1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.363 2006/06/28 16:00:23 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.
63 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
64 other effects) broke the use of negated acl sub-conditions.
66 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
69 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
70 "Deny" causes Exim to reject the incoming connection with a 554 error.
71 Unfortunately, if there is a major crisis, such as a disk failure,
72 tcp-wrappers gives "deny", whereas what one would like would be some
73 kind of temporary error. A kludge has been added to help with this.
74 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
75 554 error is used if errno is still zero or contains ENOENT (which occurs
76 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
79 PH/03 Add -lutil to the default FreeBSD LIBS setting.
81 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
82 errors. Otherwise a message that provokes a temporary error (when other
83 messages do not) can cause a whole host to time out.
85 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
86 addresses were routed directly to files or pipes from a redirect router.
87 File deliveries just didn't batch; pipe deliveries might have suffered
90 PH/06 A failure to get a lock for a hints database would erroneously always say
91 "Failed to get write lock", even when it was really a read lock.
93 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
94 of 0600. This has been changed to use the value of the lockfile_mode
95 option (which defaults to 0600).
97 PH/08 Applied small patch from the Sieve maintainer.
99 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
100 folder from quota calculations, a direct delivery into this folder messed
101 up the contents of the maildirsize file. This was because the regex was
102 used only to exclude .Trash (or whatever) when the size of the mailbox
103 was calculated. There was no check that a delivery was happening into an
104 excluded directory. This bug has been fixed by ignoring all quota
105 processing for deliveries into excluded directories.
107 PH/10 Added the maildirfolder_create_regex option to appendfile.
113 PH/01 The code for finding all the local interface addresses on a FreeBSD
114 system running IPv6 was broken. This may well have applied to all BSD
115 systems, as well as to others that have similar system calls. The broken
116 code found IPv4 interfaces correctly, but gave incorrect values for the
117 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
118 that it would not match correctly against @[] and not recognize the IPv6
121 PH/02 The ipliteral router was not recognizing addresses of the form user@
122 [ipv6:....] because it didn't know about the "ipv6:" prefix.
124 PH/03 Added disable_ipv6.
126 PH/04 Changed $reply_address to use the raw form of the headers instead of the
127 decoded form, because it is most often used to construct To: headers
128 lines in autoreplies, and the decoded form may well be syntactically
129 invalid. However, $reply_address has leading white space removed, and all
130 newlines turned into spaces so that the autoreply transport does not
133 PH/05 If group was specified without a user on a router, and no group or user
134 was specified on a transport, the group from the router was ignored.
136 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
137 for visible compile-time settings that can be used to change these
138 numbers, for those that want even more. Backwards compatibility with old
139 spool files has been maintained. However, going back to a previous Exim
140 release will lost any variables that are in spool files.
142 PH/07 Two small changes when running in the test harness: increase delay when
143 passing a TCP/IP connection to a new process, in case the original
144 process has to generate a bounce, and remove special handling of
145 127.0.0.2 (sic), which is no longer necessary.
147 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
148 be the same on different OS.
150 PH/09 Moved a debug statement in filter processing to avoid a race problem when
153 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
154 whether --show-vars was specified or not
156 JJ/02 exipick: Added support for new ACL variable spool format introduced
159 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
160 syntactically invalid From: or Reply-to: line, and a filter used this to
161 generate an autoreply, and therefore failed to obtain an address for the
162 autoreply, Exim could try to deliver to a non-existent relative file
163 name, causing unrelated and misleading errors. What now happens is that
164 it logs this as a hard delivery error, but does not attempt to create a
167 PH/11 The exinext utility has a -C option for testing purposes, but although
168 the given file was scanned by exinext itself; it wasn't being passed on
169 when Exim was called.
171 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
172 an end-of-file indication when reading a command response.
174 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
175 compiled. In many other places in Exim, IPv6 addresses are always
176 recognized, so I have changed this. It also means that IPv4 domain
177 literals of the form [IPV4:n.n.n.n] are now always recognized.
179 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
180 used if the router is not running as root, for example, when verifying at
181 ACL time, or when using -bh. The debugging output from this situation was
182 non-existent - all you got was a failure to exec. I have made two
185 (a) Failures to set uid/gid, the current directory, or a process leader
186 in a subprocess such as that created by queryprogram now generate
187 suitable debugging ouput when -d is set.
189 (b) The queryprogram router detects when it is not running as root,
190 outputs suitable debugging information if -d is set, and then runs
191 the subprocess without attempting to change uid/gid.
193 PH/15 Minor change to Makefile for building test_host (undocumented testing
196 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
197 additional section of a DNS packet that returns MX or SRV records.
198 Instead, it always explicitly searches for A/AAAA records. This avoids
199 major problems that occur when a DNS server includes only records of one
200 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
201 fixed another bug: if SRV records were looked up and the corresponding
202 address records were *not* found in the additional section, the port
203 values from the SRV records were lost.
205 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
206 using the correct key (the original address) when searching the retry
207 rules in order to find which one to use for generating the retry hint.
209 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
210 adding the default one. Similarly, if it contains a Reply-To: header, the
211 errors_reply_to option, if set, is not used.
213 PH/19 When calculating a retry time, Exim used to measure the "time since
214 failure" by looking at the "first failed" field in the retry record. Now
215 it does not use this if it is later than than the arrival time of the
216 message. Instead it uses the arrival time. This makes for better
217 behaviour in cases where some deliveries succeed, thus re-setting the
218 "first failed" field. An example is a quota failure for a huge message
219 when small messages continue to be delivered. Without this change, the
220 "time since failure" will always be short, possible causing more frequent
221 delivery attempts for the huge message than are intended.
222 [Note: This change was subsequently modified - see PH/04 for 4.62.]
224 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
225 $1, $2, $3) because the numerical variables can be reset during some
226 expansion items (e.g. "match"), thereby losing the authentication data.
228 PH/21 Make -bV show the size of off_t variables so that the test suite can
229 decide whether to run tests for quotas > 2G.
231 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
232 mailbox_size, and mailbox_filecount in the appendfile transport. If a
233 filecount value is greater than 2G or if a quota value is greater than 2G
234 on a system where the size of off_t is not greater than 4, a panic error
237 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
238 never match. The debug and -bh output now contains an explicit error
239 message indicating a malformed IPv4 address or mask.
241 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
242 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
245 PH/25 Do not write to syslog when running in the test harness. The only
246 occasion when this arises is a failure to open the main or panic logs
247 (for which there is an explicit test).
249 PH/26 Added the /no_tell option to "control=freeze".
251 PH/27 If a host name lookup failed very early in a connection, for example, if
252 the IP address matched host_lookup and the reverse lookup yielded a name
253 that did not have a forward lookup, an error message of the form "no IP
254 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
255 could be logged. Now it outputs the IP address instead of "NULL".
257 PH/28 An enabling patch from MH: add new function child_open_exim2() which
258 allows the sender and the authenticated sender to be set when
259 submitting a message from within Exim. Since child_open_exim() is
260 documented for local_scan(), the new function should be too.
262 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
263 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
264 results in an empty string is now treated as unset.
266 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
268 PH/31 Added sender_verify_fail logging option.
270 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
271 needed by RFC 822 but not by RFC 2822 was commented out. I have now
272 tidied the source and removed it altogether.
274 PH/33 When a queue run was abandoned because the load average was too high, a
275 log line was always written; now it is written only if the queue_run log
276 selector is set. In addition, the log line for abandonment now contains
277 information about the queue run such as the pid. This is always present
278 in "start" and "stop" lines but was omitted from the "abandon" line.
280 PH/34 Omit spaces between a header name and the colon in the error message that
281 is given when verify = headers_syntax fails (if there are lots of them,
282 the message gets confusing).
284 PH/35 Change the default for dns_check_names_pattern to allow slashes within
285 names, as there are now some PTR records that contain slashes. This check
286 is only to protect against broken name servers that fall over on strange
287 characters, so the fact that it applies to all lookups doesn't matter.
289 PH/36 Now that the new test suite is complete, we can remove some of the
290 special code in Exim that was needed for the old test suite. For example,
291 sorting DNS records because real resolvers return them in an arbitrary
292 order. The new test suite's fake resolver always returns records in the
295 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
296 bounces) except when queue_only is set, to avoid logging races between
297 the different processes.
299 PH/38 Panic-die if .include specifies a non-absolute path.
301 PH/39 A tweak to the "H" retry rule from its user.
303 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
304 a label. They prevented compilation on older perls.
306 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
307 a warning to be raised on newish perls.
309 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
310 on queue. Changes to match documented behaviour of showing count of
311 messages matching specified criteria.
313 PH/40 Changed the default ident timeout from 30s to 5s.
315 PH/41 Added support for the use of login_cap features, on those BSD systems
316 that have them, for controlling the resources used by pipe deliveries.
318 PH/42 The content-scanning code uses fopen() to create files in which to put
319 message data. Previously it was not paying any attention to the mode of
320 the files. Exim runs with umask(0) because the rest of the code creates
321 files with open(), and sets the required mode explicitly. Thus, these
322 files were ending up world-writeable. This was not a big issue, because,
323 being within the spool directory, they were not world-accessible. I have
324 created a function called modefopen, which takes an additional mode
325 argument. It sets umask(777), creates the file, chmods it to the required
326 mode, then resets the umask. All the relevant calls to fopen() in the
327 content scanning code have been changed to use this function.
329 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
330 to 24 hours. This avoids potential overflow problems when processing G
331 and H retry rules. I suspect nobody ever tinkers with this value.
333 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
335 PH/45 When the plaintext authenticator is running as a client, the server's
336 challenges are checked to ensure they are valid base64 strings. By
337 default, the authentication attempt is cancelled if an invalid string is
338 received. Setting client_ignore_invalid_base64 true ignores these errors.
339 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
340 they are received. Thus, the responses can be made to depend on the
341 challenges. If an invalid string is ignored, an empty string is placed in
344 PH/46 Messages that are created by the autoreply transport now contains a
345 References: header, in accordance with RFCs 2822 and 3834.
347 PH/47 Added authenticated_sender_force to the smtp transport.
349 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
351 PH/49 Installed latest patch from the Sieve maintainer.
353 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
354 was also set, the appendfile transport was unnecessarily scanning a
355 directory of message files (e.g. for maildir delivery) to find the count
356 of files (along with the size), even though it did not need this
357 information. It now does the scan only if it needs to find either the
358 size of the count of files.
360 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
362 PH/52 Two bugs concerned with error handling when the smtp transport is
365 (i) Exim was not creating retry information for temporary errors given
366 for individual recipients after the DATA command when the smtp transport
367 was used in LMTP mode. This meant that they could be retried too
368 frequently, and not timed out correctly.
370 (ii) Exim was setting the flag that allows error details to be returned
371 for LMTP errors on RCPT commands, but not for LMTP errors for individual
372 recipients that were returned after the DATA command.
374 PH/53 This is related to PH/52, but is more general: for any failing address,
375 when detailed error information was permitted to be returned to the
376 sender, but the error was temporary, then after the final timeout, only
377 "retry timeout exceeded" was returned. Now it returns the full error as
378 well as "retry timeout exceeded".
380 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
381 do this, and (what is worse) MTAs that accept it.
383 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
384 will now be deprecated.
386 PH/56 New os.c-cygwin from the Cygwin maintainer.
388 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
389 formats (previously only available in exim formats via -bpr, -bpru,
390 and -bpra. Now also available in native and exiqgrep formats)
392 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
393 with very large, slow to parse queues
395 JJ/08 exipick: added ! as generic prefix to negate any criteria format
397 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
399 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
400 responses to authentication challenges, though it was showing the
401 challenges; (ii) I've removed the CR characters from the debug output for
404 PH/58 Allow for the insertion of a newline as well as a space when a string
405 is turned into more than one encoded-word during RFC 2047 encoding. The
406 Sieve code now uses this.
408 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
409 data_4xx, lost_connection, tls_required.
411 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
412 message was being sent as an SMTP response.
414 PH/61 Add -l and -k options to exicyclog.
416 PH/62 When verifying, if an address was redirected to one new address, so that
417 verification continued, and the new address failed or deferred after
418 having set something in $address_data, the value of $address_data was not
419 passed back to the ACL. This was different to the case when no
420 redirection occurred. The value is now passed back in both cases.
422 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
423 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
424 use login_cap.h, so on its own it isn't the distinguishing feature. The
425 new name refers directly to the setclassresources() function.
427 PH/65 Added configuration files for NetBSD3.
429 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
431 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
432 is preferred over IPv4.
434 PH/68 The bounce_return_message and bounce_return_body options were not being
435 honoured for bounces generated during the reception of non-SMTP messages.
436 In particular, this applied to messages rejected by the ACL. This bug has
437 been fixed. However, if bounce_return_message is true and bounce_return_
438 body is false, the headers that are returned for a non-SMTP message
439 include only those that have been read before the error was detected.
440 (In the case of an ACL rejection, they have all been read.)
442 PH/69 The HTML version of the specification is now built in a directory called
443 spec_html instead of spec.html, because the latter looks like a path with
444 a MIME-type, and this confuses some software.
446 PH/70 Catch two compiler warnings in sieve.c.
448 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
449 function verify_get_ident() calls ip_connect() to connect a socket, but
450 if the "connect()" function timed out, ip_connect() used to close the
451 socket. However, verify_get_ident() also closes the socket later, and in
452 between Exim writes to the log, which may get opened at this point. When
453 the socket was closed in ip_connect(), the log could get the same file
454 descriptor number as the socket. This naturally causes chaos. The fix is
455 not to close the socket in ip_connect(); the socket should be closed by
456 the function that creates it. There was only one place in the code where
457 this was missing, in the iplookup router, which I don't think anybody now
458 uses, but I've fixed it anyway.
460 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
461 well as to direct DNS lookups. Otherwise the handling of names in host
462 lists is inconsistent and therefore confusing.
468 PH/01 Two changes to the default runtime configuration:
470 (1) Move the checks for relay_from_hosts and authenticated clients from
471 after to before the (commented out) DNS black list checks.
473 (2) Add control=submission to the relay_from_hosts and authenticated
474 clients checks, on the grounds that messages accepted by these
475 statements are most likely to be submissions.
477 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
479 (1) Generate an error if the third argument for the ${prvs expansion is
482 (2) Treat a missing third argument of ${prvscheck as if it were an empty
485 (3) Reset the variables that are obtained from the first argument of
486 ${prvscheck and used in the second argument before leaving the code,
487 because their memory is reclaimed, so using them afterwards may do
490 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
491 one (it's much easier than Tom thought :-).
493 (5) Because of (4), we can now allow for the use of $prvscheck_result
494 inside the third argument.
496 PH/03 For some reason, the default setting of PATH when running a command from
497 a pipe transport was just "/usr/bin". I have changed it to
500 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
501 anything to be listed in the output from -bV.
503 PH/05 When a filter generated an autoreply, the entire To: header line was
504 quoted in the delivery log line, like this:
506 => >A.N.Other <ano@some.domain> <original@ddress> ...
508 This has been changed so that it extracts the operative address. There
509 may be more than one such address. If so, they are comma-separated, like
512 => >ano@some.domain,ona@other.domain <original@ddress> ...
514 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
515 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
516 not being looked up in the rDNS to get a host name, Exim was showing the
517 IP address twice in Received: lines, even though the IP addresses were
518 identical. For example:
520 Received: from [1.2.3.4] (helo=[1.2.3.4])
522 However, if the real host name was known, it was omitting the HELO data
523 if it matched the actual IP address. This has been tidied up so that it
524 doesn't show the same IP address twice.
526 PH/07 When both +timestamp and +memory debugging was on, the value given by
527 $tod_xxx expansions could be wrong, because the tod_stamp() function was
528 called by the debug printing, thereby overwriting the timestamp buffer.
529 Debugging no longer uses the tod_stamp() function when +timestamp is set.
531 PH/08 When the original message was included in an autoreply transport, it
532 always said "this is a copy of the message, including all the headers",
533 even if body_only or headers_only was set. It now gives an appropriate
536 PH/09 Applied a patch from the Sieve maintainer which:
538 o fixes some comments
539 o adds the (disabled) notify extension core
540 o adds some debug output for the result of if/elsif tests
541 o points to the current vacation draft in the documentation
542 and documents the missing references header update
546 o fixes a bug in processing the envelope test (when testing
547 multiple envelope elements, the last element determinted the
550 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
551 Electronic Mail") by including:
553 Auto-submitted: auto-generated
555 in the messages that it generates (bounce messages and others, such as
556 warnings). In the case of bounce messages for non-SMTP mesages, there was
557 also a typo: it was using "Auto_submitted" (underscore instead of
558 hyphen). Since every message generated by Exim is necessarily in response
559 to another message, thes have all been changed to:
561 Auto-Submitted: auto-replied
563 in accordance with these statements in the RFC:
565 The auto-replied keyword:
567 - SHOULD be used on messages sent in direct response to another
568 message by an automatic process,
570 - MUST NOT be used on manually-generated messages,
572 - MAY be used on Delivery Status Notifications (DSNs) and Message
573 Disposition Notifications (MDNs),
575 - MUST NOT be used on messages generated by automatic or periodic
576 processes, except for messages which are automatic responses to
579 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
580 to the default Received: header definition.
582 PH/12 Added log selector acl_warn_skipped (default on).
584 PH/13 After a successful wildlsearch lookup, discard the values of numeric
585 variables because (a) they are in the wrong storage pool and (b) even if
586 they were copied, it wouldn't work properly because of the caching.
588 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
589 checking when decoding. Apparently there are clients that generate
590 overlong encoded strings. Why am I not surprised?
592 PH/15 If the first argument of "${if match_address" was not empty, but did not
593 contain an "@" character, Exim crashed. Now it writes a panic log message
594 and treats the condition as false.
596 PH/16 In autoreply, treat an empty string for "once" the same as unset.
598 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
599 extension "envelope-auth". The code is finished and in agreement with
600 other implementations, but there is no documentation so far and in fact,
601 nobody wrote the draft yet. This extension is currently #undef'ed, thus
602 not changing the active code.
604 Print executed "if" and "elsif" statements when debugging is used. This
605 helps a great deal to understand what a filter does.
607 Document more things not specified clearly in RFC3028. I had all this
608 sorted out, when out of a sudden new issues came to my mind. Oops."
610 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
613 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
614 canonical form (as documented). However, after a host name lookup from
615 the IP address, check_host() was doing a simple string comparison with
616 addresses acquired from the DNS when checking that the found name did
617 have the original IP as one of its addresses. Since any found IPv6
618 addresses are likely to be in abbreviated form, the comparison could
619 fail. Luckily, there already exists a function for doing the comparison
620 by converting both addresses to binary, so now that is used instead of
623 PH/20 There was another similar case to PH/19, when a complete host name was
624 given in a host list; looking up its IP address could give an abbreviated
625 form, whereas the current host's name might or might not be abbreviated.
626 The same fix has been applied.
632 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
633 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
636 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
637 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
639 PH/03 Typo: missing ".o" in src/pcre/Makefile.
641 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
642 header line, restrict the check to what is listed in RFCs 2369 and 2929.
643 Also, for "Auto-Submitted", treat anything other than "no" as
644 non-personal, in accordance with RFC 3834. (Previously it treated
645 anything starting "auto-" as non-personal.)
647 TF/01 The control=submission/name=... option had a problem with syntax
648 errors if the name included a slash character. The /name= option
649 now slurps the rest of the string, so it can include any characters
650 but it must come last in the list of options (after /sender_retain
653 PH/05 Some modifications to the interface to the fake nameserver for the new
661 TK/01 Added the "success_on_redirect" address verification option. See
662 NewStuff for rationale and an example.
664 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
666 PH/02 Patch to exigrep to allow it to work on syslog lines.
668 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
669 fread() to skip over the body file's header line, because in Cygwin the
670 header line is locked and is inaccessible.
672 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
673 co-exist for some time) to make it clear that it is the Exim ID that is
674 referenced, not the Message-ID: header line.
676 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
677 string_format() function, because snprintf() does not exist on all
680 PH/06 The use of forbid_filter_existstest now also locks out the use of the
681 ${stat: expansion item.
683 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
684 protocol synchronization error", to keep the pedants happy.
686 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
687 well as for IRIX systems, when gcc is being used. See the host.c source
690 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
692 PH/10 Named domain lists were not working if used in a queue_smtp_domains
695 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
696 transport and to the smtp transport in LMTP mode.
698 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
700 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
701 run a filter in a subprocess. This could lead to confusion in subsequent
702 lookups in the parent process. There should also be a search_tidyup() at
703 the end of the subprocess.
705 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
706 only if the host matched helo_try_verify_hosts, which caused the
707 verification to occur when the EHLO/HELO command was issued. The ACL just
708 tested the remembered result. Now, if a previous verification attempt has
709 not happened, "verify = helo" does it there and then.
711 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
713 TK/03 Fix log output including CR from clamd.
715 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
716 exist provoked a memory error which could cause a segfault.
718 PH/15 Installed PCRE 6.2
720 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
722 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
723 of the problem. Specifically, suggested +O2 rather than +O1 for the
726 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
728 PH/20 If a delivery was routed to a non-standard port by means of an SRV
729 record, the port was not correctly logged when the outgoing_port log
730 selector was set (it logged the transort's default port).
732 PH/21 Added support for host-specific ports to manualroute, queryprogram,
733 fallback_hosts, and "hosts" in the smtp transport.
735 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
736 host errors such as "Connection refused".
738 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
739 authentication with radiusclient 0.4.9:
741 - Error returned from rc_read_config was caught wrongly
742 - Username/password not passed on to radius server due to wrong length.
744 The presumption is that some radiusclient API changes for 4.51/PH/17
745 were not taken care of correctly. The code is still untested by me (my
746 Linux distribution still has 0.3.2 of radiusclient), but it was
747 contributed by a Radius user.
749 PH/24 When doing a callout, the value of $domain wasn't set correctly when
750 expanding the "port" option of the smtp transport.
752 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
753 while reading a MIME header. Thanks to Tom Hughes for a patch.
755 PH/24 Include config.h inside local_scan.h so that configuration settings are
758 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
759 that in an ACL for RCPT (for example), you can examine exactly what was
762 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
763 commands, but it was not correctly comparing the address with the actual
764 client host address. Thus, it would show the EHLO address in Received:
765 header lines when this was not necessary.
767 PH/27 Added the % operator to ${eval:}.
769 PH/28 Exim tries to create and chdir to its spool directory when it starts;
770 it should be ignoring failures (because with -C, for example, it has lost
771 privilege). It wasn't ignoring creation failures other than "already
774 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
777 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
778 because an input line was too long, either on its own, or by virtue of
779 too many continuations, the temporary file was not being removed, and the
780 return code was incorrect.
782 PH/31 Missing "BOOL" in function definition in filtertest.c.
784 PH/32 Applied Sieve patches from the maintainer.
786 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
788 PH/33 Added "verify = not_blind".
790 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
791 Local/Makefile (with some defaults set). These are used in built scripts
792 such as exicyclog, but they have never been used in the exim_install
793 script (though there are many overriding facilities there). I have
794 arranged that the exim_install script now takes note of these two
797 PH/35 Installed configuration files for Dragonfly.
799 PH/36 When a locally submitted message by a trusted user did not contain a
800 From: header, and the sender address was obtained from -f or from an SMTP
801 MAIL command, and the trusted user did not use -F to supply a sender
802 name, $originator_name was incorrectly used when constructing a From:
803 header. Furthermore, $originator_name was used for submission mode
804 messages from external hosts without From: headers in a similar way,
805 which is clearly wrong.
807 PH/37 Added control=suppress_local_fixups.
809 PH/38 When log_selector = +received_sender was set, and the addition of the
810 sender made the log line's construction buffer exactly full, or one byte
811 less than full, an overflow happened when the terminating "\n" was
814 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
815 when the result of a list match is failure because a DNS lookup failed.
817 PH/40 RM_COMMAND is now used in the building process.
819 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
820 the "build-* directories that it finds.
822 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
823 address in a domain literal was a prefix of an interface address.
825 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
826 when verifying a sender address, unless rewrite_headers is false.
828 PH/44 Wrote a long comment about why errors_to addresses are verified as
829 recipients, not senders.
831 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
832 the ratelimit ACL was added.
834 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
836 PH/46 Added extra information about PostgreSQL errors to the error string.
838 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
839 avoiding the need to install special zones in a real server. This is
840 backwards compatible; if it can't find the fake resolver, it drops back.
841 Thus, both old and new test suites can be run.
843 TF/02 Added util/ratelimit.pl
845 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
846 clock is set back in time.
848 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
849 Candler <B.Candler@pobox.com>.
851 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
852 OK for addresses that are the result of redirections.
854 PH/48 A number of further additions for the benefit of the new test suite,
855 including a fake gethostbyname() that interfaces to the fake DNS resolver
858 TF/06 The fix for widen_domains has also been applied to qualify_single and
859 search_parents which are the other dnslookup options that can cause
862 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
865 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
867 TF/07 Exim produced the error message "an SRV record indicated no SMTP
868 service" if it encountered an MX record with an empty target hostname.
869 The message is now "an MX or SRV record indicated no SMTP service".
871 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
872 if the DNS of the sending site is misconfigured. This is quite a
873 common situation. This change restores the behaviour of treating a
874 helo verification defer as a failure.
876 PH/51 If self=fail was set on a router, the bounce message did not include the
877 actual error message.
883 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
885 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
886 command itself ended in error, the underlying message about the transport
887 filter timeout was being overwritten with the pipe command error. Now the
888 underlying error message should be appended to the second error message.
890 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
892 PH/02 Reduce the amount of output that "make" produces by default. Full output
893 can still be requested.
895 PH/03 The warning log line about a condition test deferring for a "warn" verb
896 was being output only once per connection, rather than after each
897 occurrence (because it was using the same function as for successful
898 "warn" verbs). This seems wrong, so I have changed it.
900 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
901 it should not have, which might have caused a crash in the right
902 circumstances, but probably never did.
904 PH/04 Installed a modified version of Tony Finch's patch to make submission
905 mode fix the return path as well as the Sender: header line, and to
906 add a /name= option so that you can make the user's friendly name appear
909 TF/03 Added the control = fakedefer ACL modifier.
911 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
912 Mark Lowes for thorough testing.
914 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
916 TK/03 Merged latest SRS patch from Miles Wilton.
918 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
919 with the definition in sysexits.h (which is #included earlier).
920 Fortunately, Exim does not actually use EX_OK. The code used to try to
921 preserve the sysexits.h value, by assumimg that macro definitions were
922 scanned for macro replacements. I have been disabused of this notion,
923 so now the code just undefines EX_OK before #including unistd.h.
925 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
926 in the smtp transport. When a block could not be written in a single
927 write() function, the timeout was being re-applied to each part-write.
928 This seems wrong - if the receiver was accepting one byte at a time it
929 would take for ever. The timeout is now adjusted when this happens. It
930 doesn't have to be particularly precise.
932 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
933 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
935 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
936 without a domain if the check to <postmaster@domain> fails.
938 SC/01 Eximstats: added -xls and the ability to specify output files
939 (patch written by Frank Heydlauf).
941 SC/02 Eximstats: use FileHandles for outputing results.
943 SC/03 Eximstats: allow any combination of xls, txt, and html output.
945 SC/04 Eximstats: fixed display of large numbers with -nvr option
947 SC/05 Eximstats: fixed merging of reports with empty tables.
949 SC/06 Eximstats: added the -include_original_destination flag
951 SC/07 Eximstats: removed tabs and trailing whitespace.
953 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
955 TK/06 MBOX spool code: Add real "From " MBOX separator line
956 so the .eml file is really in mbox format (even though
957 most programs do not really care). Patch from Alex Miller.
959 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
960 The latter is generated from $received_to and is only set if the
961 message has one envelope recipient. SA can use these headers,
962 obviously out-of-the-box. Patch from Alex Miller.
964 PH/08 The ${def test on a variable was returning false if the variable's
965 value was "0", contrary to what the specification has always said!
966 The result should be true unless the variable is empty.
968 PH/09 The syntax error of a character other than { following "${if
969 def:variable_name" (after optional whitespace) was not being diagnosed.
970 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
971 accidental colon was present, for example, could give incorrect results.
973 PH/10 Tidied the code in a number of places where the st_size field of a stat()
974 result is used (not including appendfile, where other changes are about
977 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
978 This involved changing a lot of size variables from int to off_t. It
979 should work with maildirs and everything.
981 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
982 spamd dying while we are connected to it.
984 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
987 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
988 function for BDB 4.3.
990 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
992 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
993 the sources for PCRE have been re-arranged and more files are now
996 PH/15 The code I had for printing potentially long long variables in PH/11
997 above was not the best (it lost precision). The length of off_t variables
998 is now inspected at build time, and an appropriate printing format (%ld
999 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1000 to be "long long int" or "long int". This is needed for the internal
1001 formatting function string_vformat().
1003 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1004 the configuration file to be ":syslog", then the script "guesses" where
1005 the logs files are, rather than using the compiled in default. In our
1006 case the guess is not the same as the compiled default, so the script
1007 suddenly stopped working when I started to use syslog. The patch checks
1008 to see if log_file_path is "". If so, it attempts to read it from exim
1009 with no configuration file to get the compiled in version, before it
1010 falls back to the previous guessing code."
1012 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1013 implementing BATV in an Exim configuration. See NewStuff for the gory
1016 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1017 Makefile that are specific to HP-UX.
1019 PH/18 If the "use_postmaster" option was set for a recipient callout together
1020 with the "random" option, the postmaster address was used as the MAIL
1021 FROM address for the random test, but not for the subsequent recipient
1022 test. It is now used for both.
1024 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1025 patch removes a few documentation additions to RFC 3028, because the
1026 latest draft now contains them. It adds the new en;ascii-case comparator
1027 and a new error check for 8bit text in MIME parts. Comparator and
1028 require names are now matched exactly. I enabled the subaddress
1029 extension, but it is not well tested yet (read: it works for me)."
1031 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1032 rework some of the code of TK/09 above to avoid the hardwired use of
1033 "%lld" and "long long". Replaced the call to snprintf() with a call to
1036 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1037 records point to non-existent hosts", "retry timeout exceeded", and
1038 "retry time not reached for any host after a long failure period".
1040 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1041 experimental DomainKeys support:
1043 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1044 (2) On an error such as an illegally used "control", the wrong name for
1045 the control was given.
1047 These problems did NOT occur unless DomainKeys support was compiled.
1049 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1051 PH/24 Added ${if match_ip condition.
1053 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1054 databases so that it will be absolutely obvious if a crash occurs in the
1055 DB library. This is a regular occurrence (often caused by mis-matched
1058 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1059 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1060 picked up on a user's system that detects such things. There doesn't seem
1061 to be a gcc warning option for this - only an attribute that has to be
1062 put on the function's prototype. It seems that in Fedora Core 4 they have
1063 set this on a number of new functions. No doubt there will be more in due
1066 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1067 specify a transport. However, if an address that was verified by such a
1068 router was the subject of a callout, Exim crashed because it tried to
1069 read the rcpt_include_affixes from the non-existent transport. Now it
1070 just assumes that the setting of that option is false. This bug was
1071 introduced by 4.51/PH/31.
1073 PH/28 Changed -d+all to exclude +memory, because that information is very
1074 rarely of interest, but it makes the output a lot bigger. People tend to
1075 do -d+all out of habit.
1077 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1078 code in os-type was giving problems when libc.so lives in lib64, like on
1081 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1082 aren't the modern standard, and it seems that some systems' include files
1083 don't always have them. Exim was already checking for some of the newer
1084 ones like T_AAAA, and defining it itself. I've added checks for all the
1085 record types that Exim uses.
1087 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1088 not automatically generating a new one, as it is supposed to. This
1089 prevented TLS from working. If the file did exist, but contained invalid
1090 data, a new version was generated, as expected. It was only the case of a
1091 non-existent file that was broken.
1093 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1094 with a change in libdomainkeys > 0.64.
1096 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1097 from DNS. If the selector record carries the flag, it now has
1098 precedence over the domain-wide flag.
1100 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1102 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1103 the use of a port name that isn't defined in /etc/services) occurred, the
1104 message was deferred as in a normal delivery, and thus remained on the
1105 spool, instead of being failed because of the mua_wrapper setting. This
1106 is now fixed, and I tidied up some of the mua_wrapper messages at the
1109 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1110 the messages in a hash of arrays rather than using individual hashes.
1111 This is a bit cleaner and results in dramatic memory savings, albeit
1112 at a slight CPU cost.
1114 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1115 as requested by Marc Sherman.
1117 SC/10 Eximstats: added histograms for user specified patterns as requested
1120 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1122 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1123 fopen() in the content-scanning modules that did not already have it.
1129 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1130 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1132 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1134 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1136 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1137 to test Sieve filters that use "vacation".
1139 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1140 that changes the way the GnuTLS parameters are stored in the cache file.
1141 The new format can be generated externally. For backward compatibility,
1142 if the data in the cache doesn't make sense, Exim assumes it has read an
1143 old-format file, and it generates new data and writes a new file. This
1144 means that you can't go back to an older release without removing the
1147 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1148 work if there are any delivery delays because "one_time" forces the
1149 parent to be marked "delivered", so its unseen clone is never tried
1150 again. For this reason, Exim now forbids the simultaneous setting of
1153 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1154 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1155 entry if you want to know the details. The fix, however, neglected to
1156 consider the case when local delivery batching is involved. The test for
1157 "previously delivered" was not happening when checking to see if an
1158 address could be batched with a previous (undelivered) one; under
1159 certain circumstances this could lead to multiple deliveries to the same
1162 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1163 in its include files, and this causes problems building Exim.
1165 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1166 header_syntax) but Exim was just ignoring anything given after a slash.
1167 In particular, this caused confusion with an attempt to use "verify =
1168 reverse_host_lookup/defer_ok". An error is now given when options are
1169 supplied for verify items that do not have them. (Maybe reverse_host_
1170 lookup should have a defer_ok option, but that's a different point.)
1172 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1173 defined by RFC 821) to 2048, because there were problems with some AUTH
1174 commands, and RFC 1869 says the size should be increased for extended
1175 SMTP commands that take arguments.
1177 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1180 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1181 "unknown" error; now it says that the functionality isn't in the binary.
1183 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1184 an address' error message when a string expansion fails (syntax or
1185 whatever). Otherwise the password may appear in the log. Following change
1186 PH/42 below, there is no longer a chance of it appearing in a bounce
1189 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1191 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1192 of its addresses was ever tried. (Bugzilla bug #2.)
1194 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1195 the result incorrectly in the debug output. (It correctly added a newline
1196 to what was transported.)
1198 TF/01 Added $received_time.
1200 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1201 commented out examples of how to interface to a virus scanner and to
1202 SpamAssassin. Also added commented examples of av_scanner and
1203 spamd_address settings.
1205 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1206 and controls are allowed in which ACLs. There were a couple of minor
1207 errors. Some of the entries in the conditions table (which is a table of
1208 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1209 negation of where the condition IS allowed.
1211 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1213 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1214 header file does not have a version number, so I've had to invent a new
1215 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1216 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1217 radiusclient), but it was contributed by a Radius user.
1219 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1220 files or directories, for OpenSSL.
1222 PH/19 When an Exim process that is running as root has to create an Exim log
1223 file, it does so in a subprocess that runs as exim:exim so as to get the
1224 ownership right at creation (otherwise, other Exim processes might see
1225 the file with the wrong ownership). There was no test for failure of this
1226 fork() call, which would lead to the process getting stuck as it waited
1227 for a non-existent subprocess. Forks do occasionally fail when resources
1228 run out. I reviewed all the other calls to fork(); they all seem to check
1231 PH/20 When checking for unexpected SMTP input at connect time (before writing
1232 the banner), Exim was not dealing correctly with a non-positive return
1233 from the read() function. If the client had disconnected by this time,
1234 the result was a log entry for a synchronization error with an empty
1235 string after "input=" when read() returned zero. If read() returned -1
1236 (an event I could not check), uninitialized data bytes were printed.
1237 There were reports of junk text (parts of files, etc) appearing after
1240 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1242 PH/22 Added support for macro redefinition, and (re)definition in between
1243 driver and ACL definitions.
1245 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1246 forgetting to use the resulting value; it was using the unexpanded value.
1248 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1249 hadn't been configured. The fix is from Juergen Kreileder, who
1250 understands it better than I do:
1252 "Here's what I see happening with three configured cyrus_sasl
1253 authenticators configured (plain, login, cram-md5):
1255 On startup auth_cyrus_sasl_init() gets called for each of these.
1256 This means three calls to sasl_listmech() without a specified mech_list.
1257 => SASL tests which mechs of all available mechs actually work
1258 => three warnings about OTP not working
1259 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1261 With the patch, sasl_listmech() also gets called three times. But now
1262 SASL's mech_list option is set to the server_mech specified in the the
1263 authenticator. Or in other words, the answer from sasl_listmech()
1264 gets limited to just the mech you're testing for (which is different
1266 => the return list contains just 'plain' or 'login', 'cram-md5' or
1267 nothing depending on the value of ob->server_mech.
1269 I've just tested the patch: Authentication still works fine,
1270 unavailable mechs specified in the exim configuration are still
1271 caught, and the auth.log warnings about OTP are gone."
1273 PH/25 When debugging is enabled, the contents of the command line are added
1274 to the debugging output, even when log_selector=+arguments is not
1277 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1278 answer is "GNU", and only if the return is "GNU/something" is the answer
1281 PH/27 $acl_verify_message is now set immediately after the failure of a
1282 verification in an ACL, and so is available in subsequent modifiers. In
1283 particular, the message can be preserved by coding like this:
1285 warn !verify = sender
1286 set acl_m0 = $acl_verify_message
1288 Previously, $acl_verify_message was set only while expanding "message"
1289 and "log_message" when a very denied access.
1291 PH/28 Modified OS/os.c-Linux with
1293 -#ifndef OS_LOAD_AVERAGE
1294 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1296 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1297 nomenclature these days.)
1299 PH/29 Installed patch from the Sieve maintainer that adds the options
1300 sieve_useraddress and sieve_subaddress to the redirect router.
1302 PH/30 In these circumstances:
1303 . Two addresses routed to the same list of hosts;
1304 . First host does not offer TLS;
1305 . First host accepts first address;
1306 . First host gives temporary error to second address;
1307 . Second host offers TLS and a TLS session is established;
1308 . Second host accepts second address.
1309 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1310 and peerdn, if requested) that were in fact used only for the second
1313 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1314 attention to any local part prefix or suffix that was matched by the
1315 router that accepted the address. It now behaves in the same way as it
1316 does for delivery: the affixes are removed from the local part unless
1317 rcpt_include_affixes is set on the transport.
1319 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1320 timeout during the DATA phase of an incoming message.
1322 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1323 applied a patch sanctioned by the Sieve maintainer.
1325 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1326 the uid or gid is negative. A case of a negative gid caused this to be
1327 noticed. The fix allows for either to be negative.
1329 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1330 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1333 PH/36 The change PH/12 above was broken. Fixed it.
1335 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1336 the grounds that routing the same address twice would always produce the
1337 same answer. This might have been true once, but it is certainly no
1338 longer true now. Routing a child address may depend on the previous
1339 routing that produced that child. Some complicated redirection strategies
1340 went wrong when messages had multiple recipients, and made Exim's
1341 behaviour dependent on the order in which the addresses were given.
1343 I have moved the duplicate checking until after the routing is complete.
1344 Exim scans the addresses that are assigned to local and remote
1345 transports, and removes any duplicates. This means that more work will be
1346 done, as duplicates will always all be routed, but duplicates are
1347 presumably rare, so I don't expect this is of any significance.
1349 For deliveries to pipes, files, and autoreplies, the duplicate checking
1350 still happens during the routing process, since they are not going to be
1353 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1354 It corrects a timeout issue with spamd. This is Ian's comment: "The
1355 background is that sometimes spamd either never reads data from a
1356 connection it has accepted, or it never writes response data. The exiscan
1357 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1358 blindly assumes that writes won't block so it may never time out."
1360 PH/39 Allow G after quota size as well as K and M.
1362 PH/40 The value set for $authenticated_id in an authenticator may not contain
1363 binary zeroes or newlines because the value is written to log lines and
1364 to spool files. There was no check on this. Now the value is run through
1365 the string_printing() function so that such characters are converted to
1366 printable escape sequences.
1368 PH/41 $message_linecount is a new variable that contains the total number of
1369 lines in the message. Compare $body_linecount, which is the count for the
1372 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1373 bounce and delay warning messages, except in certain special cases, which
1376 (a) An SMTP error message from a remote host;
1377 (b) A message specified in a :fail: redirection;
1378 (c) A message specified in a "fail" command in a system filter;
1379 (d) A message specified in a FAIL return from the queryprogram router;
1380 (e) A message specified by the cannot_route_message router option.
1382 In these cases only, Exim does include the error details in bounce and
1383 warning messages. There are also a few cases where bland messages such
1384 as "unrouteable address" or "local delivery error" are given.
1386 PH/43 $value is now also set for the "else" part of a ${run expansion.
1388 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1389 being worked on, but at least Exim now implements the latest version to
1392 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1393 process to complete was treated as a delivery failure, a timeout while
1394 writing the message to the pipe was logged, but erroneously treated as a
1395 successful delivery. Such timeouts include transport filter timeouts. For
1396 consistency with the overall process timeout, these timeouts are now
1397 treated as errors, giving rise to delivery failures by default. However,
1398 there is now a new Boolean option for the pipe transport called
1399 timeout_defer, which, if set TRUE, converts the failures into defers for
1400 both kinds of timeout. A transport filter timeout is now identified in
1403 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1404 systems where "make" and "gmake" are different, calling "gmake" at top
1405 level broke things. I've arranged for the value of $(MAKE) to be passed
1406 from the Makefile to this script so that it can call the same version of
1410 A note about Exim versions 4.44 and 4.50
1411 ----------------------------------------
1413 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1414 changes of various kinds. As a consequence, a big documentation update was
1415 needed. This delayed the release for rather longer than seemed good, especially
1416 in the light of a couple of (minor) security issues. Therefore, the changes
1417 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1418 release. So 4.44 and 4.50 are in effect two different branches that both start
1421 I have left the 4.50 change log unchanged; it contains all the changes since
1422 4.43. The change log for 4.44 is below; many of its items are identical to
1423 those for 4.50. This seems to be the most sensible way to preserve the
1424 historical information.
1430 1. Minor wording change to the doc/README.SIEVE file.
1432 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1433 computation of the current number of files was incorrect.
1435 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1436 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1439 4. Give more explanation in the error message when the command for a transport
1440 filter fails to execute.
1442 5. There are several places where Exim runs a non-Exim command in a
1443 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1444 was being done only for the command run by the queryprogram router. It is
1445 now done for all such subprocesses. The other cases are: ${run, transport
1446 filters, and the commands run by the lmtp and pipe transports.
1448 6. Added CONFIGURE_GROUP build-time option.
1450 7. Some older OS have a limit of 256 on the maximum number of file
1451 descriptors. Exim was using setrlimit() to set 1000 as a large value
1452 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1453 systems. I've change it so that if it can't get 1000, it tries for 256.
1455 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1456 was an oversight, and furthermore, ever since the addition of extra
1457 controls (e.g. 4.43/32), the checks on when to allow different forms of
1458 "control" were broken. There should now be diagnostics for all cases when a
1459 control that does not make sense is encountered.
1461 9. Added the /retain_sender option to "control=submission".
1463 10. $recipients is now available in the predata ACL (oversight).
1465 11. Tidy the search cache before the fork to do a delivery from a message
1466 received from the command line. Otherwise the child will trigger a lookup
1467 failure and thereby defer the delivery if it tries to use (for example) a
1468 cached ldap connection that the parent has called unbind on.
1470 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1471 of $address_data from the recipient verification was clobbered by the
1472 sender verification.
1474 13. The value of address_data from a sender verification is now available in
1475 $sender_address_data in subsequent conditions in the ACL statement.
1477 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1479 15. Added a new option "connect=<time>" to callout options, to set a different
1482 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1483 was its contents. (It was OK if the option was not defined at all.)
1485 17. A "Completed" log line is now written for messages that are removed from
1486 the spool by the -Mrm option.
1488 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1489 information about exactly what failed.
1491 19. Added -dd to debug only the daemon process.
1493 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1494 handles timeouts, both on the server side and network timeouts. Renamed the
1495 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1497 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1498 It is now set to "smtps".
1500 22. $host_address is now set to the target address during the checking of
1501 ignore_target_hosts.
1503 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1504 being passed; this would have caused $sender_host_name to have been used if
1505 matching the list had actually called for a host name (not very likely,
1506 since this list is usually IP addresses). A host name is now passed as
1509 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1510 code that specifies a non-restarting handler (typically sigaction() in
1511 modern systems) in an attempt to fix a rare and obscure crash bug.
1513 25. Narrowed the window for a race in the daemon that could cause it to ignore
1514 SIGCHLD signals. This is not a major problem, because they are used only to
1515 wake it up if nothing else does.
1517 26. A malformed maildirsize file could cause Exim to calculate negative values
1518 for the mailbox size or file count. Odd effects could occur as a result.
1519 The maildirsize information is now recalculated if the size or filecount
1522 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1523 support for a long time. Removed HAVE_SYS_VFS_H.
1525 28. Installed the latest version of exipick from John Jetmore.
1527 29. In an address list, if the pattern was not a regular expression, an empty
1528 subject address (from a bounce message) matched only if the pattern was an
1529 empty string. Non-empty patterns were not even tested. This was the wrong
1530 because it is perfectly reasonable to use an empty address as part of a
1531 database query. An empty address is now tested by patterns that are
1532 lookups. However, all the other forms of pattern expect the subject to
1533 contain a local part and a domain, and therefore, for them, an empty
1534 address still always fails if the pattern is not itself empty.
1536 30. Exim went into a mad DNS loop when attempting to do a callout where the
1537 host was specified on an smtp transport, and looking it up yielded more
1538 than one IP address.
1540 31. Re-factored the code for checking spool and log partition space into a
1541 function that finds that data and another that does the check. The former
1542 is then used to implement four new variables: $spool_space, $log_space,
1543 $spool_inodes, and $log_inodes.
1545 32. The RFC2047 encoding function was originally intended for short strings
1546 such as real names; it was not keeping to the 75-character limit for
1547 encoded words that the RFC imposes. It now respects the limit, and
1548 generates multiple encoded words if necessary. To be on the safe side, I
1549 have increased the buffer size for the ${rfc2047: expansion operator from
1552 33. It is now permitted to omit both strings after an "if" condition; if the
1553 condition is true, the result is "true". As before, when the second string
1554 is omitted, a false condition yields an empty string. This makes it less
1555 cumbersome to write custom ACL and router conditions.
1557 34. Failure to deliver a bounce message always caused it to be frozen, even if
1558 there was an errors_to setting on the router. The errors_to setting is now
1561 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1562 canonical form (fully expanded) before being placed in
1563 $sender_host_address.
1565 36. The table in the code that translates DNS record types into text (T_A to
1566 "A" for instance) was missing entries for NS and CNAME. It is just possible
1567 that this could have caused confusion if both these types were looked up
1568 for the same domain, because the text type is used as part of Exim's
1569 per-process caching. But the chance of anyone hitting this buglet seems
1572 37. The dnsdb lookup has been extended in a number of ways.
1574 (1) There is a new type, "zns", which walks up the domain tree until it
1575 finds some nameserver records. It should be used with care.
1577 (2) There is a new type, "mxh", which is like "mx" except that it returns
1578 just the host names, not the priorities.
1580 (3) It is now possible to give a list of domains (or IP addresses) to be
1581 looked up. The behaviour when one of the lookups defers can be
1582 controlled by a keyword.
1584 (4) It is now possible to specify the separator character for use when
1585 multiple records are returned.
1587 38. The dnslists ACL condition has been extended: it is now possible to supply
1588 a list of IP addresses and/or domains to be looked up in a particular DNS
1591 39. Added log_selector=+queue_time_overall.
1593 40. When running the queue in the test harness, wait just a tad after forking a
1594 delivery process, to get repeatability of debugging output.
1596 41. Include certificate and key file names in error message when GnuTLS fails
1597 to set them up, because the GnuTLS error message doesn't include the name
1598 of the failing file when there is a problem reading it.
1600 42. Allow both -bf and -bF in the same test run.
1602 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1604 44. The "Exiscan patch" is now merged into the mainline Exim source.
1606 45. Sometimes the final signoff response after QUIT could fail to get
1607 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1608 < 0 before doing a fflush(). This bug looks as though it goes back to the
1609 introduction of TLS in release 3.20, but "sometimes" must have been rare
1610 because the tests only now provoked it.
1612 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1613 (this can affect the format of dates).
1615 47. exim_tidydb, when checking for the continued existence of a message for
1616 which it has found a message-specific retry record, was not finding
1617 messages that were in split spool directories. Consequently, it was
1618 deleting retry records that should have stayed in existence.
1620 48. Steve fixed some bugs in eximstats.
1622 49. The SPA authentication driver was not abandoning authentication and moving
1623 on to the next authenticator when an expansion was forced to fail,
1624 contradicting the general specification for all authenticators. Instead it
1625 was generating a temporary error. It now behaves as specified.
1627 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1628 (the order specifies the preference for clients). The order is now AES256,
1629 AES128, 3DES, ARCFOUR128.
1631 51. Small patch to Sieve code - explicitly set From: when generating an
1634 52. Exim crashed if a remote delivery caused a very long error message to be
1635 recorded - for instance if somebody sent an entire SpamAssassin report back
1636 as a large number of 550 error lines. This bug was coincidentally fixed by
1637 increasing the size of one of Exim's internal buffers (big_buffer) that
1638 happened as part of the Exiscan merge. However, to be on the safe side, I
1639 have made the code more robust (and fixed the comments that describe what
1642 53. Now that there can be additional text after "Completed" in log lines (if
1643 the queue_time_overall log selector is set), a one-byte patch to exigrep
1644 was needed to allow it to recognize "Completed" as not the last thing in
1647 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1648 patch that reportedly fixes this has been added. I am not expert enough to
1649 create a test for it. This is what the patch creator wrote:
1651 "I found a little strange behaviour of ldap code when working with
1652 Windows 2003 AD Domain, where users was placed in more than one
1653 Organization Units. When I tried to give exim partial DN, the exit code
1654 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1655 But simultaneously result of request was absolutely normal ldap result,
1656 so I produce this patch..."
1658 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1659 REFERENCE, so I have modified the code to exclude the patch when that macro
1662 55. Some experimental protocols are using DNS PTR records for new purposes. The
1663 keys for these records are domain names, not reversed IP addresses. The
1664 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1665 leaves it alone. Component reversal etc. now happens only for IP addresses.
1668 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1670 57. Double the size of the debug message buffer (to 2048) so that more of very
1671 long debug lines gets shown.
1673 58. The exicyclog utility now does better if the number of log files to keep
1674 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1676 59. Two changes related to the smtp_active_hostname option:
1678 (1) $smtp_active_hostname is now available as a variable.
1679 (2) The default for smtp_banner uses $smtp_active_hostname instead
1680 of $primary_hostname.
1682 60. The host_aton() function is supposed to be passed a string that is known
1683 to be a valid IP address. However, in the case of IPv6 addresses, it was
1684 not checking this. This is a hostage to fortune. Exim now panics and dies
1685 if the condition is not met. A case was found where this could be provoked
1686 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1687 components; fortuitously, this particular loophole had already been fixed
1688 by change 4.50/55 above.
1690 If there are any other similar loopholes, the new check in host_aton()
1691 itself should stop them being exploited. The report I received stated that
1692 data on the command line could provoke the exploit when Exim was running as
1693 exim, but did not say which command line option was involved. All I could
1694 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1695 running as the user.
1698 61. There was a buffer overflow vulnerability in the SPA authentication code
1699 (which came originally from the Samba project). I have added a test to the
1700 spa_base64_to_bits() function which I hope fixes it.
1703 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1704 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1706 63. The daemon start-up calls getloadavg() while still root for those OS that
1707 need the first call to be done as root, but it missed one case: when
1708 deliver_queue_load_max is set with deliver_drop_privilege. This is
1709 necessary for the benefit of the queue runner, because there is no re-exec
1710 when deliver_drop_privilege is set.
1712 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1713 This has been fixed.
1715 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1716 was in use, was not putting the data itself into the right store pool;
1717 consequently, it could be overwritten for a subsequent message in the same
1718 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1721 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1723 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1724 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1725 the function were treating the return as a boolean value, which happened to
1726 work because 0=false and not-0=true, but is not correct code.
1728 68. The host_aton() function was not handling scoped IPv6 addresses (those
1729 with, for example, "%eth0" on the end) correctly.
1731 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1732 negated items (that is, ~something) in unsigned ints. Some compilers
1733 apparently mutter when there is no cast.
1735 70. If an address verification called from an ACL failed, and did not produce a
1736 user-specific message (i.e. there was only a "system" message), nothing was
1737 put in $acl_verify_message. In this situation, it now puts the system
1740 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1741 session; change 4.31/43 added the unwanted input to the log line - except
1742 that it did not do this in the start of session case. It now does.
1744 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1745 This is wrong and can cause the other end to generate a synchronization
1746 error if it is another Exim or anything else that does the synchronization
1747 check. A QUIT command is no longer sent after a timeout.
1749 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1750 during host lookups.
1752 74. The defer_ok option of callout verification was not working if it was used
1753 when verifying addresses in header lines, that is, for this case:
1755 verify = header_sender/callout=defer_ok
1757 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1758 those file descriptors could be used for SMTP connections. If anything
1759 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1760 could be sent to the SMTP client, causing chaos. The daemon now opens
1761 stdin, stdout, and stderr to /dev/null when it puts itself into the
1764 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1765 log by default. The user can override this with suitable Perl magic.
1767 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1768 the log message when discard triggers, was not working for the DATA ACL or
1769 for the non-SMTP ACL.
1771 78. Error message wording change in sieve.c.
1773 79. If smtp_accept_max_per_host was set, the number of connections could be
1774 restricted to fewer than expected, because the daemon was trying to set up
1775 a new connection before checking whether the processes handling previous
1776 connections had finished. The check for completed processes is now done
1777 earlier. On busy systems, this bug wouldn't be noticed because something
1778 else would have woken the daemon, and it would have reaped the completed
1781 80. If a message was submitted locally by a user whose login name contained one
1782 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1783 It caused a spool format error. I have fixed the spool reading code. A
1784 related problem was that the "from" clause in the Received: line became
1785 illegal because of the space(s). It is now covered by ${quote_local_part.
1787 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1790 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1792 83. Patch from Sieve maintainer to fix unterminated string problem in
1793 "vacation" handling.
1795 84. Some minor changes to the Linux configuration files to help with other
1796 OS variants using glibc.
1798 85. One more patch for Sieve to update vacation handling to latest spec.
1801 ----------------------------------------------------
1802 See the note above about the 4.44 and 4.50 releases.
1803 ----------------------------------------------------
1809 1. Change 4.43/35 introduced a bug that caused file counts to be
1810 incorrectly computed when quota_filecount was set in an appendfile
1813 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1814 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1817 3. Give more explanation in the error message when the command for a transport
1818 filter fails to execute.
1820 4. There are several places where Exim runs a non-Exim command in a
1821 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1822 was being done only for the command run by the queryprogram router. It is
1823 now done for all such subprocesses. The other cases are: ${run, transport
1824 filters, and the commands run by the lmtp and pipe transports.
1826 5. Some older OS have a limit of 256 on the maximum number of file
1827 descriptors. Exim was using setrlimit() to set 1000 as a large value
1828 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1829 systems. I've change it so that if it can't get 1000, it tries for 256.
1831 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1832 was an oversight, and furthermore, ever since the addition of extra
1833 controls (e.g. 4.43/32), the checks on when to allow different forms of
1834 "control" were broken. There should now be diagnostics for all cases when a
1835 control that does not make sense is encountered.
1837 7. $recipients is now available in the predata ACL (oversight).
1839 8. Tidy the search cache before the fork to do a delivery from a message
1840 received from the command line. Otherwise the child will trigger a lookup
1841 failure and thereby defer the delivery if it tries to use (for example) a
1842 cached ldap connection that the parent has called unbind on.
1844 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1845 of $address_data from the recipient verification was clobbered by the
1846 sender verification.
1848 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1849 was its contents. (It was OK if the option was not defined at all.)
1851 11. A "Completed" log line is now written for messages that are removed from
1852 the spool by the -Mrm option.
1854 12. $host_address is now set to the target address during the checking of
1855 ignore_target_hosts.
1857 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1858 being passed; this would have caused $sender_host_name to have been used if
1859 matching the list had actually called for a host name (not very likely,
1860 since this list is usually IP addresses). A host name is now passed as
1863 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1864 code that specifies a non-restarting handler (typically sigaction() in
1865 modern systems) in an attempt to fix a rare and obscure crash bug.
1867 15. Narrowed the window for a race in the daemon that could cause it to ignore
1868 SIGCHLD signals. This is not a major problem, because they are used only to
1869 wake it up if nothing else does.
1871 16. A malformed maildirsize file could cause Exim to calculate negative values
1872 for the mailbox size or file count. Odd effects could occur as a result.
1873 The maildirsize information is now recalculated if the size or filecount
1876 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1877 support for a long time. Removed HAVE_SYS_VFS_H.
1879 18. Updated exipick to current release from John Jetmore.
1881 19. Allow an empty sender to be matched against a lookup in an address list.
1882 Previously the only cases considered were a regular expression, or an
1885 20. Exim went into a mad DNS lookup loop when doing a callout where the
1886 host was specified on the transport, if the DNS lookup yielded more than
1889 21. The RFC2047 encoding function was originally intended for short strings
1890 such as real names; it was not keeping to the 75-character limit for
1891 encoded words that the RFC imposes. It now respects the limit, and
1892 generates multiple encoded words if necessary. To be on the safe side, I
1893 have increased the buffer size for the ${rfc2047: expansion operator from
1896 22. Failure to deliver a bounce message always caused it to be frozen, even if
1897 there was an errors_to setting on the router. The errors_to setting is now
1900 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1901 canonical form (fully expanded) before being placed in
1902 $sender_host_address.
1904 24. Updated eximstats to version 1.33
1906 25. Include certificate and key file names in error message when GnuTLS fails
1907 to set them up, because the GnuTLS error message doesn't include the name
1908 of the failing file when there is a problem reading it.
1910 26. Expand error message when OpenSSL has problems setting up cert/key files.
1913 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1914 (this can affect the format of dates).
1916 28. exim_tidydb, when checking for the continued existence of a message for
1917 which it has found a message-specific retry record, was not finding
1918 messages that were in split spool directories. Consequently, it was
1919 deleting retry records that should have stayed in existence.
1921 29. eximstats updated to version 1.35
1922 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1923 1.35 - bugfix such that pie charts by volume are generated correctly
1925 30. The SPA authentication driver was not abandoning authentication and moving
1926 on to the next authenticator when an expansion was forced to fail,
1927 contradicting the general specification for all authenticators. Instead it
1928 was generating a temporary error. It now behaves as specified.
1930 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1931 (the order specifies the preference for clients). The order is now AES256,
1932 AES128, 3DES, ARCFOUR128.
1934 31. Small patch to Sieve code - explicitly set From: when generating an
1937 32. Exim crashed if a remote delivery caused a very long error message to be
1938 recorded - for instance if somebody sent an entire SpamAssassin report back
1939 as a large number of 550 error lines. This bug was coincidentally fixed by
1940 increasing the size of one of Exim's internal buffers (big_buffer) that
1941 happened as part of the Exiscan merge. However, to be on the safe side, I
1942 have made the code more robust (and fixed the comments that describe what
1945 33. Some experimental protocols are using DNS PTR records for new purposes. The
1946 keys for these records are domain names, not reversed IP addresses. The
1947 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1948 leaves it alone. Component reversal etc. now happens only for IP addresses.
1951 34. The host_aton() function is supposed to be passed a string that is known
1952 to be a valid IP address. However, in the case of IPv6 addresses, it was
1953 not checking this. This is a hostage to fortune. Exim now panics and dies
1954 if the condition is not met. A case was found where this could be provoked
1955 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1956 components; fortuitously, this particular loophole had already been fixed
1957 by change 4.50/55 or 4.44/33 above.
1959 If there are any other similar loopholes, the new check in host_aton()
1960 itself should stop them being exploited. The report I received stated that
1961 data on the command line could provoke the exploit when Exim was running as
1962 exim, but did not say which command line option was involved. All I could
1963 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1964 running as the user.
1967 35. There was a buffer overflow vulnerability in the SPA authentication code
1968 (which came originally from the Samba project). I have added a test to the
1969 spa_base64_to_bits() function which I hope fixes it.
1972 36. The daemon start-up calls getloadavg() while still root for those OS that
1973 need the first call to be done as root, but it missed one case: when
1974 deliver_queue_load_max is set with deliver_drop_privilege. This is
1975 necessary for the benefit of the queue runner, because there is no re-exec
1976 when deliver_drop_privilege is set.
1978 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1979 was in use, was not putting the data itself into the right store pool;
1980 consequently, it could be overwritten for a subsequent message in the same
1981 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1984 38. Sometimes the final signoff response after QUIT could fail to get
1985 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1986 < 0 before doing a fflush(). This bug looks as though it goes back to the
1987 introduction of TLS in release 3.20, but "sometimes" must have been rare
1988 because the tests only now provoked it.
1994 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1995 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1996 Now it returns an integer. A number of calls were still expecting a T/F
1997 return. Fortuitously, in all cases, the tests worked in OK situations,
1998 which is the norm. However, things would have gone wrong on any write
1999 failures on the smtp file descriptor. This function is used when sending
2000 messages over SMTP and also when doing verify callouts.
2002 2. When Exim is called to do synchronous delivery of a locally submitted
2003 message (the -odf or -odi options), it no longer closes stderr before doing
2006 3. Implemented the mua_wrapper option.
2008 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2010 5. Implemented the functions header_remove(), header_testname(),
2011 header_add_at_position(), and receive_remove_recipient(), and exported them
2014 6. If an ACL "warn" statement specified the addition of headers, Exim already
2015 inserted X-ACL-Warn: at the start if there was no header name. However, it
2016 was not making this test for the second and subsequent header lines if
2017 there were newlines in the string. This meant that an invalid header could
2018 be inserted if Exim was badly configured.
2020 7. Allow an ACL "warn" statement to add header lines at the start or after all
2021 the Received: headers, as well as at the end.
2023 8. Added the rcpt_4xx retry error code.
2025 9. Added postmaster_mailfrom=xxx to callout verification option.
2027 10. Added mailfrom=xxxx to the callout verification option, for verify=
2030 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2031 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2033 12. Inserted some casts to stop certain compilers warning when using pointer
2034 differences as field lengths or precisions in printf-type calls (mostly
2035 affecting debugging statements).
2037 13. Added optional readline() support for -be (dynamically loaded).
2039 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2040 same clock tick as a message's arrival, so that its received time was the
2041 same as the "first fail" time on the retry record, and that message
2042 remained on the queue past the ultimate address timeout, every queue runner
2043 would try a delivery (because it was past the ultimate address timeout) but
2044 after another failure, the ultimate address timeout, which should have then
2045 bounced the address, did not kick in. This was a "< instead of <=" error;
2046 in most cases the first failure would have been in the next clock tick
2047 after the received time, and all would be well.
2049 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2050 being recognized when the domain list was tested by the match_domain
2051 condition in an expansion string.
2053 16. Added the ${str2b64: operator.
2055 17. Exim was always calling setrlimit() to set a large limit for the number of
2056 processes, without checking whether the existing limit was already
2057 adequate. (It did check for the limit on file descriptors.) Furthermore,
2058 errors from getrlimit() and setrlimit() were being ignored. Now they are
2059 logged to the main and panic logs, but Exim does carry on, to try to do its
2060 job under whatever limits there are.
2062 18. Imported PCRE 5.0.
2064 19. Trivial typo in log message " temporarily refused connection" (the leading
2067 20. If the log selector return_path_on_delivery was set and an address was
2068 redirected to /dev/null, the delivery process crashed because it assumed
2069 that a return path would always be set for a "successful" delivery. In this
2070 case, the whole delivery is bypassed as an optimization, and therefore no
2073 21. Internal re-arrangement: the function for sending a challenge and reading
2074 a response while authentication was assuming a zero-terminated challenge
2075 string. It's now changed to take a pointer and a length, to allow for
2076 binary data in such strings.
2078 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2080 23. Exim was not respecting finduser_retries when seeking the login of the
2081 uid under which it was called; it was always trying 10 times. (The default
2082 setting of finduser_retries is zero.) Also, it was sleeping after the final
2083 failure, which is pointless.
2085 24. Implemented tls_on_connect_ports.
2087 25. Implemented acl_smtp_predata.
2089 26. If the domain in control=submission is set empty, Exim assumes that the
2090 authenticated id is a complete email address when it generates From: or
2091 Sender: header lines.
2093 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2094 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2095 chown and chgrp in /bin and hostname in /usr/bin.
2097 28. Exim was keeping the "process log" file open after each use, just as it
2098 does for the main log. This opens the possibility of it remaining open for
2099 long periods when the USR1 signal hits a daemon. Occasional processlog
2100 errors were reported, that could have been caused by this. Anyway, it seems
2101 much more sensible not to leave this file open at all, so that is what now
2104 29. The long-running daemon process does not normally write to the log once it
2105 has entered its main loop, and it closes the log before doing so. This is
2106 so that log files can straightforwardly be renamed and moved. However,
2107 there are a couple of unusual error situations where the daemon does write
2108 log entries, and I had neglected to close the log afterwards.
2110 30. The text of an SMTP error response that was received during a remote
2111 delivery was being truncated at 512 bytes. This is too short for some of
2112 the long messages that one sometimes sees. I've increased the limit to
2115 31. It is now possible to make retry rules that apply only when a message has a
2116 specific sender, in particular, an empty sender.
2118 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2119 it possible to be selective about when SMTP synchronization is enforced.
2121 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2123 32. Implemented hosts_connection_nolog.
2125 33. Added an ACL for QUIT.
2127 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2130 35. Added mailbox_size and mailbox_filecount to appendfile.
2132 36. Added control = no_multiline_responses to ACLs.
2134 37. There was a bug in the logic of the code that waits for the clock to tick
2135 in the case where the clock went backwards by a substantial amount such
2136 that the microsecond fraction of "now" was more than the microsecond
2137 fraction of "then" (but the whole seconds number was less).
2139 38. Added support for the libradius Radius client library this is found on
2140 FreeBSD (previously only the radiusclient library was supported).
2146 1. When certain lookups returned multiple values in the form name=value, the
2147 quoting of the values was not always being done properly. Specifically:
2148 (a) If the value started with a double quote, but contained no whitespace,
2150 (b) If the value contained whitespace other than a space character (i.e.
2151 tabs or newlines or carriage returns) it was not quoted.
2152 This fix has been applied to the mysql and pgsql lookups by writing a
2153 separate quoting function and calling it from the lookup code. The fix
2154 should probably also be applied to nisplus, ibase and oracle lookups, but
2155 since I cannot test any of those, I have not disturbed their existing code.
2157 2. A hit in the callout cache for a specific address caused a log line with no
2158 reason for rejecting RCPT. Now it says "Previous (cached) callout
2159 verification failure".
2161 3. There was an off-by-one bug in the queryprogram router. An over-long
2162 return line was truncated at 256 instead of 255 characters, thereby
2163 overflowing its buffer with the terminating zero. As well as fixing this, I
2164 have increased the buffer size to 1024 (and made a note to document this).
2166 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2167 when Exim is waiting for an SMTP response from a remote server, Exim
2168 restarts its select() call on the socket, thereby resetting its timeout.
2169 This is not a problem when such interrupts are rare. Somebody set up a cron
2170 job to run exiwhat every 2 minutes, which is less than the normal select()
2171 timeout (5 or 10 minutes). This meant that the select() timeout never
2172 kicked in because it was always reset. I have fixed this by comparing the
2173 time when an interrupt arrives with the time at the start of the first call
2174 to select(). If more time than the timeout has elapsed, the interrupt is
2175 treated as a timeout.
2177 5. Some internal re-factoring in preparation for the addition of Sieve
2178 extensions (by MH). In particular, the "personal" test is moved to a
2179 separate function, and given an option for scanning Cc: and Bcc: (which is
2180 not set for Exim filters).
2182 6. When Exim created an email address using the login of the caller as the
2183 local part (e.g. when creating a From: or Sender: header line), it was not
2184 quoting the local part when it contained special characters such as @.
2186 7. Installed new OpenBSD configuration files.
2188 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2189 try to make them clearer.
2191 9. Callout options, other than the timeout value, were being ignored when
2192 verifying sender addresses in header lines. For example, when using
2194 verify = header_sender/callout=no_cache
2196 the cache was (incorrectly) being used.
2198 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2199 only the Cygwin environment.
2201 11. When return_path_on_delivery was set as a log selector, if different remote
2202 addresses in the same message used different return paths and parallel
2203 remote delivery occurred, the wrong values would sometimes be logged.
2204 (Whenever a remote delivery process finished, the return path value from
2205 the most recently started remote delivery process was logged.)
2207 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2208 lines when AUTH and/or TLS are in use. This is the "received protocol"
2209 field. Exim used to use "asmtp" for authenticated SMTP, without any
2210 indication (in the protocol name) for TLS use. Now it follows the RFC and
2211 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2212 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2213 names appear in log lines as well as in Received: header lines.
2215 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2216 extensions, and comparison tests, and to fix some bugs.
2218 14. Changes to the "personal" filter test:
2220 (1) The test was buggy in that it was just doing the equivalent of
2221 "contains" tests on header lines. For example, if a user's address was
2222 anne@some.where, the "personal" test would incorrectly be true for
2224 To: susanne@some.where
2226 This test is now done by extracting each address from the header in turn,
2227 and checking the entire address. Other tests that are part of "personal"
2228 are now done using regular expressions (for example, to check local parts
2229 of addresses in From: header lines).
2231 (2) The list of non-personal local parts in From: addresses has been
2232 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2233 taken from the Sieve specification recommendations.
2235 (3) If the message contains any header line starting with "List-" it is
2236 treated as non-personal.
2238 (4) The test for "circular" in the Subject: header line has been removed
2239 because it now seems ill-conceived.
2241 15. Minor typos in src/EDITME comments corrected.
2243 16. Installed latest exipick from John Jetmore.
2245 17. If headers_add on a router specified a text string that was too long for
2246 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2247 of string_sprintf() is now avoided.
2249 18. $message_body_size was not set (it was always zero) when running the DATA
2250 ACL and the local_scan() function.
2252 19. For the "mail" command in an Exim filter, no default was being set for
2253 the once_repeat time, causing a random time value to be used if "once" was
2254 specified. (If the value happened to be <= 0, no repeat happened.) The
2255 default is now 0s, meaning "never repeat". The "vacation" command was OK
2256 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2257 (I found it when inspecting the code).
2259 20. There is now an overall timeout for performing a callout verification. It
2260 defaults to 4 times the callout timeout, which applies to individual SMTP
2261 commands during the callout. The overall timeout applies when there is more
2262 than one host that can be tried. The timeout is checked before trying the
2263 next host. This prevents very long delays if there are a large number of
2264 hosts and all are timing out (e.g. when the network connections are timing
2265 out). The value of the overall timeout can be changed by specifying an
2266 additional sub-option for "callout", called "maxwait". For example:
2268 verify = sender/callout=5s,maxwait=20s
2270 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2271 to the end before writing, but this should make it even safer).
2273 22. Exim was forgetting that it had advertised PIPELINING for the second and
2274 subsequent messages on an SMTP connection. It was also not resetting its
2275 memory on STARTTLS and an internal HELO.
2277 23. When Exim logs an SMTP synchronization error within a session, it now
2278 records whether PIPELINING has been advertised or not.
2280 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2281 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2282 rather than long int.
2284 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2286 26. Added the never_mail option to autoreply.
2292 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2293 crash if the getsockname() call failed; this can happen if a connection is
2294 closed very soon after it is established. The problem was simply in the
2295 order in which certain operations were done, causing Exim to try to write
2296 to the SMTP stream before it had set up the file descriptor. The bug has
2297 been fixed by making things happen in the correct order.
2303 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2304 before the connection was closed, thus losing the rejection response.
2306 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2307 some early Solaris releases, but causes trouble in current releases where
2308 socklen_t is defined.
2310 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2313 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2316 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2317 that releasing the top of it at the end releases what was used for sub-
2318 expansions (unless the block got too big). However, discard this block if
2319 the first thing is a variable or header, so that we can use its block when
2320 it is dynamic (useful for very large $message_headers, for example).
2322 6. Lookups now cache *every* query, not just the most recent. A new, separate
2323 store pool is used for this. It can be recovered when all lookup caches are
2324 flushed. Lookups now release memory at the end of their result strings.
2325 This has involved some general refactoring of the lookup sources.
2327 7. Some code has been added to the store_xxx() functions to reduce the amount
2328 of flapping under certain conditions.
2330 8. log_incoming_interface used to affect only the <= reception log lines. Now
2331 it causes the local interface and port to be added to several more SMTP log
2332 lines, for example "SMTP connection from", and rejection lines.
2334 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2336 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2338 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2339 could be overwritten at the end of the current message (or the start of a
2340 new message if it was set in a HELO ACL). The value is now preserved for
2341 the duration of the SMTP connection.
2343 12. If a transport had a headers_rewrite setting, and a matching header line
2344 contained an unqualified address, that address was qualified, even if it
2345 did not match any rewriting rules. The underlying bug was that the values
2346 of the flags that permit the existence of unqualified sender and recipient
2347 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2348 non-local messages, and by -bnq for local messages) were not being
2349 preserved with the message after it was received.
2351 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2352 "next input=" as part of the text comprising the host identity instead of
2353 the correct text. The code was using the same buffer for two different
2354 strings. However, depending on which order the printing function evaluated
2355 its arguments, the bug did not always show up. Under Linux, for example, my
2356 test suite worked just fine.
2358 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2359 doesn't work with some older versions of Perl. It has been changed to "my",
2360 which in any case is probably the better facility to use.
2362 15. A really picky compiler found some instances of statements for creating
2363 error messages that either had too many or two few arguments for the format
2366 16. The size of the buffer for calls to the DNS resolver has been increased
2367 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2368 for addresses that have a lot of PTR records. This alleviates a problem; it
2369 does not fully solve it.
2371 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2372 buffer now truncates the list and logs the incident, which is the same
2373 action as happens when Exim is looking up a host name and its aliases.
2374 Previously in this situation something unpredictable would happen;
2375 sometimes it was "internal error: store_reset failed".
2377 18. If a server dropped the connection unexpectedly when an Exim client was
2378 using GnuTLS and trying to read a response, the client delivery process
2379 crashed while trying to generate an error log message.
2381 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2382 string, for example:
2384 warn message = H1: something\nH2: something
2386 the text was added as a single header line from Exim's point of view
2387 though it ended up OK in the delivered message. However, searching for the
2388 second and subsequent header lines using $h_h2: did not work. This has been
2389 fixed. Similarly, if a system filter added multiple headers in this way,
2390 the routers could not see them.
2392 20. Expanded the error message when iplsearch is called with an invalid key to
2393 suggest using net-iplsearch in a host list.
2395 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2396 ACLs are no longer actually imposed (and a message to that effect is
2399 22. If a "gecos" field in a passwd entry contained escaped characters, in
2400 particular, if it contained a \" sequence, Exim got it wrong when building
2401 a From: or a Sender: header from that name. A second bug also caused
2402 incorrect handling when an unquoted " was present following a character
2403 that needed quoting.
2405 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2406 was not being matched caselessly.
2408 24. Arranged for all hyphens in the exim.8 source to be escaped with
2411 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2412 properly. Recipient callout cache records were still being keyed to include
2413 the sender, even when use_sender was set false. This led to far more
2414 callouts that were necessary. The sender is no longer included in the key
2415 when use_sender is false.
2417 26. Added "control = submission" modifier to ACLs.
2419 27. Added the ${base62d: operator to decode base 62 numbers.
2421 28. dnsdb lookups can now access SRV records.
2423 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2424 the configuration file.
2426 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2427 (-v) mode. This makes the output for a verbose queue run more intelligible.
2429 31. Added a use_postmaster feature to recipient callouts.
2431 32. Added the $body_zerocount variable, containing the number of binary zero
2432 bytes in the message body.
2434 33. The time of last modification of the "new" subdirectory is now used as the
2435 "mailbox time last read" when there is a quota error for a maildir
2438 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2440 35. Added +ignore_unknown as a special item in host lists.
2442 36. Code for decoding IPv6 addresses in host lists is now included, even if
2443 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2444 address was recognized as an IP address, but was then not correctly decoded
2445 into binary, causing unexpected and incorrect effects when compared with
2452 1. Very minor rewording of debugging text in manualroute to say "list of
2453 hosts" instead of "hostlist".
2455 2. If verify=header_syntax was set, and a header line with an unqualified
2456 address (no domain) and a large number of spaces between the end of the
2457 name and the colon was received, the reception process suffered a buffer
2458 overflow, and (when I tested it) crashed. This was caused by some obsolete
2459 code that should have been removed. The fix is to remove it!
2461 3. When running in the test harness, delay a bit after writing a bounce
2462 message to get a bit more predictability in the log output.
2464 4. Added a call to search_tidyup() just before forking a reception process. In
2465 theory, someone could use a lookup in the expansion of smtp_accept_max_
2466 per_host which, without the tidyup, could leave open a database connection.
2468 5. Added the variables $recipient_data and $sender_data which get set from a
2469 lookup success in an ACL "recipients" or "senders" condition, or a router
2470 "senders" option, similar to $domain_data and $local_part_data.
2472 6. Moved the writing of debug_print from before to after the "senders" test
2475 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2476 problems for message scanning, either using a data ACL, or using
2477 local_scan() because the Received: header was not generated till after they
2478 were called (in order to set the time as the time of reception completion).
2479 I have revised the way this works. The header is now generated after the
2480 body is received, but before the ACL or local_scan() are called. After they
2481 are run, the timestamp in the header is updated.
2487 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2488 before starting a queue runner without re-exec. This happened only when
2489 deliver_drop_privilege was set or when the Exim user was set to root. The
2490 effect of the bug was that timeouts during subsequent deliveries caused
2491 crashes instead of being properly handled. The handler is now left at its
2492 default (and expected) setting.
2494 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2495 message, again when deliver_drop_privilege is set or Exim is run as root.
2496 The bug described in (1) was not present in this case, but the tidying up
2497 of the other signals was missing. I have made the two cases consistent.
2499 3. The ignore_target_hosts setting on a manualroute router was being ignored
2500 for hosts that were looked up using the /MX notation.
2502 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2505 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2506 operated on the sender address. After changing the $sender_address to <>
2507 for the sender address verify, Exim was re-instated it as the original
2508 (before rewriting) address, but remembering that it had rewritten it, so it
2509 wasn't rewriting it again. This bug also had the effect of breaking the
2510 sender address verification caching when the sender address was rewritten.
2512 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2513 This has been changed so that if the ip literal address matches
2514 ignore_target_hosts, the router declines.
2516 7. Added expansion conditions match_domain, match_address, and match_local_
2517 part (NOT match_host).
2519 8. The placeholder for the Received: header didn't have a length field set.
2521 9. Added code to Exim itself and to exim_lock to test for a specific race
2522 condition that could lead to file corruption when using MBX delivery. The
2523 issue is with the lockfile that is created in /tmp. If this file is removed
2524 after a process has opened it but before that process has acquired a lock,
2525 there is the potential for a second process to recreate the file and also
2526 acquire a lock. This could lead to two Exim processes writing to the file
2527 at the same time. The added code performs the same test as UW imapd; it
2528 checks after acquiring the lock that its file descriptor still refers to
2529 the same named file.
2531 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2532 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2535 11. Added the smtp_active_hostname option. If used, this will typically be made
2536 to depend on the incoming interface address. Because $interface_address is
2537 not set up until the daemon has forked a reception process, error responses
2538 that can happen earlier (such as "too many connections") no longer contain
2541 12. If an expansion in a condition on a "warn" statement fails because a lookup
2542 defers, the "warn" statement is abandoned, and the next ACL statement is
2543 processed. Previously this caused the whole ACL to be aborted.
2545 13. Added the iplsearch lookup type.
2547 14. Added ident_timeout as a log selector.
2549 15. Added tls_certificate_verified as a log selector.
2551 16. Added a global option tls_require_ciphers (compare the smtp transport
2552 option of the same name). This controls incoming TLS connections.
2554 17. I finally figured out how to make tls_require_ciphers do a similar thing
2555 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2556 before starting the TLS session.
2558 18. Tabs are now shown as \t in -bP output.
2560 19. If the log selector return_path_on_delivery was set, Exim crashed when
2561 bouncing a message because it had too many Received: header lines.
2563 20. If two routers both had headers_remove settings, and the first one included
2564 a superfluous trailing colon, the final name in the first list and the
2565 first name in the second list were incorrectly joined into one item (with a
2566 colon in the middle).
2572 1. Added -C and -D options to the exinext utility, mainly to make it easier
2573 to include in the automated testing, but these could be helpful when
2574 multiple configurations are in use.
2576 2. The exinext utility was not formatting the output nicely when there was
2577 an alternate port involved in the retry record key, nor when there was a
2578 message id as well (for retries that were specific to a specific message
2579 and a specific host). It was also confused by IPv6 addresses, because of
2580 the additional colons they contain. I have fixed the IPv4 problem, and
2581 patched it up to do a reasonable job for IPv6.
2583 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2584 delivery, the log line now contains "pipelined" if PIPELINING was used.
2586 4. An SMTP transport process used to panic and die if the bind() call to set
2587 an explicit outgoing interface failed. This has been changed; it is now
2588 treated in the same way as a connect() failure.
2590 5. A reference to $sender_host_name in the part of a conditional expansion
2591 that was being skipped was still causing a DNS lookup. This no longer
2594 6. The def: expansion condition was not recognizing references to header lines
2595 that used bh_ and bheader_.
2597 7. Added the _cache feature to named lists.
2599 8. The code for checking quota_filecount in the appendfile transport was
2600 allowing one more file than it should have been.
2602 9. For compatibility with Sendmail, the command line option
2610 and sets the incoming protocol and host name (for trusted callers). The
2611 host name and its colon can be omitted when only the protocol is to be set.
2612 Note the Exim already has two private options, -pd and -ps, that refer to
2613 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2614 "s", but I don't think that's a major issue.
2616 10. A number of refactoring changes to the code, none of which should affect
2619 (a) The number of logging options was getting close to filling up the
2620 32-bit word that was used as a bit map. I have split them into two classes:
2621 those that are passed in the argument to log_write(), and those that are
2622 only ever tested independently outside of that function. These are now in
2623 separate 32-bit words, so there is plenty of room for expansion again.
2624 There is no change in the user interface or the logging behaviour.
2626 (b) When building, for example, log lines, the code previously used a
2627 macro that called string_cat() twice, in order to add two strings. This is
2628 not really sufficiently general. Furthermore, there was one instance where
2629 it was actually wrong because one of the argument was used twice, and in
2630 one call a function was used. (As it happened, calling the function twice
2631 did not affect the overall behaviour.) The macro has been replaced by a
2632 function that can join an arbitrary number of extra strings onto a growing
2635 (c) The code for expansion conditions now uses a table and a binary chop
2636 instead of a serial search (which was left over from when there were very
2637 few conditions). Also, it now recognizes conditions like "pam" even when
2638 the relevant support is not compiled in: a suitably worded error message is
2639 given if an attempt is made to use such a condition.
2641 11. Added ${time_interval:xxxxx}.
2643 12. A bug was causing one of the ddress fields not to be passed back correctly
2644 from remote delivery subprocesses. The field in question was not being
2645 subsequently used, so this caused to problems in practice.
2647 13. Added new log selectors queue_time and deliver_time.
2649 14. Might have fixed a bug in maildirsizefile handling that threw up
2650 "unexpected character" debug warnings, and recalculated the data
2651 unnecessarily. In any case, I expanded the warning message to give more
2654 15. Added the message "Restricted characters in address" to the statements in
2655 the default ACL that block characters like @ and % in local parts.
2657 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2658 Three changes have been made:
2660 (a) There was a serious bug; a negative response to MAIL caused the whole
2661 recipient domain to be cached as invalid, thereby blocking all messages
2662 to all local parts at the same domain, from all senders. This bug has
2663 been fixed. The domain is no longer cached after a negative response to
2664 MAIL if the sender used is not empty.
2666 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2669 (c) A new callout option, "use_sender" has been added for people who want
2670 the modified behaviour.
2676 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2679 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2680 indeed breaks things for older releases.
2682 3. Added additional logging to the case where there is a problem reading data
2683 from a filter that is running in a subprocess using a pipe, in order to
2684 try to track down a specific problem.
2686 4. Testing facility fudge: when running in the test harness and attempting
2687 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2688 getting "No route to host". Convert this to a timeout.
2690 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2693 6. Some OS don't have socklen_t but use size_t instead. This affects the
2694 fifth argument of getsockopt() amongst other things. This is now
2695 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2696 can be set for individual OS. I have set it for SunOS5, OSF1, and
2697 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2698 some earlier ones do not.
2700 7. Change 4.30/15 was not doing the test caselessly.
2702 8. The standard form for an IPv6 address literal was being rejected by address
2703 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2704 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2705 this, as well as the form without the "IPv6" on the front (but only when
2706 address literals are enabled, of course).
2708 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2710 10. Exim crashed if a message with an empty sender address specified by -f
2711 encountered a router with an errors_to setting. This could be provoked only
2712 by a command such as
2716 where an empty string was supplied; "<>" did not hit this bug.
2718 11. Installed PCRE release 4.5.
2720 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2721 remained set. It is now erased.
2723 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2724 times from message ids (which are base 36 rather than the normal 62).
2726 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2727 were being counted as actual protocol errors, and logged if the log
2728 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2729 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2730 and DATA following a set of rejected RCPTs do not count as protocol errors.
2731 In other words, Exim assumes they were pipelined, though this may not
2732 actually be the case. Of course, in all cases the client gets an
2733 appropriate error code.
2735 15. If a lookup fails in an ACL condition, a message about the failure may
2736 be available; it is used if testing the ACL cannot continue, because most
2737 such messages specify what the cause of the deferral is. However, some
2738 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2739 that caused an old message to be retained and used if a later statement
2740 caused a defer, replacing the real cause of the deferral.
2742 16. If an IP address had so many PTR records that the DNS lookup buffer
2743 was not large enough to hold them, Exim could crash while trying to process
2744 the truncated data. It now detects and logs this case.
2746 17. Further to 4.21/58, another change has been made: if (and only if) the
2747 first line of a message (the first header line) ends with CRLF, a bare LF
2748 in a subsequent header line has a space inserted after it, so as not to
2749 terminate the header.
2751 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2752 unnecessarily, used atoi() instead of strtol(), and didn't check the
2753 termination when getting file sizes from file names by regex.
2755 19. Completely re-implemented the support for maildirsize files, in the light
2756 of a number of problems with the previous contributed implementation
2757 (4.30/29). In particular:
2759 . If the quota is zero, the maildirsize file is maintained, but no quota is
2762 . If the maildir directory does not exist, it is created before any attempt
2763 to write a maildirsize file.
2765 . The quota value in the file is just a cache; if the quota is changed in
2766 the transport, the new value overrides.
2768 . A regular expression is available for excluding directories from the
2771 20. The autoreply transport checks the characters in options that define the
2772 message's headers; it allows continued headers, but it was checking with
2773 isspace() after an embedded newline instead of explicitly looking for a
2776 21. If all the "regular" hosts to which an address was routed had passed their
2777 expiry times, and had not reached their retry times, the address was
2778 bounced, even if fallback hosts were defined. Now Exim should go on to try
2781 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2782 equivalent code in the SMTP transport. Some hosts send humungous responses
2783 to HELO/EHLO, more than 1024 it seems.
2785 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2786 gives compiler warnings in some environments. I've now done it "properly",
2789 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2790 (because of problems with the built-in one) was declared to return uschar *
2791 instead of char *, causing compiler failure.
2793 25. Fixed a file descriptor leak when processing alias/forward files.
2795 26. Fixed a minor format string issue in dbfn.c.
2797 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2799 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2800 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2803 29. When a sender address is verified, it is cached, to save repeating the test
2804 when there is more than one recipient in a message. However, when the
2805 verification involves a callout, it is possible for different callout
2806 options to be set for different recipients. It is too complicated to keep
2807 track of this in the cache, so now Exim always runs a verification when a
2808 callout is required, relying on the callout cache for the optimization.
2809 The overhead is duplication of the address routing, but this should not be
2812 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2813 to be verified with callout=postmaster, and the main callout worked but the
2814 postmaster check failed, the verification correctly failed. However, if a
2815 subsequent RCPT command asked for sender verification *without* the
2816 postmaster check, incorrect caching caused this verification also to fail,
2819 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2820 it was not caching the DNS options (qualify_single, search_parents) that
2821 were used when the lookup failed. A subsequent lookup with different
2822 options therefore always gave the same answer, though there were cases
2823 where it should not have. (Example: a "domains = !$mx_any" option on a
2824 dnslookup router: the "domains" option is always processed without any
2825 widening, but the router might have qualify_single set.) Now Exim uses the
2826 cached value only when the same options are set.
2828 32. Added John Jetmore's "exipick" utility to the distribution.
2830 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2831 than a timeout (e.g. a certificate is required, and is not provided), an
2832 Exim server now closes the connection immediately. Previously it waited for
2833 the client to close - but if the client is SSL, it seems that they each
2834 wait for each other, leading to a delay before one of them times out.
2836 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2837 maintained 0.8.x compatibility because I don't think many are using it, and
2838 it is clearly obsolete.
2840 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2843 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2844 client certificate was expired. A simple patch fixes this, though I don't
2845 understand the full logic of why the verify callback is called multiple
2848 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2849 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2850 which causes problems with some clients (such as the Certicom SSL Plus
2851 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2852 disables the coutermeasure allowing Eudora to connect."
2854 38. Exim was not checking that a write() to a log file succeeded. This could
2855 lead to Bad Things if a log got too big, in particular if it hit a file
2856 size limit. Exim now panics and dies if it cannot write to a log file, just
2857 as it does if it cannot open a log file.
2859 39. Modified OS/Makefile-Linux so that it now contains
2861 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2863 The two -D definitions ensure that Exim is compiled with large file
2864 support, which makes it possible to handle log files that are bigger than
2867 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2868 instance) a domain was checked against a named list that involved a lookup,
2869 causing $domain_data to be set, then another domain was checked against the
2870 same list, then the first domain was re-checked, the value of $domain_data
2871 after the final check could be wrong. In particular, if the second check
2872 failed, it could be set empty. This bug probably also applied to
2875 41. The strip_trailing_dot option was not being applied to the address given
2876 with the -f command-line option.
2878 42. The code for reading a message's header from the spool was incrementing
2879 $received_count, but never initializing it. This meant that the value was
2880 incorrect (doubled) while delivering a message in the same process in which
2881 it was received. In the most common configuration of Exim, this never
2882 happens - a fresh exec is done - but it can happen when
2883 deliver_drop_privilege is set.
2885 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2886 it now includes up to 150 characters of the unexpected data in the log
2889 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2890 and building data strings. The size of both of these buffers was 10 000
2891 bytes - far larger than anybody would *ever* want, thought I. Needless to
2892 say, somebody hit the limit. I have increased the maximum line length to
2893 20 000 and the maximum data length of concatenated lines to 100 000. I have
2894 also fixed two bugs, because there was no checking on these buffers. Tsk,
2895 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2896 buffer is too small.
2898 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2899 lsearch lookups. Now it does.
2901 46. When parsing a route_list item in a manualroute router, a fixed-length
2902 buffer was used for the list of hosts. I made this 1024 bytes long,
2903 thinking that nobody would ever have a list of hosts that long. Wrong.
2904 Somebody had a whole pile of complicated expansion conditions, and the
2905 string was silently truncated, leading to an expansion error. It turns out
2906 that it is easier to change to an unlimited length (owing to other changes
2907 that have happened since this code was originally written) than to build
2908 structure for giving a limitation error. The length of the item that
2909 expands into the list of hosts is now unlimited.
2911 47. The lsearch lookup could not handle data where the length of text line was
2912 more than 4095 characters. Such lines were truncated, leading to shortened
2913 data being returned. It should now handle lines of any length.
2915 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2916 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2919 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2920 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2921 no longer happens in comment lines. A list of replacements is now placed
2922 at the head of all of the source files, except those whose only change is
2923 to replace PERL_COMMAND in the very first #! line.
2925 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2926 messages on the queue, with a bottom-up merge sort, using code contributed
2927 by Michael Haardt. This should make operations like -bp somewhat faster on
2928 large queues. It won't affect queue runners, except when queue_run_in_order
2931 51. Installed eximstats 1.31 in the distribution.
2933 52. Added support for SRV lookups to the dnslookup router.
2935 53. If an ACL referred to $message_body or $message_body_end, the value was not
2936 reset for any messages that followed in the same SMTP session.
2938 54. The store-handling optimization for building very long strings was not
2939 differentiating between the different store pools. I don't think this
2940 actually made any difference in practice, but I've tidied it.
2942 55. While running the routers to verify a sender address, $sender_address
2943 was still set to the sender address. This is wrong, because when routing to
2944 send a bounce to the sender, it would be empty. Therefore, I have changed
2945 it so that, while verifying a sender address, $sender_address is set to <>.
2946 (There is no change to what happens when verifying a recipient address.)
2948 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2949 target A or AAAA records (if not already returned) without resetting the
2950 qualify_single or search_parents options of the DNS resolver. These are
2951 inappropriate in this case because the targets of MX and SRV records must
2952 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2953 target that, when qualified, matched something in the local domain. These
2954 two options are now turned off when doing these lookups.
2956 57. It seems that at least some releases of Reiserfs (which does not have the
2957 concept of a fixed number of inodes) returns zero and not -1 for the
2958 number of available inodes. This interacted badly with check_spool_inodes,
2959 which assumed that -1 was the "no such thing" setting. What I have done is
2960 to check that the total number of inodes is greater than zero before doing
2961 the test of how many are available.
2963 58. When a "warn" ACL statement has a log_message modifier, the message is
2964 remembered, and not repeated. This is to avoid a lot of repetition when a
2965 message has many recipients that cause the same warning to be written.
2966 Howewer, Exim was preserving the list of already written lines for an
2967 entire SMTP session, which doesn't seem right. The memory is now reset if a
2968 new message is started.
2970 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2971 debugging output unless log_rewrite was also set.
2973 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2974 of (int)(handle) when we know that handle contains (void *)(-1).
2976 61. The Exim daemon panic-logs an error return when it closes the incoming
2977 connection. However "connection reset by peer" seems to be common, and
2978 isn't really an error worthy of noting specially, so that particular error
2981 62. When Exim is trying to find all the local interfaces, it used to panic and
2982 die if the ioctl to get the interface flags failed. However, it seems that
2983 on at least one OS (Solaris 9) it is possible to have an interface that is
2984 included in the list of interfaces, but for which you get a failure error
2985 for this call. This happens when the interface is not "plumbed" into a
2986 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2987 failure of the "get flags" call assumes that the interface is down.
2989 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2990 makes life easier for people who are doing arithmetic on fields extracted
2991 from dates, where you often get leading zeros that should not be
2992 interpreted as octal.
2994 64. Added qualify_domain to the redirect router, to override the global
2997 65. If a pathologically long header line contained very many addresses (the
2998 report of this problem mentioned 10 000) and each of them was rewritten,
2999 Exim could use up a very large amount of memory. (It kept on making new
3000 copies of the header line as it rewrote, and never released the old ones.)
3001 At the expense of a bit more processing, the header rewriting function has
3002 been changed so that it no longer eats memory in this way.
3004 66. The generation of the Received: header has been moved from the time that a
3005 message starts to be received, to the time that it finishes. The timestamp
3006 in the Received: header should now be very close to that of the <= log
3007 line. There are two side-effects of this change:
3009 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3010 logged header lines no longer include the local Received: line, because
3011 it has not yet been created. The same applies to a copy of the message
3012 that is returned to a non-SMTP sender when a message is rejected.
3014 (b) When a filter file is tested using -bf, no additional Received: header
3015 is added to the test message. After some thought, I decided that this
3018 This change does not affect the value of $received_for. It is still set
3019 after address rewriting, but before local_scan() is called.
3021 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3023 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3024 gave an unhelpful panic error message, and a defer error. I have managed to
3025 change this behaviour so that it now rejects any supplied certificate,
3026 which seems right, as the list of acceptable certificates is empty.
3028 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3029 gave an unhelpful defer error. I have not managed to make this reject any
3030 supplied certificates, but the error message it gives is "no certificate
3031 supplied", which is not helpful.
3033 70. exigrep's output now also includes lines that are not associated with any
3034 message, but which match the given pattern. Implemented by a patch from
3035 Martin Sluka, which also tidied up the Perl a bit.
3037 71. Recipient callout verification, like sender verification, was using <> in
3038 the MAIL FROM command. This isn't really the right thing, since the actual
3039 sender may affect whether the remote host accepts the recipient or not. I
3040 have changed it to use the actual sender in the callout; this means that
3041 the cache record is now keyed on a recipient/sender pair, not just the
3042 recipient address. There doesn't seem to be a real danger of callout loops,
3043 since a callout by the remote host to check the sender would use <>.
3044 [SEE ABOVE: changed after hitting problems.]
3046 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3047 temporary errors. However, in the case of such a code being given after
3048 the end of a data transmission (i.e. after ".") Exim was failing to write
3049 a retry record for the message. (Yes, there was some broken host that was
3050 actually sending 8xx at this point.)
3052 73. An unknown lookup type in a host list could cause Exim to panic-die when
3053 the list was checked. (An example that provoked this was putting <; in the
3054 middle of a list instead of at the start.) If this happened during a DATA
3055 ACL check, a -D file could be left lying around. This kind of configuration
3056 error no longer causes Exim to die; instead it causes a defer errror. The
3057 incident is still logged to the main and panic logs.
3059 74. Buglet left over from Exim 3 conversion. The message "too many messages
3060 in one connection" was written to the rejectlog but not the mainlog, except
3061 when address rewriting (yes!) was being logged.
3063 75. Added write_rejectlog option.
3065 76. When a system filter was run not as root (that is, when system_filter_user
3066 was set), the values of the $n variables were not being returned to the
3067 main process; thus, they were not subsequently available in the $sn
3070 77. Added +return_path_on_delivery log selector.
3072 78. A connection timeout was being treated differently from recipients deferred
3073 when testing hosts_max_try with a message that was older than the host's
3074 retry timeout. (The host should not be counted, thus allowing all hosts to
3075 be tried at least once before bouncing.) This may have been the cause of an
3076 occasionally reported bug whereby a message would remain on the queue
3077 longer than the retry timeout, but would be bounced if a delivery was
3078 forced. I say "may" because I never totally pinned down the problem;
3079 setting up timeout/retry tests is difficult. See also the next item.
3081 79. The ultimate address timeout was not being applied to errors that involved
3082 a combination of host plus message (for example, a timeout on a MAIL
3083 command). When an address resolved to a number of possible hosts, and they
3084 were not all tried for each delivery (e.g. because of hosts_max_try), a
3085 message could remain on the queue longer than the retry timeout.
3087 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3090 81. Fixed an obscure SMTP outgoing bug which required at least the following
3091 conditions: (a) there was another message waiting for the same server;
3092 (b) the server returned 5xx to all RCPT commands in the first message so
3093 that the message was not completed; (c) the server dropped the connection
3094 or gave a negative response to the RSET that Exim sends to abort the
3095 transaction. The observed case was a dropped connection after DATA that had
3096 been sent in pipelining mode. That is, the server had advertised PIPELINING
3097 but was not implementing it correctly. The effect of the bug was incorrect
3098 behaviour, such as trying another host, and this could lead to a crash.
3104 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3105 and daemon.c were passed as pointers to ints; they should have been
3106 pointers to socklen_t variables (which are typically unsigned ints).
3108 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3111 3. Fixed a really odd bug that affected only the testing scheme; patching a
3112 certain fixed string in the binary changed the value of another string that
3113 happened to be identical to the end of the original first string.
3115 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3116 name", it returns that address as the IP address. On some operating
3117 systems (e.g. Solaris), it also passes back the IP address string as the
3118 "host name". However, on others (e.g. Linux), it passes back an empty
3119 string. Exim wasn't checking for this, and was changing the host name to an
3120 empty string, assuming it had been canonicized.
3122 5. Although rare, it is permitted to have more than one PTR record for a given
3123 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3124 all the names associated with an address, because they do in Solaris.
3125 However, it seems that they do not in Linux for data that comes from the
3126 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3127 I found this out when I moved to a new Linux workstation and tried to run
3128 the Exim test suite.
3130 To get round this problem I have changed the code so that it now does its
3131 own call to the DNS to look up PTR records when searching for a host name.
3132 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3133 addresses that are only in /etc/hosts are still found.
3135 This behaviour is, however, controlled by an option called host_lookup_
3136 order, which defaults to "bydns:byaddr". If people want to use the other
3137 order, or indeed, just use one or the other means of lookup, they can
3138 specify it in this variable.
3140 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3141 some operating systems, this comes back from gethostbyaddr() as an empty
3142 string, and this is what Exim used to test for. However, it seems that in
3143 other systems, "." is yielded. Exim now tests for this case too.
3145 7. The values of check_spool_space and check_log_space are now held internally
3146 as a number of kilobytes instead of an absolute number of bytes. If a
3147 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3148 kilobyte. This means that much larger values can be stored.
3150 8. Exim monitor: an attempt to get the action menu when not actually pointing
3151 at a message produces an empty menu entitled "No message selected". This
3152 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3153 no entries in it ("Shell widget menu has zero width and/or height"). So I
3154 have added a single, blank menu entry in this case.
3156 9. Added ${quote_local_part.
3158 10. MIME decoding is now applied to the contents of Subject: header lines when
3161 11. Now that a reference to $sender_host_address automatically causes a reverse
3162 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3163 host lookup before query-style lookups in lists that might use this
3164 variable. This has therefore been abolished, and the "net-" prefix is no
3165 longer necessary for query-style lookups.
3167 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3168 have been a typo for LFLAGS, so it has been changed.
3170 13. The install script calls Exim with "-C /dev/null" in order to find the
3171 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3172 to be output. Howeve, since Exim outputs its version number before the
3173 error, it didn't break the script. It just looked ugly. I fixed this by
3174 always allowing "-C /dev/null" if the caller is root.
3176 14. Ignore overlarge ACL variable number when reading spool file - insurance
3177 against a later release with more variables having written the file.
3179 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3180 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3181 this, as well as the form without the "IPv6" on the front.
3183 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3184 OS/Makefile-Darwin file.
3186 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3187 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3189 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3190 variable. However, some OS use ints for inodes. I've added cast to long int
3191 to get rid of the compiler warning.
3193 19. I had forgotten to lock out "/../" in configuration file names when
3194 ALT_CONFIG_PREFIX was set.
3196 20. Routers used for verification do not need to specify transports. However,
3197 if such a router generated a host list, and callout was configured, Exim
3198 crashed, because it could not find a port number from the (non-existent)
3199 transport. It now assumes port 25 in this circumstance.
3201 21. Added the -t option to exigrep.
3203 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3204 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3205 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3206 them is set, LOOKUP_LSEARCH is forced.
3208 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3210 24. Added sender and host information to the "rejected by local_scan()" log
3211 line; previously there was no indication of these.
3213 25. Added .include_if_exists.
3215 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3216 Linux. It turns out that not all file systems support this. Apparently some
3217 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3218 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3219 not supported on the file descriptor, is now ignored when Exim is trying to
3220 sync a directory. This applies only to Linux.
3222 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3224 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3225 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3228 29. Added support for maildirsize files from supplied patch (modified a bit).
3230 30. The use of :fail: followed by an empty string could lead Exim to respond to
3231 sender verification failures with (e.g.):
3233 550 Verification failed for <xxx>
3234 550 Sender verify failed
3236 where the first response line was missing the '-' that indicates it is not
3237 the final line of the response.
3239 31. The loop for finding the name of the user that called Exim had a hardwired
3240 limit of 10; it now uses the value of finduser_retries, which is used for
3241 all other user lookups.
3243 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3246 33. Exim was neglecting to zero errno before one call of strtol() when
3247 expanding a string and expecting an integer value. On some systems this
3248 resulted in spurious "integer overflow" errors. Also, it was casting the
3249 result into an int without checking.
3251 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3252 did not work. The code looks as if it has *never* worked, though it appears
3253 to have been documented since at least releast 1.62. I have made it work.
3255 35. The "timeout_DNS" error in retry rules, also documented since at least
3256 1.62, also never worked. As it isn't clear exactly what this means, and
3257 clearly it isn't a major issue, I have abolished the feature by treating it
3258 as "timeout", and writing a warning to the main and panic logs.
3260 36. The display of retry rules for -brt wasn't always showing the error code
3263 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3264 timeout_connect_A, timeout_connect_MX.
3266 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3267 to the empty sender.
3269 39. The daemon was not analysing the content of -oX till after it had closed
3270 stderr and disconnected from the controlling terminal. This meant that any
3271 syntax errors were only noted on the panic log, and the return code from
3272 the command was 0. By re-arranging the code a little, I've made the
3273 decoding happen first, so such errors now appear on stderr, and the return
3274 code is 1. However, the actual setting up of the sockets still happens in
3275 the disconnected process, so errors there are still only recorded on the
3278 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3279 connections (as happens on some IP stacks) was logged at start up time as
3280 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3281 it from "IPv6 and IPv4", which means that two separate sockets are being
3284 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3285 says whether AF_INET or AF_INET6 was passed as an argument.
3287 42. Exiwhat output was messed up when time zones were included in log
3290 43. Exiwhat now gives more information about the daemon's listening ports,
3291 and whether -tls-on-connect was used.
3293 44. The "port" option of the smtp transport is now expanded.
3295 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3296 silently ignored. Now an error message is written to the main and panic
3299 46. There's a new ACL modifier called "logwrite" which writes to a log file
3300 as soon as it is encountered.
3302 47. Added $local_user_uid and $local_user_gid at routing time.
3304 48. Exim crashed when trying to verify a sender address that was being
3307 49. Exim was recognizing only a space character after ".include". It now also
3308 recognizes a tab character.
3310 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3311 extracting the relevant information from the specification. The man page no
3312 longer contains scrambled data for the -d option, and I've added a section
3313 at the front about calling Exim under different names.
3315 51. Added "extra_headers" argument to the "mail" command in filter files.
3317 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3320 53. Installed eximstats 1.29.
3322 54. Added transport_filter_timeout as a generic transport option.
3324 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3325 Cc: header lines. This was required by RFC 822, but it not required by RFC
3328 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3329 incoming messages that did not have them. Now it does so only if the
3330 message originates locally, that is, if there is no associated remote host
3331 address. When Resent- header lines are present, this applies to the Resent-
3332 lines rather than the non-Resent- lines.
3334 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3335 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3337 58. Messages for configuration errors now include the name of the main
3338 configuration file - useful now that there may be more than one file in a
3339 list (.included file names were always shown).
3341 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3342 for those rare installations that do not start the daemon as root or run it
3343 setuid root. I've cut out the call to initgroups() if the daemon is not
3346 60. The Exim user and group can now be bound into the binary as text strings
3347 that are looked up at the start of Exim's processing.
3349 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3351 62. Added $mailstore_basename variable.
3353 63. Installed patch to sieve.c from Michael Haardt.
3355 64. When Exim failed to open the panic log after failing to open the main log,
3356 the original message it was trying to log was written to stderr and debug
3357 output, but if they were not available (the usual case in production), it
3358 was lost. Now it is written to syslog before the two lines that record the
3359 failures to open the logs.
3361 65. Users' Exim filters run in subprocesses under the user's uid. It is
3362 possible for a "deliver" command or an alias in a "personal" command to
3363 provoke an address rewrite. If logging of address rewriting is configured,
3364 this fails because the process is not running as root or exim. There may be
3365 a better way of dealing with this, but for the moment (because 4.30 needs
3366 to be released), I have disabled address rewrite logging when running a
3367 filter in a non-root, non-exim process.
3373 1. The buildconfig auxiliary program wasn't quoting the value set for
3374 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3375 not defined. This bug was masked in 4.22 by the effect that was fixed in
3378 2. Some messages that were rejected after a message id was allocated were
3379 shown as "incomplete" by exigrep. It no longer does this for messages that
3380 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3382 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3383 have allow_domain_literals set, the ID did not get logged in the <= line.
3384 Domain literals are now always recognized in Message-ID: header lines.
3386 4. The first argument for a ${extract expansion item is the key name or field
3387 number. Leading and trailing spaces in this item were not being ignored,
3388 causing some misleading effects.
3390 5. When deliver_drop_privilege was set, single queue runner processes started
3391 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3392 same command in the process it spins off) were not dropping privilege.
3394 6. When the daemon running as "exim" started a queue runner, it always
3395 re-executed Exim in the spun-off process. This is a waste of effort when
3396 deliver_drop_privilege is set. The new process now just calls the
3397 queue-runner function directly.
3403 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3406 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3407 ignored. Though the use of -oP was forcing the writing of a pid file, it
3408 was always written to the default place.
3410 3. If the message "no IP address found for host xxxx" is generated during
3411 incoming verification, it is now followed by identification of the incoming
3412 connection (so you can more easily find what provoked it).
3414 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3416 5. Added some features to "harden" Exim a bit more against certain attacks:
3418 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3419 be put in Local/Makefile. This is like the never_users runtime option,
3420 but it cannot be overridden. The default setting is "root".
3422 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3423 prefix string with which any file named in a -C command line option
3426 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3427 is retained for -C and -D only if the caller of Exim is root. Without
3428 it, the exim user may also use -C and -D and retain privilege.
3430 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3431 command line option is disabled.
3433 6. Macro names set by the -D option must start with an upper case letter, just
3434 like macro names defined in the configuration file.
3436 7. Added "dereference=" facility to LDAP.
3438 8. Two instances of the typo "uknown" in the source files are fixed.
3440 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3441 the Configure-Makefile script screwed up while processing it.
3443 10. Incorporated PCRE 4.4.
3445 11. The SMTP synchronization check was not operating right at the start of an
3446 SMTP session. For example, it could not catch a HELO sent before the client
3447 waited for the greeting. There is now a check for outstanding input at the
3448 point when the greeting is written. Because of the duplex, asynchronous
3449 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3450 way, but not yet received, when the check is performed.
3452 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3453 on TCP/IP sockets, because this apparently causes some broken clients to
3456 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3457 unchanged) from the Cygwin maintainer.
3459 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3460 was supported instead of testing for mbx. Effectively a typo.
3462 15. The spa authenticator server code was not checking that the input it
3463 received was valid base64.
3465 16. The debug output line for the "set" modifier in ACLs was not showing the
3466 name of the variable that was being set.
3468 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3470 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3471 lookup on its own. Something else was needed to trigger the lookup. For
3472 example, a match in host_lookup or the need for a host name in a host list.
3473 Now, if $sender_host_name is referenced and the host name has not yet been
3474 looked up, a lookup is performed. If the lookup fails, the variable remains
3475 empty, and $host_lookup_failed is set to "1".
3477 19. Added "eqi" as a case-independent comparison operator.
3479 20. The saslauthd authentication condition could segfault if neither service
3480 nor realm was specified.
3482 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3483 error message that was logged was misleading, and incoming SMTP
3484 connections were dropped. The message is now more accurate, and temporary
3485 errors are given to SMTP connections.
3487 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3488 (except RSET) with a 421 error, until QUIT is received. However, it was
3489 failing to send a response to QUIT.
3491 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3492 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3493 ACL is now run after the helo_try_verify_hosts code.
3495 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3496 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3497 case-independent, so other case variants are also recognized). Apparently
3498 some systems use these upper case variants.
3500 25. If more than two messages were waiting for the same host, and a transport
3501 filter was specified for the transport, Exim sent two messages over the
3502 same TCP/IP connection, and then failed with "socket operation on non-
3503 socket" when it tried to send the third.
3505 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3507 27. The extern definition of crypt16() in expand.c was not being excluded when
3508 the OS had its own crypt16() function.
3510 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3511 as a preferred synonym for return_size_limit, both as an option and as an
3514 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3516 30. Changed the default configuration ACL to relax the local part checking rule
3517 for addresses that are not in any local domains. For these addresses,
3518 slashes and pipe symbols are allowed within local parts, but the sequence
3519 /../ is explicitly forbidden.
3521 31. SPA server authentication was not clearing the challenge buffer before
3524 32. log_message in a "warn" ACL statement was writing to the reject log as
3525 well as to the main log, which contradicts the documentation and doesn't
3526 seem right (because no rejection is happening). So I have stopped it.
3528 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3529 However, I am unable to do any testing of this.
3531 34. Fixed an infelicity in the appendfile transport. When checking directories
3532 for a mailbox, to see if any needed to be created, it was accidentally
3533 using path names with one or more superfluous leading slashes; tracing
3534 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3536 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3537 contents are added to the log line that is written for every discarded
3538 recipient. (Previously a log_message setting was ignored.)
3540 36. The ${quote: operator now quotes the string if it is empty.
3542 37. The install script runs exim in order to find its version number. If for
3543 some reason other than non-existence or emptiness, which it checks, it
3544 could not run './exim', it was installing it with an empty version number,
3545 i.e. as "exim-". This error state is now caught, and the installation is
3548 38. An argument was missing from the function that creates an error message
3549 when Exim fails to connect to the socket for saslauthd authentication.
3550 This could cause Exim to crash, or give a corrupted message.
3552 39. Added isip, isip4, and isip6 to ${if conditions.
3554 40. The ACL variables $acl_xx are now saved with the message, and can be
3555 accessed later in routers, transports, and filters.
3557 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3558 strings in the file are not string-expanded.
3560 42. If a MAIL command specified a SIZE value that was too large to fit into an
3561 int variable, the check against message_size_limit failed. Such values are
3562 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3563 day this will have to be increased, but I don't think I want to be around
3564 when emails are that large.
3571 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3572 iconv() is not standard in FreeBSD.
3574 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3575 IPv6 enabled. The observed symptom was a segmentation fault on return from
3576 the function os_common_find_running_interfaces() in src/os.c.
3578 3. In the check_special_case() function in daemon.c I had used "errno" as an
3579 argument name, which causes warnings on some systems. This was basically a
3580 typo, since it was named "eno" in the comments!
3582 4. The code that waits for the clock to tick (at a resolution of some fraction
3583 of a second) so as to ensure message-id uniqueness was always waiting for
3584 at least one whole tick, when it could have waited for less. [This is
3585 almost certainly not relevant at current processor speeds, where it is
3586 unlikely to ever wait at all. But we try to future-proof.]
3588 5. The function that sleeps for a time interval that includes fractions of a
3589 second contained a race. It did not block SIGALRM between setting the
3590 timer, and suspending (a couple of lines later). If the interval was short
3591 and the sigsuspend() was delayed until after it had expired, the suspension
3592 never ended. On busy systems this could lead to processes getting stuck for
3595 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3596 process, before it forks any delivery processes. The open lookup caching
3597 mechanism meant that the open file or database connection was passed into
3598 the delivery process. The problem was that delivery processes always tidy
3599 up cached lookup data. This could cause a problem for the next delivery
3600 process started by the queue runner, because the external queue runner
3601 process does not know about the closure. So the next delivery process
3602 still has data in the lookup cache. In the case of a file lookup, there was
3603 no problem because closing a file descriptor in a subprocess doesn't affect
3604 the parent. However, if the lookup was caching a connection to a database,
3605 the connection was closed, and the second delivery process was likely to
3606 see errors such as "PGSQL: query failed: server closed the connection
3607 unexpectedly". The problem has been fixed by closing all cached lookups
3608 in a queue runner before running a delivery process.
3610 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3611 seem to have the "const" qualifier which it has on other OS. I've
3614 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3615 *of the same type* (client or server) with the same public name that an
3616 error should be diagnosed.
3618 9. When Exim looked up a host name for an IP address, but failed to find the
3619 original IP address when looking up the host name (a safety check), it
3620 output the message "<ip address> does not match any IP for NULL", which was
3621 confusing, to say the least. The bug was that the host name should have
3622 appeared instead of "NULL".
3624 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3625 user that is built into the binary, and the -C or -D options is used, root
3626 privilege is dropped before the configuration file is read. In addition,
3627 logging is switched to stderr instead of the normal log files. If the
3628 configuration then re-defines the Exim user, the unprivileged environment
3629 is probably not what is expected, so Exim logs a panic warning message (but
3632 However, if deliver_drop_privilege is set, the unprivileged state may well
3633 be exactly what is intended, so the warning has been cut out in that case,
3634 and Exim is allowed to try to write to its normal log files.
3640 1. smtp_return_error_details was not giving details for temporary sender
3641 or receiver verification errors.
3643 2. Diagnose a configuration error if two authenticators have the same public
3646 3. Exim used not to create the message log file for a message until the first
3647 delivery attempt. This could be confusing when incoming messages were held
3648 for policy or load reasons. The message log file is now created at the time
3649 the message is received, and an initial "Received" line is written to it.
3651 4. The automatically generated man page for command line options had a minor
3652 bug that caused no ill effects; however, a more serious problem was that
3653 the procedure for building the man page automatically didn't always
3654 operate. Consequently, release 4.20 contains an out-of-date version. This
3655 shouldn't happen again.
3657 5. When building Exim with embedded Perl support, the script that builds the
3658 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3659 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3661 6. The freeze_tell option was not being used for messages that were frozen on
3662 arrival, either by an ACL or by local_scan().
3664 7. Added the smtp_incomplete_transaction log selector.
3666 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3667 was accepting AUTH without a new EHLO.
3669 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3670 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3673 10. Logging of TCP/IP connections (when configured) now happens in the main
3674 daemon process instead of the child process, so that the TCP/IP connection
3675 count is more accurate (but it can never be perfect).
3677 11. The use of "drop" in a nested ACL was not being handled correctly in the
3678 outer ACL. Now, if condition failure induced by the nested "drop" causes
3679 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3680 or "require"), the connection is dropped.
3682 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3683 that yield "discard" can now be used with an "accept" or a "discard" verb,
3684 but an error is generated for any others (because I can't see a useful way
3685 to define what should happen).
3687 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3688 are now processed in the same way as lines in the Exim configuration file.
3689 In particular, continuation lines are supported.
3691 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3693 15. Added -ti meaning -t -i.
3695 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3696 domains, and warn by logging if others are found.
3698 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3699 returned from GIFCONF when Exim is trying to find the list of interfaces on
3700 a host. The code in os.c has been modified to copy each ifreq to an aligned
3701 structure in all cases.
3703 Also, in some cases, the returned ifreq's were being copied to a 'struct
3704 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3705 means the last couple of bytes of an IPv6 address could be chopped if the
3706 ifreq contained only a normal sockaddr (14 bytes storage).
3708 18. Named domain lists were not supported in the hosts_treat_as_local option.
3709 An entry such as +xxxx was not recognized, and was treated as a literal
3712 19. Ensure that header lines added by a DATA ACL are included in the reject log
3713 if the ACL subsequently rejects the message.
3715 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3716 MD5 (which is deprecated).
3718 21. When testing a filter file using -bf, Exim was writing a message when it
3719 took the sender from a "From " line in the message, but it was not doing so
3720 when it took $return_path from a Return-Path: header line. It now does.
3722 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3723 with a valid header line field name (a series of printing characters
3724 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3726 23. Changed "disc" in the source to "disk" to conform to the documentation and
3727 the book and for uniformity.
3729 24. Ignore Sendmail's -Ooption=value command line item.
3731 25. When execve() failed while trying to run a command in a pipe transport,
3732 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3733 could be confused with a return value of 69 from the command itself. This
3734 has been changed to 127, the value the shell returns if it is asked to run
3735 a non-existent command. The wording for the related log line suggests a
3736 non-existent command as the problem.
3738 26. If received_header_text expands to an empty string, do not add a Received:
3739 header line to the message. (Well, it adds a token one on the spool, but
3740 marks it "old" so that it doesn't get used or transmitted.)
3742 27. Installed eximstats 1.28 (addition of -nt option).
3744 28. There was no check for failure on the call to getsockname() in the daemon
3745 code. This can fail if there is a shortage of resources on the system, with
3746 ENOMEM, for example. A temporary error is now given on failure.
3748 29. Contrary to the C standard, it seems that in some environments, the
3749 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3750 program. Exim now does this explicitly; it affects the formatting of
3751 timestamps using strftime().
3753 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3754 complaints. I've now initialized all the variables, to avoid this.
3756 32. Header lines added by a system filter were not being "seen" during
3757 transport-time rewrites.
3759 33. The info_callback() function passed to OpenSSL is set up with type void
3760 (*)(SSL *, int, int), as described somewhere. However, when calling the
3761 function (actually a macro) that sets it up, the type void(*)() is
3762 expected. I've put in a cast to prevent warnings from picky compilers.
3764 34. If a DNS black list lookup found a CNAME record, but there were no A
3765 records associated with the domain it pointed at, Exim crashed.
3767 35. If a DNS black list lookup returned more than one A record, Exim ignored
3768 all but the first. It now scans all returned addresses if a particular IP
3769 value is being sought. In this situation, the contents of the
3770 $dnslist_value variable are a list of all the addresses, separated by a
3773 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3774 to accept a host name and all its aliases if the forward lookup for any of
3775 them yielded the IP address of the incoming connection. Now it accepts only
3776 those names whose forward lookup yields the correct IP address. Any other
3777 names are discarded. This closes a loophole whereby a rogue DNS
3778 administrator could create reverse DNS records to break through a
3779 wildcarded host restriction in an ACL.
3781 37. If a user filter or a system filter that ran in a subprocess used any of
3782 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3783 the wrong values were passed to the pipe command ($thisaddress had the
3784 value of $0, $0 had the value of $1, etc). This bug was introduced by
3785 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3787 38. Improved the line breaking for long SMTP error messages from ACLs.
3788 Previously, if there was no break point between 40 and 75 characters, Exim
3789 left the rest of the message alone. Two changes have been made: (a) I've
3790 reduced the minimum length to 35 characters; (b) if it can't find a break
3791 point between 35 and 75 characters, it looks ahead and uses the first one
3792 that it finds. This may give the occasional overlong line, but at least the
3793 remaining text gets split now.
3795 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3796 file descriptors might be low, and that setting 1000 would always raise it.
3797 It turns out that in some environments, the limit is already over 1000 and
3798 that lowering it causes trouble. So now Exim takes care not to decrease it.
3800 40. When delivering a message, the value of $return_path is set to $sender_
3801 address at the start of routing (routers may change the value). By an
3802 oversight, this default was not being set up when an address was tested by
3803 -bt or -bv, which affected the outcome if any router or filter referred to
3806 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3807 log only when "message" or "log_message" are set. However, if one of the
3808 conditions was an address verification, or a call to a nested ACL, the
3809 messages generated by the underlying test were being passed through. This
3810 no longer happens. The underlying message is available in $acl_verify_
3811 message for both "message" and "log_message" expansions, so it can be
3812 passed through if needed.
3814 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3815 new expansion $bh_ to give the encoded byte string without charset
3816 translation. Translation happens only if iconv() is available; HAVE_ICONV
3817 indicates this at build time. HEADERS_CHARSET gives the charset to
3818 translate to; headers_charset can change it in the configuration, and
3819 "headers charset" can change it in an individual filter file.
3821 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3822 that creates RFC 2047 encoded "words" labels them as that charset instead
3823 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3824 expansion operator; (ii) when Exim creates a From: line for a local
3825 message; (iii) when a header line is rewritten to include a "phrase" part.
3827 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3828 buggy, causing it to skip the first lines of messages whose message ID
3829 ended in 'D'. This would not have bitten before Exim release 4.14, because
3830 message IDs were unlikely to end in 'D' before then. The effect was to have
3831 incorrect size information for certain domains.
3833 45. #include "config.h" was missing at the start of the crypt16.c module. This
3834 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3837 46. If there was a timeout during a "random" callout check, Exim treated it as
3838 a failure of the random address, and carried on sending RSET and the real
3839 address. If the delay was just some slowness somewhere, the response to the
3840 original RCPT would be taken as a response to RSET and so on, causing
3841 mayhem of various kinds.
3843 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3844 when I implemented it. It didn't allow for the fact that some option values
3845 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3846 the right test for positive values.
3848 48. Domain names in DNS records are case-independent. Exim always looks them up
3849 in lower case. Some resolvers return domain names in exactly the case they
3850 appear in the zone file, that is, they may contain uppercase letters. Not
3851 all resolvers do this - some return always lower case. Exim was treating a
3852 change of case by a resolver as a change of domain, similar to a widening
3853 of a domain abbreviation. This triggered its re-routing code and so it was
3854 trying to route what was effectively the same domain again. This normally
3855 caused routing to fail (because the router wouldn't handle the domain
3856 twice). Now Exim checks for this case specially, and just changes the
3857 casing of the domain that it ultimately uses when it transmits the message
3860 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3863 50. If a filter generated a file delivery with a non-absolute name (possible if
3864 no home directory exists for the router), the forbid_file option was not
3867 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3868 the existing equality matching.
3870 52. Exim was using ints instead of ino_t variables in some places where it was
3871 dealing with inode numbers.
3873 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3874 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3875 called TMPDIR, and if it finds it is different, it changes its value.
3877 54. The smtp_printf() function is now made available to local_scan() so
3878 additional output lines can be written before returning. There is also an
3879 smtp_fflush() function to enable the detection of a dropped connection.
3880 The variables smtp_input and smtp_batched_input are exported to
3883 55. Changed the default runtime configuration: the message "Unknown user"
3884 has been removed from the ACL, and instead placed on the localuser router,
3885 using the cannot_route_message feature. This means that any verification
3886 failures that generate their own messages won't get overridden. Similarly,
3887 the "Unrouteable address" message that was in the ACL for unverifiable
3888 relay addresses has also been removed.
3890 56. Added hosts_avoid_esmtp to the smtp transport.
3892 57. The exicyclog script was not checking for the esoteric option
3893 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3894 will work only if exicyclog is run under the appropriate euid.
3896 58. Following a discussion on the list, the rules by which Exim recognises line
3897 endings on incoming messages have been changed. The -dropcr and drop_cr
3898 options are now no-ops, retained only for backwards compatibility. The
3899 following line terminators are recognized: LF CRLF CR. However, special
3900 processing applies to CR:
3902 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3903 nor a local message in the state where . is a terminator.
3905 (ii) If a bare CR is encountered in a header line, an extra space is added
3906 after the line terminator so as not to end the header. The reasoning
3907 behind this is that bare CRs in header lines are most likely either
3908 to be mistakes, or people trying to play silly games.
3910 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3911 was being incorrectly given as 18 bytes larger than it should have been.
3912 This is a VOB (very old bug).
3914 60. This may never have affected anything current, but just in case it has:
3915 When the local host is found other than at the start of a list of hosts,
3916 the local host, those with the same MX, and any that follow, are discarded.
3917 When the list in question was part of a longer list of hosts, the following
3918 hosts (not currently being processed) were also being discarded. This no
3919 longer happens. I'm not sure if this situation could ever has previously
3922 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3925 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3926 header. This is something that is recommended in a new Internet Draft, and
3927 is something that is documented as being done by Sendmail. There are two
3928 possible values. For messages generated by the autoreply transport, Exim
3931 Auto-Submitted: auto-replied
3933 whereas for all other generated messages (e.g. bounces) it adds
3935 Auto-Submitted: auto-generated
3937 63. The "personal" condition in filters now includes a test for the
3938 Auto-Submitted: header. If it contains the string "auto-" the message it
3939 not considered personal.
3941 64. Added rcpt_include_affixes as a generic transport option.
3943 65. Added queue_only_override (default true).
3945 66. Added the syslog_duplication option.
3947 67. If what should have been the first header line of a message consisted of
3948 a space followed by a colon, Exim was mis-interpreting it as a header line.
3949 It isn't of course - it is syntactically invalid and should therefore be
3950 treated as the start of the message body. The misbehaviour could have
3951 caused a number of strange effects, including loss of data in subsequent
3952 header lines, and spool format errors.
3954 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3955 client host had authenticated. This control can now be exercised by an ACL
3956 for more flexibility.
3958 69. By default, callouts do not happen when testing with -bh. There is now a
3959 variant, -bhc, which does actually run the callout code, including
3960 consulting and updating the callout cache.
3962 70. Added support for saslauthd authentication, courtesy of Alexander
3965 71. If statvfs() failed on the spool or log directories while checking their
3966 size for availability, Exim confusingly gave the error "space shortage".
3967 Furthermore, in debugging mode it crashed with a floating point exception.
3968 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3969 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3970 serious problem, Exim now writes to the main and panic logs when this
3971 happens, with details of the failure. It then refuses to accept the
3972 incoming message, giving the message "spool directory problem" or "log
3973 directory problem" with a 421 code for SMTP messages.
3975 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3976 0, 1, and 2 exist, because some OS complain for execs without them (see
3977 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3978 descriptors. However, the code omitted to check that the open succeeded,
3979 causing mysterious errors if for some reason the permissions on /dev/null
3980 got screwed. Now Exim writes a message to the main and panic logs, and
3981 bombs out if it can't open /dev/null.
3983 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3984 interact so that it is all more flexible. It is supposed to remain
3985 backwards compatible. Also added extra_local_interfaces.
3987 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3988 to bomb out with an assertion failure - to the client this appears as a
3989 connection drop. This problem occurs in the part of the code that was taken
3990 from the Samba project. Fortunately, the assertion is in a very simple
3991 function, so I have fixed this by reproducing the function inline in the
3992 one place where it is called, and arranging for authentication to fail
3993 instead of killing the process with assert().
3995 75. The SPA client code was not working when the server requested OEM rather
3996 than Unicode encoding.
3998 76. Added code to make require_files with a specific uid setting more usable in
3999 the case where statting the file as root fails - usually a non-root-mounted
4000 NFS file system. When this happens and the failure is EACCES, Exim now
4001 forks a subprocess and does the per-uid checking as the relevant uid.
4003 77. Added process_log_path.
4005 78. If log_file_path was not explicitly set, a setting of check_log_space or
4006 check_log_inodes was ignored.
4008 79. If a space check for the spool or log partitions fails, the incident is now
4009 logged. Of course, in the latter case the data may get lost...
4011 80. Added the %p formatting code to string_format() so that it can be used to
4012 print addresses in debug_print(). Adjusted all the address printing in the
4013 debugging in store.c to use %p rather than %d.
4015 81. There was a concern that a line of code in smtp_in.c could overflow a
4016 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4017 initially expressed, the concern was not well-founded, because trailing
4018 spaces are removed early. However, if the trailing spaces were followed by
4019 a NULL, they did not get removed, so the overflow was possible. Two fixes
4022 (a) I re-wrote the offending code in a cleaner fashion.
4023 (b) If an incoming SMTP command contains a NULL character, it is rejected
4026 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4027 runs initgroups(), so that if the Exim user is in any additional groups,
4028 they will be used during message reception.
4034 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob