling scoped IPv6 addresses (those
1745 with, for example, "%eth0" on the end) correctly.
1747 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1748 negated items (that is, ~something) in unsigned ints. Some compilers
1749 apparently mutter when there is no cast.
1751 70. If an address verification called from an ACL failed, and did not produce a
1752 user-specific message (i.e. there was only a "system" message), nothing was
1753 put in $acl_verify_message. In this situation, it now puts the system
1756 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1757 session; change 4.31/43 added the unwanted input to the log line - except
1758 that it did not do this in the start of session case. It now does.
1760 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1761 This is wrong and can cause the other end to generate a synchronization
1762 error if it is another Exim or anything else that does the synchronization
1763 check. A QUIT command is no longer sent after a timeout.
1765 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1766 during host lookups.
1768 74. The defer_ok option of callout verification was not working if it was used
1769 when verifying addresses in header lines, that is, for this case:
1771 verify = header_sender/callout=defer_ok
1773 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1774 those file descriptors could be used for SMTP connections. If anything
1775 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1776 could be sent to the SMTP client, causing chaos. The daemon now opens
1777 stdin, stdout, and stderr to /dev/null when it puts itself into the
1780 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1781 log by default. The user can override this with suitable Perl magic.
1783 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1784 the log message when discard triggers, was not working for the DATA ACL or
1785 for the non-SMTP ACL.
1787 78. Error message wording change in sieve.c.
1789 79. If smtp_accept_max_per_host was set, the number of connections could be
1790 restricted to fewer than expected, because the daemon was trying to set up
1791 a new connection before checking whether the processes handling previous
1792 connections had finished. The check for completed processes is now done
1793 earlier. On busy systems, this bug wouldn't be noticed because something
1794 else would have woken the daemon, and it would have reaped the completed
1797 80. If a message was submitted locally by a user whose login name contained one
1798 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1799 It caused a spool format error. I have fixed the spool reading code. A
1800 related problem was that the "from" clause in the Received: line became
1801 illegal because of the space(s). It is now covered by ${quote_local_part.
1803 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1806 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1808 83. Patch from Sieve maintainer to fix unterminated string problem in
1809 "vacation" handling.
1811 84. Some minor changes to the Linux configuration files to help with other
1812 OS variants using glibc.
1814 85. One more patch for Sieve to update vacation handling to latest spec.
1817 ----------------------------------------------------
1818 See the note above about the 4.44 and 4.50 releases.
1819 ----------------------------------------------------
1825 1. Change 4.43/35 introduced a bug that caused file counts to be
1826 incorrectly computed when quota_filecount was set in an appendfile
1829 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1830 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1833 3. Give more explanation in the error message when the command for a transport
1834 filter fails to execute.
1836 4. There are several places where Exim runs a non-Exim command in a
1837 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1838 was being done only for the command run by the queryprogram router. It is
1839 now done for all such subprocesses. The other cases are: ${run, transport
1840 filters, and the commands run by the lmtp and pipe transports.
1842 5. Some older OS have a limit of 256 on the maximum number of file
1843 descriptors. Exim was using setrlimit() to set 1000 as a large value
1844 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1845 systems. I've change it so that if it can't get 1000, it tries for 256.
1847 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1848 was an oversight, and furthermore, ever since the addition of extra
1849 controls (e.g. 4.43/32), the checks on when to allow different forms of
1850 "control" were broken. There should now be diagnostics for all cases when a
1851 control that does not make sense is encountered.
1853 7. $recipients is now available in the predata ACL (oversight).
1855 8. Tidy the search cache before the fork to do a delivery from a message
1856 received from the command line. Otherwise the child will trigger a lookup
1857 failure and thereby defer the delivery if it tries to use (for example) a
1858 cached ldap connection that the parent has called unbind on.
1860 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1861 of $address_data from the recipient verification was clobbered by the
1862 sender verification.
1864 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1865 was its contents. (It was OK if the option was not defined at all.)
1867 11. A "Completed" log line is now written for messages that are removed from
1868 the spool by the -Mrm option.
1870 12. $host_address is now set to the target address during the checking of
1871 ignore_target_hosts.
1873 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1874 being passed; this would have caused $sender_host_name to have been used if
1875 matching the list had actually called for a host name (not very likely,
1876 since this list is usually IP addresses). A host name is now passed as
1879 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1880 code that specifies a non-restarting handler (typically sigaction() in
1881 modern systems) in an attempt to fix a rare and obscure crash bug.
1883 15. Narrowed the window for a race in the daemon that could cause it to ignore
1884 SIGCHLD signals. This is not a major problem, because they are used only to
1885 wake it up if nothing else does.
1887 16. A malformed maildirsize file could cause Exim to calculate negative values
1888 for the mailbox size or file count. Odd effects could occur as a result.
1889 The maildirsize information is now recalculated if the size or filecount
1892 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1893 support for a long time. Removed HAVE_SYS_VFS_H.
1895 18. Updated exipick to current release from John Jetmore.
1897 19. Allow an empty sender to be matched against a lookup in an address list.
1898 Previously the only cases considered were a regular expression, or an
1901 20. Exim went into a mad DNS lookup loop when doing a callout where the
1902 host was specified on the transport, if the DNS lookup yielded more than
1905 21. The RFC2047 encoding function was originally intended for short strings
1906 such as real names; it was not keeping to the 75-character limit for
1907 encoded words that the RFC imposes. It now respects the limit, and
1908 generates multiple encoded words if necessary. To be on the safe side, I
1909 have increased the buffer size for the ${rfc2047: expansion operator from
1912 22. Failure to deliver a bounce message always caused it to be frozen, even if
1913 there was an errors_to setting on the router. The errors_to setting is now
1916 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1917 canonical form (fully expanded) before being placed in
1918 $sender_host_address.
1920 24. Updated eximstats to version 1.33
1922 25. Include certificate and key file names in error message when GnuTLS fails
1923 to set them up, because the GnuTLS error message doesn't include the name
1924 of the failing file when there is a problem reading it.
1926 26. Expand error message when OpenSSL has problems setting up cert/key files.
1929 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1930 (this can affect the format of dates).
1932 28. exim_tidydb, when checking for the continued existence of a message for
1933 which it has found a message-specific retry record, was not finding
1934 messages that were in split spool directories. Consequently, it was
1935 deleting retry records that should have stayed in existence.
1937 29. eximstats updated to version 1.35
1938 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1939 1.35 - bugfix such that pie charts by volume are generated correctly
1941 30. The SPA authentication driver was not abandoning authentication and moving
1942 on to the next authenticator when an expansion was forced to fail,
1943 contradicting the general specification for all authenticators. Instead it
1944 was generating a temporary error. It now behaves as specified.
1946 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1947 (the order specifies the preference for clients). The order is now AES256,
1948 AES128, 3DES, ARCFOUR128.
1950 31. Small patch to Sieve code - explicitly set From: when generating an
1953 32. Exim crashed if a remote delivery caused a very long error message to be
1954 recorded - for instance if somebody sent an entire SpamAssassin report back
1955 as a large number of 550 error lines. This bug was coincidentally fixed by
1956 increasing the size of one of Exim's internal buffers (big_buffer) that
1957 happened as part of the Exiscan merge. However, to be on the safe side, I
1958 have made the code more robust (and fixed the comments that describe what
1961 33. Some experimental protocols are using DNS PTR records for new purposes. The
1962 keys for these records are domain names, not reversed IP addresses. The
1963 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1964 leaves it alone. Component reversal etc. now happens only for IP addresses.
1967 34. The host_aton() function is supposed to be passed a string that is known
1968 to be a valid IP address. However, in the case of IPv6 addresses, it was
1969 not checking this. This is a hostage to fortune. Exim now panics and dies
1970 if the condition is not met. A case was found where this could be provoked
1971 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1972 components; fortuitously, this particular loophole had already been fixed
1973 by change 4.50/55 or 4.44/33 above.
1975 If there are any other similar loopholes, the new check in host_aton()
1976 itself should stop them being exploited. The report I received stated that
1977 data on the command line could provoke the exploit when Exim was running as
1978 exim, but did not say which command line option was involved. All I could
1979 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1980 running as the user.
1983 35. There was a buffer overflow vulnerability in the SPA authentication code
1984 (which came originally from the Samba project). I have added a test to the
1985 spa_base64_to_bits() function which I hope fixes it.
1988 36. The daemon start-up calls getloadavg() while still root for those OS that
1989 need the first call to be done as root, but it missed one case: when
1990 deliver_queue_load_max is set with deliver_drop_privilege. This is
1991 necessary for the benefit of the queue runner, because there is no re-exec
1992 when deliver_drop_privilege is set.
1994 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1995 was in use, was not putting the data itself into the right store pool;
1996 consequently, it could be overwritten for a subsequent message in the same
1997 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2000 38. Sometimes the final signoff response after QUIT could fail to get
2001 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2002 < 0 before doing a fflush(). This bug looks as though it goes back to the
2003 introduction of TLS in release 3.20, but "sometimes" must have been rare
2004 because the tests only now provoked it.
2010 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2011 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2012 Now it returns an integer. A number of calls were still expecting a T/F
2013 return. Fortuitously, in all cases, the tests worked in OK situations,
2014 which is the norm. However, things would have gone wrong on any write
2015 failures on the smtp file descriptor. This function is used when sending
2016 messages over SMTP and also when doing verify callouts.
2018 2. When Exim is called to do synchronous delivery of a locally submitted
2019 message (the -odf or -odi options), it no longer closes stderr before doing
2022 3. Implemented the mua_wrapper option.
2024 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2026 5. Implemented the functions header_remove(), header_testname(),
2027 header_add_at_position(), and receive_remove_recipient(), and exported them
2030 6. If an ACL "warn" statement specified the addition of headers, Exim already
2031 inserted X-ACL-Warn: at the start if there was no header name. However, it
2032 was not making this test for the second and subsequent header lines if
2033 there were newlines in the string. This meant that an invalid header could
2034 be inserted if Exim was badly configured.
2036 7. Allow an ACL "warn" statement to add header lines at the start or after all
2037 the Received: headers, as well as at the end.
2039 8. Added the rcpt_4xx retry error code.
2041 9. Added postmaster_mailfrom=xxx to callout verification option.
2043 10. Added mailfrom=xxxx to the callout verification option, for verify=
2046 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2047 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2049 12. Inserted some casts to stop certain compilers warning when using pointer
2050 differences as field lengths or precisions in printf-type calls (mostly
2051 affecting debugging statements).
2053 13. Added optional readline() support for -be (dynamically loaded).
2055 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2056 same clock tick as a message's arrival, so that its received time was the
2057 same as the "first fail" time on the retry record, and that message
2058 remained on the queue past the ultimate address timeout, every queue runner
2059 would try a delivery (because it was past the ultimate address timeout) but
2060 after another failure, the ultimate address timeout, which should have then
2061 bounced the address, did not kick in. This was a "< instead of <=" error;
2062 in most cases the first failure would have been in the next clock tick
2063 after the received time, and all would be well.
2065 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2066 being recognized when the domain list was tested by the match_domain
2067 condition in an expansion string.
2069 16. Added the ${str2b64: operator.
2071 17. Exim was always calling setrlimit() to set a large limit for the number of
2072 processes, without checking whether the existing limit was already
2073 adequate. (It did check for the limit on file descriptors.) Furthermore,
2074 errors from getrlimit() and setrlimit() were being ignored. Now they are
2075 logged to the main and panic logs, but Exim does carry on, to try to do its
2076 job under whatever limits there are.
2078 18. Imported PCRE 5.0.
2080 19. Trivial typo in log message " temporarily refused connection" (the leading
2083 20. If the log selector return_path_on_delivery was set and an address was
2084 redirected to /dev/null, the delivery process crashed because it assumed
2085 that a return path would always be set for a "successful" delivery. In this
2086 case, the whole delivery is bypassed as an optimization, and therefore no
2089 21. Internal re-arrangement: the function for sending a challenge and reading
2090 a response while authentication was assuming a zero-terminated challenge
2091 string. It's now changed to take a pointer and a length, to allow for
2092 binary data in such strings.
2094 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2096 23. Exim was not respecting finduser_retries when seeking the login of the
2097 uid under which it was called; it was always trying 10 times. (The default
2098 setting of finduser_retries is zero.) Also, it was sleeping after the final
2099 failure, which is pointless.
2101 24. Implemented tls_on_connect_ports.
2103 25. Implemented acl_smtp_predata.
2105 26. If the domain in control=submission is set empty, Exim assumes that the
2106 authenticated id is a complete email address when it generates From: or
2107 Sender: header lines.
2109 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2110 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2111 chown and chgrp in /bin and hostname in /usr/bin.
2113 28. Exim was keeping the "process log" file open after each use, just as it
2114 does for the main log. This opens the possibility of it remaining open for
2115 long periods when the USR1 signal hits a daemon. Occasional processlog
2116 errors were reported, that could have been caused by this. Anyway, it seems
2117 much more sensible not to leave this file open at all, so that is what now
2120 29. The long-running daemon process does not normally write to the log once it
2121 has entered its main loop, and it closes the log before doing so. This is
2122 so that log files can straightforwardly be renamed and moved. However,
2123 there are a couple of unusual error situations where the daemon does write
2124 log entries, and I had neglected to close the log afterwards.
2126 30. The text of an SMTP error response that was received during a remote
2127 delivery was being truncated at 512 bytes. This is too short for some of
2128 the long messages that one sometimes sees. I've increased the limit to
2131 31. It is now possible to make retry rules that apply only when a message has a
2132 specific sender, in particular, an empty sender.
2134 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2135 it possible to be selective about when SMTP synchronization is enforced.
2137 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2139 32. Implemented hosts_connection_nolog.
2141 33. Added an ACL for QUIT.
2143 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2146 35. Added mailbox_size and mailbox_filecount to appendfile.
2148 36. Added control = no_multiline_responses to ACLs.
2150 37. There was a bug in the logic of the code that waits for the clock to tick
2151 in the case where the clock went backwards by a substantial amount such
2152 that the microsecond fraction of "now" was more than the microsecond
2153 fraction of "then" (but the whole seconds number was less).
2155 38. Added support for the libradius Radius client library this is found on
2156 FreeBSD (previously only the radiusclient library was supported).
2162 1. When certain lookups returned multiple values in the form name=value, the
2163 quoting of the values was not always being done properly. Specifically:
2164 (a) If the value started with a double quote, but contained no whitespace,
2166 (b) If the value contained whitespace other than a space character (i.e.
2167 tabs or newlines or carriage returns) it was not quoted.
2168 This fix has been applied to the mysql and pgsql lookups by writing a
2169 separate quoting function and calling it from the lookup code. The fix
2170 should probably also be applied to nisplus, ibase and oracle lookups, but
2171 since I cannot test any of those, I have not disturbed their existing code.
2173 2. A hit in the callout cache for a specific address caused a log line with no
2174 reason for rejecting RCPT. Now it says "Previous (cached) callout
2175 verification failure".
2177 3. There was an off-by-one bug in the queryprogram router. An over-long
2178 return line was truncated at 256 instead of 255 characters, thereby
2179 overflowing its buffer with the terminating zero. As well as fixing this, I
2180 have increased the buffer size to 1024 (and made a note to document this).
2182 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2183 when Exim is waiting for an SMTP response from a remote server, Exim
2184 restarts its select() call on the socket, thereby resetting its timeout.
2185 This is not a problem when such interrupts are rare. Somebody set up a cron
2186 job to run exiwhat every 2 minutes, which is less than the normal select()
2187 timeout (5 or 10 minutes). This meant that the select() timeout never
2188 kicked in because it was always reset. I have fixed this by comparing the
2189 time when an interrupt arrives with the time at the start of the first call
2190 to select(). If more time than the timeout has elapsed, the interrupt is
2191 treated as a timeout.
2193 5. Some internal re-factoring in preparation for the addition of Sieve
2194 extensions (by MH). In particular, the "personal" test is moved to a
2195 separate function, and given an option for scanning Cc: and Bcc: (which is
2196 not set for Exim filters).
2198 6. When Exim created an email address using the login of the caller as the
2199 local part (e.g. when creating a From: or Sender: header line), it was not
2200 quoting the local part when it contained special characters such as @.
2202 7. Installed new OpenBSD configuration files.
2204 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2205 try to make them clearer.
2207 9. Callout options, other than the timeout value, were being ignored when
2208 verifying sender addresses in header lines. For example, when using
2210 verify = header_sender/callout=no_cache
2212 the cache was (incorrectly) being used.
2214 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2215 only the Cygwin environment.
2217 11. When return_path_on_delivery was set as a log selector, if different remote
2218 addresses in the same message used different return paths and parallel
2219 remote delivery occurred, the wrong values would sometimes be logged.
2220 (Whenever a remote delivery process finished, the return path value from
2221 the most recently started remote delivery process was logged.)
2223 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2224 lines when AUTH and/or TLS are in use. This is the "received protocol"
2225 field. Exim used to use "asmtp" for authenticated SMTP, without any
2226 indication (in the protocol name) for TLS use. Now it follows the RFC and
2227 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2228 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2229 names appear in log lines as well as in Received: header lines.
2231 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2232 extensions, and comparison tests, and to fix some bugs.
2234 14. Changes to the "personal" filter test:
2236 (1) The test was buggy in that it was just doing the equivalent of
2237 "contains" tests on header lines. For example, if a user's address was
2238 anne@some.where, the "personal" test would incorrectly be true for
2240 To: susanne@some.where
2242 This test is now done by extracting each address from the header in turn,
2243 and checking the entire address. Other tests that are part of "personal"
2244 are now done using regular expressions (for example, to check local parts
2245 of addresses in From: header lines).
2247 (2) The list of non-personal local parts in From: addresses has been
2248 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2249 taken from the Sieve specification recommendations.
2251 (3) If the message contains any header line starting with "List-" it is
2252 treated as non-personal.
2254 (4) The test for "circular" in the Subject: header line has been removed
2255 because it now seems ill-conceived.
2257 15. Minor typos in src/EDITME comments corrected.
2259 16. Installed latest exipick from John Jetmore.
2261 17. If headers_add on a router specified a text string that was too long for
2262 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2263 of string_sprintf() is now avoided.
2265 18. $message_body_size was not set (it was always zero) when running the DATA
2266 ACL and the local_scan() function.
2268 19. For the "mail" command in an Exim filter, no default was being set for
2269 the once_repeat time, causing a random time value to be used if "once" was
2270 specified. (If the value happened to be <= 0, no repeat happened.) The
2271 default is now 0s, meaning "never repeat". The "vacation" command was OK
2272 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2273 (I found it when inspecting the code).
2275 20. There is now an overall timeout for performing a callout verification. It
2276 defaults to 4 times the callout timeout, which applies to individual SMTP
2277 commands during the callout. The overall timeout applies when there is more
2278 than one host that can be tried. The timeout is checked before trying the
2279 next host. This prevents very long delays if there are a large number of
2280 hosts and all are timing out (e.g. when the network connections are timing
2281 out). The value of the overall timeout can be changed by specifying an
2282 additional sub-option for "callout", called "maxwait". For example:
2284 verify = sender/callout=5s,maxwait=20s
2286 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2287 to the end before writing, but this should make it even safer).
2289 22. Exim was forgetting that it had advertised PIPELINING for the second and
2290 subsequent messages on an SMTP connection. It was also not resetting its
2291 memory on STARTTLS and an internal HELO.
2293 23. When Exim logs an SMTP synchronization error within a session, it now
2294 records whether PIPELINING has been advertised or not.
2296 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2297 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2298 rather than long int.
2300 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2302 26. Added the never_mail option to autoreply.
2308 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2309 crash if the getsockname() call failed; this can happen if a connection is
2310 closed very soon after it is established. The problem was simply in the
2311 order in which certain operations were done, causing Exim to try to write
2312 to the SMTP stream before it had set up the file descriptor. The bug has
2313 been fixed by making things happen in the correct order.
2319 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2320 before the connection was closed, thus losing the rejection response.
2322 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2323 some early Solaris releases, but causes trouble in current releases where
2324 socklen_t is defined.
2326 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2329 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2332 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2333 that releasing the top of it at the end releases what was used for sub-
2334 expansions (unless the block got too big). However, discard this block if
2335 the first thing is a variable or header, so that we can use its block when
2336 it is dynamic (useful for very large $message_headers, for example).
2338 6. Lookups now cache *every* query, not just the most recent. A new, separate
2339 store pool is used for this. It can be recovered when all lookup caches are
2340 flushed. Lookups now release memory at the end of their result strings.
2341 This has involved some general refactoring of the lookup sources.
2343 7. Some code has been added to the store_xxx() functions to reduce the amount
2344 of flapping under certain conditions.
2346 8. log_incoming_interface used to affect only the <= reception log lines. Now
2347 it causes the local interface and port to be added to several more SMTP log
2348 lines, for example "SMTP connection from", and rejection lines.
2350 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2352 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2354 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2355 could be overwritten at the end of the current message (or the start of a
2356 new message if it was set in a HELO ACL). The value is now preserved for
2357 the duration of the SMTP connection.
2359 12. If a transport had a headers_rewrite setting, and a matching header line
2360 contained an unqualified address, that address was qualified, even if it
2361 did not match any rewriting rules. The underlying bug was that the values
2362 of the flags that permit the existence of unqualified sender and recipient
2363 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2364 non-local messages, and by -bnq for local messages) were not being
2365 preserved with the message after it was received.
2367 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2368 "next input=" as part of the text comprising the host identity instead of
2369 the correct text. The code was using the same buffer for two different
2370 strings. However, depending on which order the printing function evaluated
2371 its arguments, the bug did not always show up. Under Linux, for example, my
2372 test suite worked just fine.
2374 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2375 doesn't work with some older versions of Perl. It has been changed to "my",
2376 which in any case is probably the better facility to use.
2378 15. A really picky compiler found some instances of statements for creating
2379 error messages that either had too many or two few arguments for the format
2382 16. The size of the buffer for calls to the DNS resolver has been increased
2383 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2384 for addresses that have a lot of PTR records. This alleviates a problem; it
2385 does not fully solve it.
2387 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2388 buffer now truncates the list and logs the incident, which is the same
2389 action as happens when Exim is looking up a host name and its aliases.
2390 Previously in this situation something unpredictable would happen;
2391 sometimes it was "internal error: store_reset failed".
2393 18. If a server dropped the connection unexpectedly when an Exim client was
2394 using GnuTLS and trying to read a response, the client delivery process
2395 crashed while trying to generate an error log message.
2397 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2398 string, for example:
2400 warn message = H1: something\nH2: something
2402 the text was added as a single header line from Exim's point of view
2403 though it ended up OK in the delivered message. However, searching for the
2404 second and subsequent header lines using $h_h2: did not work. This has been
2405 fixed. Similarly, if a system filter added multiple headers in this way,
2406 the routers could not see them.
2408 20. Expanded the error message when iplsearch is called with an invalid key to
2409 suggest using net-iplsearch in a host list.
2411 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2412 ACLs are no longer actually imposed (and a message to that effect is
2415 22. If a "gecos" field in a passwd entry contained escaped characters, in
2416 particular, if it contained a \" sequence, Exim got it wrong when building
2417 a From: or a Sender: header from that name. A second bug also caused
2418 incorrect handling when an unquoted " was present following a character
2419 that needed quoting.
2421 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2422 was not being matched caselessly.
2424 24. Arranged for all hyphens in the exim.8 source to be escaped with
2427 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2428 properly. Recipient callout cache records were still being keyed to include
2429 the sender, even when use_sender was set false. This led to far more
2430 callouts that were necessary. The sender is no longer included in the key
2431 when use_sender is false.
2433 26. Added "control = submission" modifier to ACLs.
2435 27. Added the ${base62d: operator to decode base 62 numbers.
2437 28. dnsdb lookups can now access SRV records.
2439 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2440 the configuration file.
2442 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2443 (-v) mode. This makes the output for a verbose queue run more intelligible.
2445 31. Added a use_postmaster feature to recipient callouts.
2447 32. Added the $body_zerocount variable, containing the number of binary zero
2448 bytes in the message body.
2450 33. The time of last modification of the "new" subdirectory is now used as the
2451 "mailbox time last read" when there is a quota error for a maildir
2454 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2456 35. Added +ignore_unknown as a special item in host lists.
2458 36. Code for decoding IPv6 addresses in host lists is now included, even if
2459 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2460 address was recognized as an IP address, but was then not correctly decoded
2461 into binary, causing unexpected and incorrect effects when compared with
2468 1. Very minor rewording of debugging text in manualroute to say "list of
2469 hosts" instead of "hostlist".
2471 2. If verify=header_syntax was set, and a header line with an unqualified
2472 address (no domain) and a large number of spaces between the end of the
2473 name and the colon was received, the reception process suffered a buffer
2474 overflow, and (when I tested it) crashed. This was caused by some obsolete
2475 code that should have been removed. The fix is to remove it!
2477 3. When running in the test harness, delay a bit after writing a bounce
2478 message to get a bit more predictability in the log output.
2480 4. Added a call to search_tidyup() just before forking a reception process. In
2481 theory, someone could use a lookup in the expansion of smtp_accept_max_
2482 per_host which, without the tidyup, could leave open a database connection.
2484 5. Added the variables $recipient_data and $sender_data which get set from a
2485 lookup success in an ACL "recipients" or "senders" condition, or a router
2486 "senders" option, similar to $domain_data and $local_part_data.
2488 6. Moved the writing of debug_print from before to after the "senders" test
2491 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2492 problems for message scanning, either using a data ACL, or using
2493 local_scan() because the Received: header was not generated till after they
2494 were called (in order to set the time as the time of reception completion).
2495 I have revised the way this works. The header is now generated after the
2496 body is received, but before the ACL or local_scan() are called. After they
2497 are run, the timestamp in the header is updated.
2503 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2504 before starting a queue runner without re-exec. This happened only when
2505 deliver_drop_privilege was set or when the Exim user was set to root. The
2506 effect of the bug was that timeouts during subsequent deliveries caused
2507 crashes instead of being properly handled. The handler is now left at its
2508 default (and expected) setting.
2510 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2511 message, again when deliver_drop_privilege is set or Exim is run as root.
2512 The bug described in (1) was not present in this case, but the tidying up
2513 of the other signals was missing. I have made the two cases consistent.
2515 3. The ignore_target_hosts setting on a manualroute router was being ignored
2516 for hosts that were looked up using the /MX notation.
2518 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2521 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2522 operated on the sender address. After changing the $sender_address to <>
2523 for the sender address verify, Exim was re-instated it as the original
2524 (before rewriting) address, but remembering that it had rewritten it, so it
2525 wasn't rewriting it again. This bug also had the effect of breaking the
2526 sender address verification caching when the sender address was rewritten.
2528 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2529 This has been changed so that if the ip literal address matches
2530 ignore_target_hosts, the router declines.
2532 7. Added expansion conditions match_domain, match_address, and match_local_
2533 part (NOT match_host).
2535 8. The placeholder for the Received: header didn't have a length field set.
2537 9. Added code to Exim itself and to exim_lock to test for a specific race
2538 condition that could lead to file corruption when using MBX delivery. The
2539 issue is with the lockfile that is created in /tmp. If this file is removed
2540 after a process has opened it but before that process has acquired a lock,
2541 there is the potential for a second process to recreate the file and also
2542 acquire a lock. This could lead to two Exim processes writing to the file
2543 at the same time. The added code performs the same test as UW imapd; it
2544 checks after acquiring the lock that its file descriptor still refers to
2545 the same named file.
2547 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2548 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2551 11. Added the smtp_active_hostname option. If used, this will typically be made
2552 to depend on the incoming interface address. Because $interface_address is
2553 not set up until the daemon has forked a reception process, error responses
2554 that can happen earlier (such as "too many connections") no longer contain
2557 12. If an expansion in a condition on a "warn" statement fails because a lookup
2558 defers, the "warn" statement is abandoned, and the next ACL statement is
2559 processed. Previously this caused the whole ACL to be aborted.
2561 13. Added the iplsearch lookup type.
2563 14. Added ident_timeout as a log selector.
2565 15. Added tls_certificate_verified as a log selector.
2567 16. Added a global option tls_require_ciphers (compare the smtp transport
2568 option of the same name). This controls incoming TLS connections.
2570 17. I finally figured out how to make tls_require_ciphers do a similar thing
2571 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2572 before starting the TLS session.
2574 18. Tabs are now shown as \t in -bP output.
2576 19. If the log selector return_path_on_delivery was set, Exim crashed when
2577 bouncing a message because it had too many Received: header lines.
2579 20. If two routers both had headers_remove settings, and the first one included
2580 a superfluous trailing colon, the final name in the first list and the
2581 first name in the second list were incorrectly joined into one item (with a
2582 colon in the middle).
2588 1. Added -C and -D options to the exinext utility, mainly to make it easier
2589 to include in the automated testing, but these could be helpful when
2590 multiple configurations are in use.
2592 2. The exinext utility was not formatting the output nicely when there was
2593 an alternate port involved in the retry record key, nor when there was a
2594 message id as well (for retries that were specific to a specific message
2595 and a specific host). It was also confused by IPv6 addresses, because of
2596 the additional colons they contain. I have fixed the IPv4 problem, and
2597 patched it up to do a reasonable job for IPv6.
2599 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2600 delivery, the log line now contains "pipelined" if PIPELINING was used.
2602 4. An SMTP transport process used to panic and die if the bind() call to set
2603 an explicit outgoing interface failed. This has been changed; it is now
2604 treated in the same way as a connect() failure.
2606 5. A reference to $sender_host_name in the part of a conditional expansion
2607 that was being skipped was still causing a DNS lookup. This no longer
2610 6. The def: expansion condition was not recognizing references to header lines
2611 that used bh_ and bheader_.
2613 7. Added the _cache feature to named lists.
2615 8. The code for checking quota_filecount in the appendfile transport was
2616 allowing one more file than it should have been.
2618 9. For compatibility with Sendmail, the command line option
2626 and sets the incoming protocol and host name (for trusted callers). The
2627 host name and its colon can be omitted when only the protocol is to be set.
2628 Note the Exim already has two private options, -pd and -ps, that refer to
2629 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2630 "s", but I don't think that's a major issue.
2632 10. A number of refactoring changes to the code, none of which should affect
2635 (a) The number of logging options was getting close to filling up the
2636 32-bit word that was used as a bit map. I have split them into two classes:
2637 those that are passed in the argument to log_write(), and those that are
2638 only ever tested independently outside of that function. These are now in
2639 separate 32-bit words, so there is plenty of room for expansion again.
2640 There is no change in the user interface or the logging behaviour.
2642 (b) When building, for example, log lines, the code previously used a
2643 macro that called string_cat() twice, in order to add two strings. This is
2644 not really sufficiently general. Furthermore, there was one instance where
2645 it was actually wrong because one of the argument was used twice, and in
2646 one call a function was used. (As it happened, calling the function twice
2647 did not affect the overall behaviour.) The macro has been replaced by a
2648 function that can join an arbitrary number of extra strings onto a growing
2651 (c) The code for expansion conditions now uses a table and a binary chop
2652 instead of a serial search (which was left over from when there were very
2653 few conditions). Also, it now recognizes conditions like "pam" even when
2654 the relevant support is not compiled in: a suitably worded error message is
2655 given if an attempt is made to use such a condition.
2657 11. Added ${time_interval:xxxxx}.
2659 12. A bug was causing one of the ddress fields not to be passed back correctly
2660 from remote delivery subprocesses. The field in question was not being
2661 subsequently used, so this caused to problems in practice.
2663 13. Added new log selectors queue_time and deliver_time.
2665 14. Might have fixed a bug in maildirsizefile handling that threw up
2666 "unexpected character" debug warnings, and recalculated the data
2667 unnecessarily. In any case, I expanded the warning message to give more
2670 15. Added the message "Restricted characters in address" to the statements in
2671 the default ACL that block characters like @ and % in local parts.
2673 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2674 Three changes have been made:
2676 (a) There was a serious bug; a negative response to MAIL caused the whole
2677 recipient domain to be cached as invalid, thereby blocking all messages
2678 to all local parts at the same domain, from all senders. This bug has
2679 been fixed. The domain is no longer cached after a negative response to
2680 MAIL if the sender used is not empty.
2682 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2685 (c) A new callout option, "use_sender" has been added for people who want
2686 the modified behaviour.
2692 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2695 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2696 indeed breaks things for older releases.
2698 3. Added additional logging to the case where there is a problem reading data
2699 from a filter that is running in a subprocess using a pipe, in order to
2700 try to track down a specific problem.
2702 4. Testing facility fudge: when running in the test harness and attempting
2703 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2704 getting "No route to host". Convert this to a timeout.
2706 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2709 6. Some OS don't have socklen_t but use size_t instead. This affects the
2710 fifth argument of getsockopt() amongst other things. This is now
2711 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2712 can be set for individual OS. I have set it for SunOS5, OSF1, and
2713 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2714 some earlier ones do not.
2716 7. Change 4.30/15 was not doing the test caselessly.
2718 8. The standard form for an IPv6 address literal was being rejected by address
2719 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2720 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2721 this, as well as the form without the "IPv6" on the front (but only when
2722 address literals are enabled, of course).
2724 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2726 10. Exim crashed if a message with an empty sender address specified by -f
2727 encountered a router with an errors_to setting. This could be provoked only
2728 by a command such as
2732 where an empty string was supplied; "<>" did not hit this bug.
2734 11. Installed PCRE release 4.5.
2736 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2737 remained set. It is now erased.
2739 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2740 times from message ids (which are base 36 rather than the normal 62).
2742 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2743 were being counted as actual protocol errors, and logged if the log
2744 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2745 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2746 and DATA following a set of rejected RCPTs do not count as protocol errors.
2747 In other words, Exim assumes they were pipelined, though this may not
2748 actually be the case. Of course, in all cases the client gets an
2749 appropriate error code.
2751 15. If a lookup fails in an ACL condition, a message about the failure may
2752 be available; it is used if testing the ACL cannot continue, because most
2753 such messages specify what the cause of the deferral is. However, some
2754 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2755 that caused an old message to be retained and used if a later statement
2756 caused a defer, replacing the real cause of the deferral.
2758 16. If an IP address had so many PTR records that the DNS lookup buffer
2759 was not large enough to hold them, Exim could crash while trying to process
2760 the truncated data. It now detects and logs this case.
2762 17. Further to 4.21/58, another change has been made: if (and only if) the
2763 first line of a message (the first header line) ends with CRLF, a bare LF
2764 in a subsequent header line has a space inserted after it, so as not to
2765 terminate the header.
2767 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2768 unnecessarily, used atoi() instead of strtol(), and didn't check the
2769 termination when getting file sizes from file names by regex.
2771 19. Completely re-implemented the support for maildirsize files, in the light
2772 of a number of problems with the previous contributed implementation
2773 (4.30/29). In particular:
2775 . If the quota is zero, the maildirsize file is maintained, but no quota is
2778 . If the maildir directory does not exist, it is created before any attempt
2779 to write a maildirsize file.
2781 . The quota value in the file is just a cache; if the quota is changed in
2782 the transport, the new value overrides.
2784 . A regular expression is available for excluding directories from the
2787 20. The autoreply transport checks the characters in options that define the
2788 message's headers; it allows continued headers, but it was checking with
2789 isspace() after an embedded newline instead of explicitly looking for a
2792 21. If all the "regular" hosts to which an address was routed had passed their
2793 expiry times, and had not reached their retry times, the address was
2794 bounced, even if fallback hosts were defined. Now Exim should go on to try
2797 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2798 equivalent code in the SMTP transport. Some hosts send humungous responses
2799 to HELO/EHLO, more than 1024 it seems.
2801 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2802 gives compiler warnings in some environments. I've now done it "properly",
2805 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2806 (because of problems with the built-in one) was declared to return uschar *
2807 instead of char *, causing compiler failure.
2809 25. Fixed a file descriptor leak when processing alias/forward files.
2811 26. Fixed a minor format string issue in dbfn.c.
2813 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2815 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2816 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2819 29. When a sender address is verified, it is cached, to save repeating the test
2820 when there is more than one recipient in a message. However, when the
2821 verification involves a callout, it is possible for different callout
2822 options to be set for different recipients. It is too complicated to keep
2823 track of this in the cache, so now Exim always runs a verification when a
2824 callout is required, relying on the callout cache for the optimization.
2825 The overhead is duplication of the address routing, but this should not be
2828 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2829 to be verified with callout=postmaster, and the main callout worked but the
2830 postmaster check failed, the verification correctly failed. However, if a
2831 subsequent RCPT command asked for sender verification *without* the
2832 postmaster check, incorrect caching caused this verification also to fail,
2835 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2836 it was not caching the DNS options (qualify_single, search_parents) that
2837 were used when the lookup failed. A subsequent lookup with different
2838 options therefore always gave the same answer, though there were cases
2839 where it should not have. (Example: a "domains = !$mx_any" option on a
2840 dnslookup router: the "domains" option is always processed without any
2841 widening, but the router might have qualify_single set.) Now Exim uses the
2842 cached value only when the same options are set.
2844 32. Added John Jetmore's "exipick" utility to the distribution.
2846 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2847 than a timeout (e.g. a certificate is required, and is not provided), an
2848 Exim server now closes the connection immediately. Previously it waited for
2849 the client to close - but if the client is SSL, it seems that they each
2850 wait for each other, leading to a delay before one of them times out.
2852 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2853 maintained 0.8.x compatibility because I don't think many are using it, and
2854 it is clearly obsolete.
2856 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2859 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2860 client certificate was expired. A simple patch fixes this, though I don't
2861 understand the full logic of why the verify callback is called multiple
2864 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2865 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2866 which causes problems with some clients (such as the Certicom SSL Plus
2867 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2868 disables the coutermeasure allowing Eudora to connect."
2870 38. Exim was not checking that a write() to a log file succeeded. This could
2871 lead to Bad Things if a log got too big, in particular if it hit a file
2872 size limit. Exim now panics and dies if it cannot write to a log file, just
2873 as it does if it cannot open a log file.
2875 39. Modified OS/Makefile-Linux so that it now contains
2877 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2879 The two -D definitions ensure that Exim is compiled with large file
2880 support, which makes it possible to handle log files that are bigger than
2883 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2884 instance) a domain was checked against a named list that involved a lookup,
2885 causing $domain_data to be set, then another domain was checked against the
2886 same list, then the first domain was re-checked, the value of $domain_data
2887 after the final check could be wrong. In particular, if the second check
2888 failed, it could be set empty. This bug probably also applied to
2891 41. The strip_trailing_dot option was not being applied to the address given
2892 with the -f command-line option.
2894 42. The code for reading a message's header from the spool was incrementing
2895 $received_count, but never initializing it. This meant that the value was
2896 incorrect (doubled) while delivering a message in the same process in which
2897 it was received. In the most common configuration of Exim, this never
2898 happens - a fresh exec is done - but it can happen when
2899 deliver_drop_privilege is set.
2901 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2902 it now includes up to 150 characters of the unexpected data in the log
2905 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2906 and building data strings. The size of both of these buffers was 10 000
2907 bytes - far larger than anybody would *ever* want, thought I. Needless to
2908 say, somebody hit the limit. I have increased the maximum line length to
2909 20 000 and the maximum data length of concatenated lines to 100 000. I have
2910 also fixed two bugs, because there was no checking on these buffers. Tsk,
2911 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2912 buffer is too small.
2914 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2915 lsearch lookups. Now it does.
2917 46. When parsing a route_list item in a manualroute router, a fixed-length
2918 buffer was used for the list of hosts. I made this 1024 bytes long,
2919 thinking that nobody would ever have a list of hosts that long. Wrong.
2920 Somebody had a whole pile of complicated expansion conditions, and the
2921 string was silently truncated, leading to an expansion error. It turns out
2922 that it is easier to change to an unlimited length (owing to other changes
2923 that have happened since this code was originally written) than to build
2924 structure for giving a limitation error. The length of the item that
2925 expands into the list of hosts is now unlimited.
2927 47. The lsearch lookup could not handle data where the length of text line was
2928 more than 4095 characters. Such lines were truncated, leading to shortened
2929 data being returned. It should now handle lines of any length.
2931 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2932 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2935 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2936 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2937 no longer happens in comment lines. A list of replacements is now placed
2938 at the head of all of the source files, except those whose only change is
2939 to replace PERL_COMMAND in the very first #! line.
2941 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2942 messages on the queue, with a bottom-up merge sort, using code contributed
2943 by Michael Haardt. This should make operations like -bp somewhat faster on
2944 large queues. It won't affect queue runners, except when queue_run_in_order
2947 51. Installed eximstats 1.31 in the distribution.
2949 52. Added support for SRV lookups to the dnslookup router.
2951 53. If an ACL referred to $message_body or $message_body_end, the value was not
2952 reset for any messages that followed in the same SMTP session.
2954 54. The store-handling optimization for building very long strings was not
2955 differentiating between the different store pools. I don't think this
2956 actually made any difference in practice, but I've tidied it.
2958 55. While running the routers to verify a sender address, $sender_address
2959 was still set to the sender address. This is wrong, because when routing to
2960 send a bounce to the sender, it would be empty. Therefore, I have changed
2961 it so that, while verifying a sender address, $sender_address is set to <>.
2962 (There is no change to what happens when verifying a recipient address.)
2964 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2965 target A or AAAA records (if not already returned) without resetting the
2966 qualify_single or search_parents options of the DNS resolver. These are
2967 inappropriate in this case because the targets of MX and SRV records must
2968 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2969 target that, when qualified, matched something in the local domain. These
2970 two options are now turned off when doing these lookups.
2972 57. It seems that at least some releases of Reiserfs (which does not have the
2973 concept of a fixed number of inodes) returns zero and not -1 for the
2974 number of available inodes. This interacted badly with check_spool_inodes,
2975 which assumed that -1 was the "no such thing" setting. What I have done is
2976 to check that the total number of inodes is greater than zero before doing
2977 the test of how many are available.
2979 58. When a "warn" ACL statement has a log_message modifier, the message is
2980 remembered, and not repeated. This is to avoid a lot of repetition when a
2981 message has many recipients that cause the same warning to be written.
2982 Howewer, Exim was preserving the list of already written lines for an
2983 entire SMTP session, which doesn't seem right. The memory is now reset if a
2984 new message is started.
2986 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2987 debugging output unless log_rewrite was also set.
2989 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2990 of (int)(handle) when we know that handle contains (void *)(-1).
2992 61. The Exim daemon panic-logs an error return when it closes the incoming
2993 connection. However "connection reset by peer" seems to be common, and
2994 isn't really an error worthy of noting specially, so that particular error
2997 62. When Exim is trying to find all the local interfaces, it used to panic and
2998 die if the ioctl to get the interface flags failed. However, it seems that
2999 on at least one OS (Solaris 9) it is possible to have an interface that is
3000 included in the list of interfaces, but for which you get a failure error
3001 for this call. This happens when the interface is not "plumbed" into a
3002 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3003 failure of the "get flags" call assumes that the interface is down.
3005 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3006 makes life easier for people who are doing arithmetic on fields extracted
3007 from dates, where you often get leading zeros that should not be
3008 interpreted as octal.
3010 64. Added qualify_domain to the redirect router, to override the global
3013 65. If a pathologically long header line contained very many addresses (the
3014 report of this problem mentioned 10 000) and each of them was rewritten,
3015 Exim could use up a very large amount of memory. (It kept on making new
3016 copies of the header line as it rewrote, and never released the old ones.)
3017 At the expense of a bit more processing, the header rewriting function has
3018 been changed so that it no longer eats memory in this way.
3020 66. The generation of the Received: header has been moved from the time that a
3021 message starts to be received, to the time that it finishes. The timestamp
3022 in the Received: header should now be very close to that of the <= log
3023 line. There are two side-effects of this change:
3025 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3026 logged header lines no longer include the local Received: line, because
3027 it has not yet been created. The same applies to a copy of the message
3028 that is returned to a non-SMTP sender when a message is rejected.
3030 (b) When a filter file is tested using -bf, no additional Received: header
3031 is added to the test message. After some thought, I decided that this
3034 This change does not affect the value of $received_for. It is still set
3035 after address rewriting, but before local_scan() is called.
3037 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3039 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3040 gave an unhelpful panic error message, and a defer error. I have managed to
3041 change this behaviour so that it now rejects any supplied certificate,
3042 which seems right, as the list of acceptable certificates is empty.
3044 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3045 gave an unhelpful defer error. I have not managed to make this reject any
3046 supplied certificates, but the error message it gives is "no certificate
3047 supplied", which is not helpful.
3049 70. exigrep's output now also includes lines that are not associated with any
3050 message, but which match the given pattern. Implemented by a patch from
3051 Martin Sluka, which also tidied up the Perl a bit.
3053 71. Recipient callout verification, like sender verification, was using <> in
3054 the MAIL FROM command. This isn't really the right thing, since the actual
3055 sender may affect whether the remote host accepts the recipient or not. I
3056 have changed it to use the actual sender in the callout; this means that
3057 the cache record is now keyed on a recipient/sender pair, not just the
3058 recipient address. There doesn't seem to be a real danger of callout loops,
3059 since a callout by the remote host to check the sender would use <>.
3060 [SEE ABOVE: changed after hitting problems.]
3062 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3063 temporary errors. However, in the case of such a code being given after
3064 the end of a data transmission (i.e. after ".") Exim was failing to write
3065 a retry record for the message. (Yes, there was some broken host that was
3066 actually sending 8xx at this point.)
3068 73. An unknown lookup type in a host list could cause Exim to panic-die when
3069 the list was checked. (An example that provoked this was putting <; in the
3070 middle of a list instead of at the start.) If this happened during a DATA
3071 ACL check, a -D file could be left lying around. This kind of configuration
3072 error no longer causes Exim to die; instead it causes a defer errror. The
3073 incident is still logged to the main and panic logs.
3075 74. Buglet left over from Exim 3 conversion. The message "too many messages
3076 in one connection" was written to the rejectlog but not the mainlog, except
3077 when address rewriting (yes!) was being logged.
3079 75. Added write_rejectlog option.
3081 76. When a system filter was run not as root (that is, when system_filter_user
3082 was set), the values of the $n variables were not being returned to the
3083 main process; thus, they were not subsequently available in the $sn
3086 77. Added +return_path_on_delivery log selector.
3088 78. A connection timeout was being treated differently from recipients deferred
3089 when testing hosts_max_try with a message that was older than the host's
3090 retry timeout. (The host should not be counted, thus allowing all hosts to
3091 be tried at least once before bouncing.) This may have been the cause of an
3092 occasionally reported bug whereby a message would remain on the queue
3093 longer than the retry timeout, but would be bounced if a delivery was
3094 forced. I say "may" because I never totally pinned down the problem;
3095 setting up timeout/retry tests is difficult. See also the next item.
3097 79. The ultimate address timeout was not being applied to errors that involved
3098 a combination of host plus message (for example, a timeout on a MAIL
3099 command). When an address resolved to a number of possible hosts, and they
3100 were not all tried for each delivery (e.g. because of hosts_max_try), a
3101 message could remain on the queue longer than the retry timeout.
3103 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3106 81. Fixed an obscure SMTP outgoing bug which required at least the following
3107 conditions: (a) there was another message waiting for the same server;
3108 (b) the server returned 5xx to all RCPT commands in the first message so
3109 that the message was not completed; (c) the server dropped the connection
3110 or gave a negative response to the RSET that Exim sends to abort the
3111 transaction. The observed case was a dropped connection after DATA that had
3112 been sent in pipelining mode. That is, the server had advertised PIPELINING
3113 but was not implementing it correctly. The effect of the bug was incorrect
3114 behaviour, such as trying another host, and this could lead to a crash.
3120 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3121 and daemon.c were passed as pointers to ints; they should have been
3122 pointers to socklen_t variables (which are typically unsigned ints).
3124 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3127 3. Fixed a really odd bug that affected only the testing scheme; patching a
3128 certain fixed string in the binary changed the value of another string that
3129 happened to be identical to the end of the original first string.
3131 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3132 name", it returns that address as the IP address. On some operating
3133 systems (e.g. Solaris), it also passes back the IP address string as the
3134 "host name". However, on others (e.g. Linux), it passes back an empty
3135 string. Exim wasn't checking for this, and was changing the host name to an
3136 empty string, assuming it had been canonicized.
3138 5. Although rare, it is permitted to have more than one PTR record for a given
3139 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3140 all the names associated with an address, because they do in Solaris.
3141 However, it seems that they do not in Linux for data that comes from the
3142 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3143 I found this out when I moved to a new Linux workstation and tried to run
3144 the Exim test suite.
3146 To get round this problem I have changed the code so that it now does its
3147 own call to the DNS to look up PTR records when searching for a host name.
3148 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3149 addresses that are only in /etc/hosts are still found.
3151 This behaviour is, however, controlled by an option called host_lookup_
3152 order, which defaults to "bydns:byaddr". If people want to use the other
3153 order, or indeed, just use one or the other means of lookup, they can
3154 specify it in this variable.
3156 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3157 some operating systems, this comes back from gethostbyaddr() as an empty
3158 string, and this is what Exim used to test for. However, it seems that in
3159 other systems, "." is yielded. Exim now tests for this case too.
3161 7. The values of check_spool_space and check_log_space are now held internally
3162 as a number of kilobytes instead of an absolute number of bytes. If a
3163 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3164 kilobyte. This means that much larger values can be stored.
3166 8. Exim monitor: an attempt to get the action menu when not actually pointing
3167 at a message produces an empty menu entitled "No message selected". This
3168 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3169 no entries in it ("Shell widget menu has zero width and/or height"). So I
3170 have added a single, blank menu entry in this case.
3172 9. Added ${quote_local_part.
3174 10. MIME decoding is now applied to the contents of Subject: header lines when
3177 11. Now that a reference to $sender_host_address automatically causes a reverse
3178 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3179 host lookup before query-style lookups in lists that might use this
3180 variable. This has therefore been abolished, and the "net-" prefix is no
3181 longer necessary for query-style lookups.
3183 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3184 have been a typo for LFLAGS, so it has been changed.
3186 13. The install script calls Exim with "-C /dev/null" in order to find the
3187 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3188 to be output. Howeve, since Exim outputs its version number before the
3189 error, it didn't break the script. It just looked ugly. I fixed this by
3190 always allowing "-C /dev/null" if the caller is root.
3192 14. Ignore overlarge ACL variable number when reading spool file - insurance
3193 against a later release with more variables having written the file.
3195 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3196 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3197 this, as well as the form without the "IPv6" on the front.
3199 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3200 OS/Makefile-Darwin file.
3202 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3203 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3205 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3206 variable. However, some OS use ints for inodes. I've added cast to long int
3207 to get rid of the compiler warning.
3209 19. I had forgotten to lock out "/../" in configuration file names when
3210 ALT_CONFIG_PREFIX was set.
3212 20. Routers used for verification do not need to specify transports. However,
3213 if such a router generated a host list, and callout was configured, Exim
3214 crashed, because it could not find a port number from the (non-existent)
3215 transport. It now assumes port 25 in this circumstance.
3217 21. Added the -t option to exigrep.
3219 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3220 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3221 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3222 them is set, LOOKUP_LSEARCH is forced.
3224 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3226 24. Added sender and host information to the "rejected by local_scan()" log
3227 line; previously there was no indication of these.
3229 25. Added .include_if_exists.
3231 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3232 Linux. It turns out that not all file systems support this. Apparently some
3233 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3234 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3235 not supported on the file descriptor, is now ignored when Exim is trying to
3236 sync a directory. This applies only to Linux.
3238 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3240 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3241 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3244 29. Added support for maildirsize files from supplied patch (modified a bit).
3246 30. The use of :fail: followed by an empty string could lead Exim to respond to
3247 sender verification failures with (e.g.):
3249 550 Verification failed for <xxx>
3250 550 Sender verify failed
3252 where the first response line was missing the '-' that indicates it is not
3253 the final line of the response.
3255 31. The loop for finding the name of the user that called Exim had a hardwired
3256 limit of 10; it now uses the value of finduser_retries, which is used for
3257 all other user lookups.
3259 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3262 33. Exim was neglecting to zero errno before one call of strtol() when
3263 expanding a string and expecting an integer value. On some systems this
3264 resulted in spurious "integer overflow" errors. Also, it was casting the
3265 result into an int without checking.
3267 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3268 did not work. The code looks as if it has *never* worked, though it appears
3269 to have been documented since at least releast 1.62. I have made it work.
3271 35. The "timeout_DNS" error in retry rules, also documented since at least
3272 1.62, also never worked. As it isn't clear exactly what this means, and
3273 clearly it isn't a major issue, I have abolished the feature by treating it
3274 as "timeout", and writing a warning to the main and panic logs.
3276 36. The display of retry rules for -brt wasn't always showing the error code
3279 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3280 timeout_connect_A, timeout_connect_MX.
3282 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3283 to the empty sender.
3285 39. The daemon was not analysing the content of -oX till after it had closed
3286 stderr and disconnected from the controlling terminal. This meant that any
3287 syntax errors were only noted on the panic log, and the return code from
3288 the command was 0. By re-arranging the code a little, I've made the
3289 decoding happen first, so such errors now appear on stderr, and the return
3290 code is 1. However, the actual setting up of the sockets still happens in
3291 the disconnected process, so errors there are still only recorded on the
3294 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3295 connections (as happens on some IP stacks) was logged at start up time as
3296 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3297 it from "IPv6 and IPv4", which means that two separate sockets are being
3300 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3301 says whether AF_INET or AF_INET6 was passed as an argument.
3303 42. Exiwhat output was messed up when time zones were included in log
3306 43. Exiwhat now gives more information about the daemon's listening ports,
3307 and whether -tls-on-connect was used.
3309 44. The "port" option of the smtp transport is now expanded.
3311 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3312 silently ignored. Now an error message is written to the main and panic
3315 46. There's a new ACL modifier called "logwrite" which writes to a log file
3316 as soon as it is encountered.
3318 47. Added $local_user_uid and $local_user_gid at routing time.
3320 48. Exim crashed when trying to verify a sender address that was being
3323 49. Exim was recognizing only a space character after ".include". It now also
3324 recognizes a tab character.
3326 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3327 extracting the relevant information from the specification. The man page no
3328 longer contains scrambled data for the -d option, and I've added a section
3329 at the front about calling Exim under different names.
3331 51. Added "extra_headers" argument to the "mail" command in filter files.
3333 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3336 53. Installed eximstats 1.29.
3338 54. Added transport_filter_timeout as a generic transport option.
3340 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3341 Cc: header lines. This was required by RFC 822, but it not required by RFC
3344 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3345 incoming messages that did not have them. Now it does so only if the
3346 message originates locally, that is, if there is no associated remote host
3347 address. When Resent- header lines are present, this applies to the Resent-
3348 lines rather than the non-Resent- lines.
3350 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3351 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3353 58. Messages for configuration errors now include the name of the main
3354 configuration file - useful now that there may be more than one file in a
3355 list (.included file names were always shown).
3357 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3358 for those rare installations that do not start the daemon as root or run it
3359 setuid root. I've cut out the call to initgroups() if the daemon is not
3362 60. The Exim user and group can now be bound into the binary as text strings
3363 that are looked up at the start of Exim's processing.
3365 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3367 62. Added $mailstore_basename variable.
3369 63. Installed patch to sieve.c from Michael Haardt.
3371 64. When Exim failed to open the panic log after failing to open the main log,
3372 the original message it was trying to log was written to stderr and debug
3373 output, but if they were not available (the usual case in production), it
3374 was lost. Now it is written to syslog before the two lines that record the
3375 failures to open the logs.
3377 65. Users' Exim filters run in subprocesses under the user's uid. It is
3378 possible for a "deliver" command or an alias in a "personal" command to
3379 provoke an address rewrite. If logging of address rewriting is configured,
3380 this fails because the process is not running as root or exim. There may be
3381 a better way of dealing with this, but for the moment (because 4.30 needs
3382 to be released), I have disabled address rewrite logging when running a
3383 filter in a non-root, non-exim process.
3389 1. The buildconfig auxiliary program wasn't quoting the value set for
3390 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3391 not defined. This bug was masked in 4.22 by the effect that was fixed in
3394 2. Some messages that were rejected after a message id was allocated were
3395 shown as "incomplete" by exigrep. It no longer does this for messages that
3396 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3398 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3399 have allow_domain_literals set, the ID did not get logged in the <= line.
3400 Domain literals are now always recognized in Message-ID: header lines.
3402 4. The first argument for a ${extract expansion item is the key name or field
3403 number. Leading and trailing spaces in this item were not being ignored,
3404 causing some misleading effects.
3406 5. When deliver_drop_privilege was set, single queue runner processes started
3407 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3408 same command in the process it spins off) were not dropping privilege.
3410 6. When the daemon running as "exim" started a queue runner, it always
3411 re-executed Exim in the spun-off process. This is a waste of effort when
3412 deliver_drop_privilege is set. The new process now just calls the
3413 queue-runner function directly.
3419 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3422 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3423 ignored. Though the use of -oP was forcing the writing of a pid file, it
3424 was always written to the default place.
3426 3. If the message "no IP address found for host xxxx" is generated during
3427 incoming verification, it is now followed by identification of the incoming
3428 connection (so you can more easily find what provoked it).
3430 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3432 5. Added some features to "harden" Exim a bit more against certain attacks:
3434 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3435 be put in Local/Makefile. This is like the never_users runtime option,
3436 but it cannot be overridden. The default setting is "root".
3438 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3439 prefix string with which any file named in a -C command line option
3442 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3443 is retained for -C and -D only if the caller of Exim is root. Without
3444 it, the exim user may also use -C and -D and retain privilege.
3446 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3447 command line option is disabled.
3449 6. Macro names set by the -D option must start with an upper case letter, just
3450 like macro names defined in the configuration file.
3452 7. Added "dereference=" facility to LDAP.
3454 8. Two instances of the typo "uknown" in the source files are fixed.
3456 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3457 the Configure-Makefile script screwed up while processing it.
3459 10. Incorporated PCRE 4.4.
3461 11. The SMTP synchronization check was not operating right at the start of an
3462 SMTP session. For example, it could not catch a HELO sent before the client
3463 waited for the greeting. There is now a check for outstanding input at the
3464 point when the greeting is written. Because of the duplex, asynchronous
3465 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3466 way, but not yet received, when the check is performed.
3468 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3469 on TCP/IP sockets, because this apparently causes some broken clients to
3472 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3473 unchanged) from the Cygwin maintainer.
3475 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3476 was supported instead of testing for mbx. Effectively a typo.
3478 15. The spa authenticator server code was not checking that the input it
3479 received was valid base64.
3481 16. The debug output line for the "set" modifier in ACLs was not showing the
3482 name of the variable that was being set.
3484 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3486 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3487 lookup on its own. Something else was needed to trigger the lookup. For
3488 example, a match in host_lookup or the need for a host name in a host list.
3489 Now, if $sender_host_name is referenced and the host name has not yet been
3490 looked up, a lookup is performed. If the lookup fails, the variable remains
3491 empty, and $host_lookup_failed is set to "1".
3493 19. Added "eqi" as a case-independent comparison operator.
3495 20. The saslauthd authentication condition could segfault if neither service
3496 nor realm was specified.
3498 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3499 error message that was logged was misleading, and incoming SMTP
3500 connections were dropped. The message is now more accurate, and temporary
3501 errors are given to SMTP connections.
3503 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3504 (except RSET) with a 421 error, until QUIT is received. However, it was
3505 failing to send a response to QUIT.
3507 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3508 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3509 ACL is now run after the helo_try_verify_hosts code.
3511 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3512 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3513 case-independent, so other case variants are also recognized). Apparently
3514 some systems use these upper case variants.
3516 25. If more than two messages were waiting for the same host, and a transport
3517 filter was specified for the transport, Exim sent two messages over the
3518 same TCP/IP connection, and then failed with "socket operation on non-
3519 socket" when it tried to send the third.
3521 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3523 27. The extern definition of crypt16() in expand.c was not being excluded when
3524 the OS had its own crypt16() function.
3526 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3527 as a preferred synonym for return_size_limit, both as an option and as an
3530 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3532 30. Changed the default configuration ACL to relax the local part checking rule
3533 for addresses that are not in any local domains. For these addresses,
3534 slashes and pipe symbols are allowed within local parts, but the sequence
3535 /../ is explicitly forbidden.
3537 31. SPA server authentication was not clearing the challenge buffer before
3540 32. log_message in a "warn" ACL statement was writing to the reject log as
3541 well as to the main log, which contradicts the documentation and doesn't
3542 seem right (because no rejection is happening). So I have stopped it.
3544 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3545 However, I am unable to do any testing of this.
3547 34. Fixed an infelicity in the appendfile transport. When checking directories
3548 for a mailbox, to see if any needed to be created, it was accidentally
3549 using path names with one or more superfluous leading slashes; tracing
3550 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3552 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3553 contents are added to the log line that is written for every discarded
3554 recipient. (Previously a log_message setting was ignored.)
3556 36. The ${quote: operator now quotes the string if it is empty.
3558 37. The install script runs exim in order to find its version number. If for
3559 some reason other than non-existence or emptiness, which it checks, it
3560 could not run './exim', it was installing it with an empty version number,
3561 i.e. as "exim-". This error state is now caught, and the installation is
3564 38. An argument was missing from the function that creates an error message
3565 when Exim fails to connect to the socket for saslauthd authentication.
3566 This could cause Exim to crash, or give a corrupted message.
3568 39. Added isip, isip4, and isip6 to ${if conditions.
3570 40. The ACL variables $acl_xx are now saved with the message, and can be
3571 accessed later in routers, transports, and filters.
3573 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3574 strings in the file are not string-expanded.
3576 42. If a MAIL command specified a SIZE value that was too large to fit into an
3577 int variable, the check against message_size_limit failed. Such values are
3578 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3579 day this will have to be increased, but I don't think I want to be around
3580 when emails are that large.
3587 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3588 iconv() is not standard in FreeBSD.
3590 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3591 IPv6 enabled. The observed symptom was a segmentation fault on return from
3592 the function os_common_find_running_interfaces() in src/os.c.
3594 3. In the check_special_case() function in daemon.c I had used "errno" as an
3595 argument name, which causes warnings on some systems. This was basically a
3596 typo, since it was named "eno" in the comments!
3598 4. The code that waits for the clock to tick (at a resolution of some fraction
3599 of a second) so as to ensure message-id uniqueness was always waiting for
3600 at least one whole tick, when it could have waited for less. [This is
3601 almost certainly not relevant at current processor speeds, where it is
3602 unlikely to ever wait at all. But we try to future-proof.]
3604 5. The function that sleeps for a time interval that includes fractions of a
3605 second contained a race. It did not block SIGALRM between setting the
3606 timer, and suspending (a couple of lines later). If the interval was short
3607 and the sigsuspend() was delayed until after it had expired, the suspension
3608 never ended. On busy systems this could lead to processes getting stuck for
3611 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3612 process, before it forks any delivery processes. The open lookup caching
3613 mechanism meant that the open file or database connection was passed into
3614 the delivery process. The problem was that delivery processes always tidy
3615 up cached lookup data. This could cause a problem for the next delivery
3616 process started by the queue runner, because the external queue runner
3617 process does not know about the closure. So the next delivery process
3618 still has data in the lookup cache. In the case of a file lookup, there was
3619 no problem because closing a file descriptor in a subprocess doesn't affect
3620 the parent. However, if the lookup was caching a connection to a database,
3621 the connection was closed, and the second delivery process was likely to
3622 see errors such as "PGSQL: query failed: server closed the connection
3623 unexpectedly". The problem has been fixed by closing all cached lookups
3624 in a queue runner before running a delivery process.
3626 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3627 seem to have the "const" qualifier which it has on other OS. I've
3630 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3631 *of the same type* (client or server) with the same public name that an
3632 error should be diagnosed.
3634 9. When Exim looked up a host name for an IP address, but failed to find the
3635 original IP address when looking up the host name (a safety check), it
3636 output the message "<ip address> does not match any IP for NULL", which was
3637 confusing, to say the least. The bug was that the host name should have
3638 appeared instead of "NULL".
3640 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3641 user that is built into the binary, and the -C or -D options is used, root
3642 privilege is dropped before the configuration file is read. In addition,
3643 logging is switched to stderr instead of the normal log files. If the
3644 configuration then re-defines the Exim user, the unprivileged environment
3645 is probably not what is expected, so Exim logs a panic warning message (but
3648 However, if deliver_drop_privilege is set, the unprivileged state may well
3649 be exactly what is intended, so the warning has been cut out in that case,
3650 and Exim is allowed to try to write to its normal log files.
3656 1. smtp_return_error_details was not giving details for temporary sender
3657 or receiver verification errors.
3659 2. Diagnose a configuration error if two authenticators have the same public
3662 3. Exim used not to create the message log file for a message until the first
3663 delivery attempt. This could be confusing when incoming messages were held
3664 for policy or load reasons. The message log file is now created at the time
3665 the message is received, and an initial "Received" line is written to it.
3667 4. The automatically generated man page for command line options had a minor
3668 bug that caused no ill effects; however, a more serious problem was that
3669 the procedure for building the man page automatically didn't always
3670 operate. Consequently, release 4.20 contains an out-of-date version. This
3671 shouldn't happen again.
3673 5. When building Exim with embedded Perl support, the script that builds the
3674 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3675 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3677 6. The freeze_tell option was not being used for messages that were frozen on
3678 arrival, either by an ACL or by local_scan().
3680 7. Added the smtp_incomplete_transaction log selector.
3682 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3683 was accepting AUTH without a new EHLO.
3685 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3686 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3689 10. Logging of TCP/IP connections (when configured) now happens in the main
3690 daemon process instead of the child process, so that the TCP/IP connection
3691 count is more accurate (but it can never be perfect).
3693 11. The use of "drop" in a nested ACL was not being handled correctly in the
3694 outer ACL. Now, if condition failure induced by the nested "drop" causes
3695 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3696 or "require"), the connection is dropped.
3698 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3699 that yield "discard" can now be used with an "accept" or a "discard" verb,
3700 but an error is generated for any others (because I can't see a useful way
3701 to define what should happen).
3703 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3704 are now processed in the same way as lines in the Exim configuration file.
3705 In particular, continuation lines are supported.
3707 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3709 15. Added -ti meaning -t -i.
3711 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3712 domains, and warn by logging if others are found.
3714 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3715 returned from GIFCONF when Exim is trying to find the list of interfaces on
3716 a host. The code in os.c has been modified to copy each ifreq to an aligned
3717 structure in all cases.
3719 Also, in some cases, the returned ifreq's were being copied to a 'struct
3720 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3721 means the last couple of bytes of an IPv6 address could be chopped if the
3722 ifreq contained only a normal sockaddr (14 bytes storage).
3724 18. Named domain lists were not supported in the hosts_treat_as_local option.
3725 An entry such as +xxxx was not recognized, and was treated as a literal
3728 19. Ensure that header lines added by a DATA ACL are included in the reject log
3729 if the ACL subsequently rejects the message.
3731 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3732 MD5 (which is deprecated).
3734 21. When testing a filter file using -bf, Exim was writing a message when it
3735 took the sender from a "From " line in the message, but it was not doing so
3736 when it took $return_path from a Return-Path: header line. It now does.
3738 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3739 with a valid header line field name (a series of printing characters
3740 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3742 23. Changed "disc" in the source to "disk" to conform to the documentation and
3743 the book and for uniformity.
3745 24. Ignore Sendmail's -Ooption=value command line item.
3747 25. When execve() failed while trying to run a command in a pipe transport,
3748 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3749 could be confused with a return value of 69 from the command itself. This
3750 has been changed to 127, the value the shell returns if it is asked to run
3751 a non-existent command. The wording for the related log line suggests a
3752 non-existent command as the problem.
3754 26. If received_header_text expands to an empty string, do not add a Received:
3755 header line to the message. (Well, it adds a token one on the spool, but
3756 marks it "old" so that it doesn't get used or transmitted.)
3758 27. Installed eximstats 1.28 (addition of -nt option).
3760 28. There was no check for failure on the call to getsockname() in the daemon
3761 code. This can fail if there is a shortage of resources on the system, with
3762 ENOMEM, for example. A temporary error is now given on failure.
3764 29. Contrary to the C standard, it seems that in some environments, the
3765 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3766 program. Exim now does this explicitly; it affects the formatting of
3767 timestamps using strftime().
3769 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3770 complaints. I've now initialized all the variables, to avoid this.
3772 32. Header lines added by a system filter were not being "seen" during
3773 transport-time rewrites.
3775 33. The info_callback() function passed to OpenSSL is set up with type void
3776 (*)(SSL *, int, int), as described somewhere. However, when calling the
3777 function (actually a macro) that sets it up, the type void(*)() is
3778 expected. I've put in a cast to prevent warnings from picky compilers.
3780 34. If a DNS black list lookup found a CNAME record, but there were no A
3781 records associated with the domain it pointed at, Exim crashed.
3783 35. If a DNS black list lookup returned more than one A record, Exim ignored
3784 all but the first. It now scans all returned addresses if a particular IP
3785 value is being sought. In this situation, the contents of the
3786 $dnslist_value variable are a list of all the addresses, separated by a
3789 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3790 to accept a host name and all its aliases if the forward lookup for any of
3791 them yielded the IP address of the incoming connection. Now it accepts only
3792 those names whose forward lookup yields the correct IP address. Any other
3793 names are discarded. This closes a loophole whereby a rogue DNS
3794 administrator could create reverse DNS records to break through a
3795 wildcarded host restriction in an ACL.
3797 37. If a user filter or a system filter that ran in a subprocess used any of
3798 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3799 the wrong values were passed to the pipe command ($thisaddress had the
3800 value of $0, $0 had the value of $1, etc). This bug was introduced by
3801 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3803 38. Improved the line breaking for long SMTP error messages from ACLs.
3804 Previously, if there was no break point between 40 and 75 characters, Exim
3805 left the rest of the message alone. Two changes have been made: (a) I've
3806 reduced the minimum length to 35 characters; (b) if it can't find a break
3807 point between 35 and 75 characters, it looks ahead and uses the first one
3808 that it finds. This may give the occasional overlong line, but at least the
3809 remaining text gets split now.
3811 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3812 file descriptors might be low, and that setting 1000 would always raise it.
3813 It turns out that in some environments, the limit is already over 1000 and
3814 that lowering it causes trouble. So now Exim takes care not to decrease it.
3816 40. When delivering a message, the value of $return_path is set to $sender_
3817 address at the start of routing (routers may change the value). By an
3818 oversight, this default was not being set up when an address was tested by
3819 -bt or -bv, which affected the outcome if any router or filter referred to
3822 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3823 log only when "message" or "log_message" are set. However, if one of the
3824 conditions was an address verification, or a call to a nested ACL, the
3825 messages generated by the underlying test were being passed through. This
3826 no longer happens. The underlying message is available in $acl_verify_
3827 message for both "message" and "log_message" expansions, so it can be
3828 passed through if needed.
3830 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3831 new expansion $bh_ to give the encoded byte string without charset
3832 translation. Translation happens only if iconv() is available; HAVE_ICONV
3833 indicates this at build time. HEADERS_CHARSET gives the charset to
3834 translate to; headers_charset can change it in the configuration, and
3835 "headers charset" can change it in an individual filter file.
3837 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3838 that creates RFC 2047 encoded "words" labels them as that charset instead
3839 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3840 expansion operator; (ii) when Exim creates a From: line for a local
3841 message; (iii) when a header line is rewritten to include a "phrase" part.
3843 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3844 buggy, causing it to skip the first lines of messages whose message ID
3845 ended in 'D'. This would not have bitten before Exim release 4.14, because
3846 message IDs were unlikely to end in 'D' before then. The effect was to have
3847 incorrect size information for certain domains.
3849 45. #include "config.h" was missing at the start of the crypt16.c module. This
3850 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3853 46. If there was a timeout during a "random" callout check, Exim treated it as
3854 a failure of the random address, and carried on sending RSET and the real
3855 address. If the delay was just some slowness somewhere, the response to the
3856 original RCPT would be taken as a response to RSET and so on, causing
3857 mayhem of various kinds.
3859 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3860 when I implemented it. It didn't allow for the fact that some option values
3861 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3862 the right test for positive values.
3864 48. Domain names in DNS records are case-independent. Exim always looks them up
3865 in lower case. Some resolvers return domain names in exactly the case they
3866 appear in the zone file, that is, they may contain uppercase letters. Not
3867 all resolvers do this - some return always lower case. Exim was treating a
3868 change of case by a resolver as a change of domain, similar to a widening
3869 of a domain abbreviation. This triggered its re-routing code and so it was
3870 trying to route what was effectively the same domain again. This normally
3871 caused routing to fail (because the router wouldn't handle the domain
3872 twice). Now Exim checks for this case specially, and just changes the
3873 casing of the domain that it ultimately uses when it transmits the message
3876 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3879 50. If a filter generated a file delivery with a non-absolute name (possible if
3880 no home directory exists for the router), the forbid_file option was not
3883 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3884 the existing equality matching.
3886 52. Exim was using ints instead of ino_t variables in some places where it was
3887 dealing with inode numbers.
3889 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3890 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3891 called TMPDIR, and if it finds it is different, it changes its value.
3893 54. The smtp_printf() function is now made available to local_scan() so
3894 additional output lines can be written before returning. There is also an
3895 smtp_fflush() function to enable the detection of a dropped connection.
3896 The variables smtp_input and smtp_batched_input are exported to
3899 55. Changed the default runtime configuration: the message "Unknown user"
3900 has been removed from the ACL, and instead placed on the localuser router,
3901 using the cannot_route_message feature. This means that any verification
3902 failures that generate their own messages won't get overridden. Similarly,
3903 the "Unrouteable address" message that was in the ACL for unverifiable
3904 relay addresses has also been removed.
3906 56. Added hosts_avoid_esmtp to the smtp transport.
3908 57. The exicyclog script was not checking for the esoteric option
3909 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3910 will work only if exicyclog is run under the appropriate euid.
3912 58. Following a discussion on the list, the rules by which Exim recognises line
3913 endings on incoming messages have been changed. The -dropcr and drop_cr
3914 options are now no-ops, retained only for backwards compatibility. The
3915 following line terminators are recognized: LF CRLF CR. However, special
3916 processing applies to CR:
3918 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3919 nor a local message in the state where . is a terminator.
3921 (ii) If a bare CR is encountered in a header line, an extra space is added
3922 after the line terminator so as not to end the header. The reasoning
3923 behind this is that bare CRs in header lines are most likely either
3924 to be mistakes, or people trying to play silly games.
3926 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3927 was being incorrectly given as 18 bytes larger than it should have been.
3928 This is a VOB (very old bug).
3930 60. This may never have affected anything current, but just in case it has:
3931 When the local host is found other than at the start of a list of hosts,
3932 the local host, those with the same MX, and any that follow, are discarded.
3933 When the list in question was part of a longer list of hosts, the following
3934 hosts (not currently being processed) were also being discarded. This no
3935 longer happens. I'm not sure if this situation could ever has previously
3938 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3941 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3942 header. This is something that is recommended in a new Internet Draft, and
3943 is something that is documented as being done by Sendmail. There are two
3944 possible values. For messages generated by the autoreply transport, Exim
3947 Auto-Submitted: auto-replied
3949 whereas for all other generated messages (e.g. bounces) it adds
3951 Auto-Submitted: auto-generated
3953 63. The "personal" condition in filters now includes a test for the
3954 Auto-Submitted: header. If it contains the string "auto-" the message it
3955 not considered personal.
3957 64. Added rcpt_include_affixes as a generic transport option.
3959 65. Added queue_only_override (default true).
3961 66. Added the syslog_duplication option.
3963 67. If what should have been the first header line of a message consisted of
3964 a space followed by a colon, Exim was mis-interpreting it as a header line.
3965 It isn't of course - it is syntactically invalid and should therefore be
3966 treated as the start of the message body. The misbehaviour could have
3967 caused a number of strange effects, including loss of data in subsequent
3968 header lines, and spool format errors.
3970 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3971 client host had authenticated. This control can now be exercised by an ACL
3972 for more flexibility.
3974 69. By default, callouts do not happen when testing with -bh. There is now a
3975 variant, -bhc, which does actually run the callout code, including
3976 consulting and updating the callout cache.
3978 70. Added support for saslauthd authentication, courtesy of Alexander
3981 71. If statvfs() failed on the spool or log directories while checking their
3982 size for availability, Exim confusingly gave the error "space shortage".
3983 Furthermore, in debugging mode it crashed with a floating point exception.
3984 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3985 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3986 serious problem, Exim now writes to the main and panic logs when this
3987 happens, with details of the failure. It then refuses to accept the
3988 incoming message, giving the message "spool directory problem" or "log
3989 directory problem" with a 421 code for SMTP messages.
3991 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3992 0, 1, and 2 exist, because some OS complain for execs without them (see
3993 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3994 descriptors. However, the code omitted to check that the open succeeded,
3995 causing mysterious errors if for some reason the permissions on /dev/null
3996 got screwed. Now Exim writes a message to the main and panic logs, and
3997 bombs out if it can't open /dev/null.
3999 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4000 interact so that it is all more flexible. It is supposed to remain
4001 backwards compatible. Also added extra_local_interfaces.
4003 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4004 to bomb out with an assertion failure - to the client this appears as a
4005 connection drop. This problem occurs in the part of the code that was taken
4006 from the Samba project. Fortunately, the assertion is in a very simple
4007 function, so I have fixed this by reproducing the function inline in the
4008 one place where it is called, and arranging for authentication to fail
4009 instead of killing the process with assert().
4011 75. The SPA client code was not working when the server requested OEM rather
4012 than Unicode encoding.
4014 76. Added code to make require_files with a specific uid setting more usable in
4015 the case where statting the file as root fails - usually a non-root-mounted
4016 NFS file system. When this happens and the failure is EACCES, Exim now
4017 forks a subprocess and does the per-uid checking as the relevant uid.
4019 77. Added process_log_path.
4021 78. If log_file_path was not explicitly set, a setting of check_log_space or
4022 check_log_inodes was ignored.
4024 79. If a space check for the spool or log partitions fails, the incident is now
4025 logged. Of course, in the latter case the data may get lost...
4027 80. Added the %p formatting code to string_format() so that it can be used to
4028 print addresses in debug_print(). Adjusted all the address printing in the
4029 debugging in store.c to use %p rather than %d.
4031 81. There was a concern that a line of code in smtp_in.c could overflow a
4032 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4033 initially expressed, the concern was not well-founded, because trailing
4034 spaces are removed early. However, if the trailing spaces were followed by
4035 a NULL, they did not get removed, so the overflow was possible. Two fixes
4038 (a) I re-wrote the offending code in a cleaner fashion.
4039 (b) If an incoming SMTP command contains a NULL character, it is rejected
4042 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4043 runs initgroups(), so that if the Exim user is in any additional groups,
4044 they will be used during message reception.
4050 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob