1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.373 2006/07/13 13:53:32 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 SC/01 Use a glob alias rather than an array ref in eximstats generated
10 parser. This improves both readability and performance.
12 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
13 Don't display local sender or destination tables in eximstats unless
14 there is data to show.
15 Added average volumes into the eximstats top table text output.
17 SC/03 Collect data on the number of addresses (recipients) as well
18 as the number of messages in eximstats.
20 TF/01 Correct an error in the documentation for the redirect router. Exim
21 does (usually) call initgroups() when daemonizing.
23 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
24 with consistent privilege compared to when running as a daemon.
26 TF/03 Note in the spec that $authenticated_id is not set for local
27 submissions from trusted users.
29 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
30 Thanks to Dean Brooks <dean@iglou.com> for the patch.
32 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
33 by adding some example configuration directives to the default
34 configuration file. A little bit of work is required to uncomment the
35 directives and define how usernames and passwords are checked, but
36 there is now a framework to start from.
38 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
39 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
40 without this. I don't know how relevant this is to other LDAP libraries.
42 PH/02 Add the verb name to the "unknown ACL verb" error.
44 PH/03 Magnus Holmgren's patch for filter_prepend_home.
46 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
48 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
49 directory not expanded when it should be if an expanded home directory
50 was set for the address (which is overridden by the transport).
52 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
55 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
56 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
57 because it is too late at that time, and has no effect.
59 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
60 security issue with \' (bugzilla #107). I could not use the
61 PQescapeStringConn() function, because it needs a PGconn value as one of
64 PH/08 When testing addresses using -bt, indicate those final addresses that
65 are duplicates that would not cause an additional delivery. At least one
66 person was confused, thinking that -bt output corresponded to deliveries.
67 (Suppressing duplicates isn't a good idea as you lose the information
68 about possibly different redirections that led to the duplicates.)
70 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
71 systems where poll() doesn't work, in particular OS X.
73 PH/10 Added more information to debugging output for retry time not reached.
75 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
76 operations in malware.c.
78 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
81 PH/13 If write_rejectlog was set false when logging was sent to syslog with
82 syslog_duplication set false, log lines that would normally be written
83 both the the main log and to the reject log were not written to syslog at
86 PH/14 In the default configuration, change the use of "message" in ACL warn
87 statements to "add_header".
89 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
90 not followed by a command (e.g. "seen endif").
92 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
93 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
100 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
101 other effects) broke the use of negated acl sub-conditions.
103 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
106 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
107 "Deny" causes Exim to reject the incoming connection with a 554 error.
108 Unfortunately, if there is a major crisis, such as a disk failure,
109 tcp-wrappers gives "deny", whereas what one would like would be some
110 kind of temporary error. A kludge has been added to help with this.
111 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
112 554 error is used if errno is still zero or contains ENOENT (which occurs
113 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
116 PH/03 Add -lutil to the default FreeBSD LIBS setting.
118 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
119 errors. Otherwise a message that provokes a temporary error (when other
120 messages do not) can cause a whole host to time out.
122 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
123 addresses were routed directly to files or pipes from a redirect router.
124 File deliveries just didn't batch; pipe deliveries might have suffered
127 PH/06 A failure to get a lock for a hints database would erroneously always say
128 "Failed to get write lock", even when it was really a read lock.
130 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
131 of 0600. This has been changed to use the value of the lockfile_mode
132 option (which defaults to 0600).
134 PH/08 Applied small patch from the Sieve maintainer.
136 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
137 folder from quota calculations, a direct delivery into this folder messed
138 up the contents of the maildirsize file. This was because the regex was
139 used only to exclude .Trash (or whatever) when the size of the mailbox
140 was calculated. There was no check that a delivery was happening into an
141 excluded directory. This bug has been fixed by ignoring all quota
142 processing for deliveries into excluded directories.
144 PH/10 Added the maildirfolder_create_regex option to appendfile.
150 PH/01 The code for finding all the local interface addresses on a FreeBSD
151 system running IPv6 was broken. This may well have applied to all BSD
152 systems, as well as to others that have similar system calls. The broken
153 code found IPv4 interfaces correctly, but gave incorrect values for the
154 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
155 that it would not match correctly against @[] and not recognize the IPv6
158 PH/02 The ipliteral router was not recognizing addresses of the form user@
159 [ipv6:....] because it didn't know about the "ipv6:" prefix.
161 PH/03 Added disable_ipv6.
163 PH/04 Changed $reply_address to use the raw form of the headers instead of the
164 decoded form, because it is most often used to construct To: headers
165 lines in autoreplies, and the decoded form may well be syntactically
166 invalid. However, $reply_address has leading white space removed, and all
167 newlines turned into spaces so that the autoreply transport does not
170 PH/05 If group was specified without a user on a router, and no group or user
171 was specified on a transport, the group from the router was ignored.
173 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
174 for visible compile-time settings that can be used to change these
175 numbers, for those that want even more. Backwards compatibility with old
176 spool files has been maintained. However, going back to a previous Exim
177 release will lost any variables that are in spool files.
179 PH/07 Two small changes when running in the test harness: increase delay when
180 passing a TCP/IP connection to a new process, in case the original
181 process has to generate a bounce, and remove special handling of
182 127.0.0.2 (sic), which is no longer necessary.
184 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
185 be the same on different OS.
187 PH/09 Moved a debug statement in filter processing to avoid a race problem when
190 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
191 whether --show-vars was specified or not
193 JJ/02 exipick: Added support for new ACL variable spool format introduced
196 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
197 syntactically invalid From: or Reply-to: line, and a filter used this to
198 generate an autoreply, and therefore failed to obtain an address for the
199 autoreply, Exim could try to deliver to a non-existent relative file
200 name, causing unrelated and misleading errors. What now happens is that
201 it logs this as a hard delivery error, but does not attempt to create a
204 PH/11 The exinext utility has a -C option for testing purposes, but although
205 the given file was scanned by exinext itself; it wasn't being passed on
206 when Exim was called.
208 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
209 an end-of-file indication when reading a command response.
211 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
212 compiled. In many other places in Exim, IPv6 addresses are always
213 recognized, so I have changed this. It also means that IPv4 domain
214 literals of the form [IPV4:n.n.n.n] are now always recognized.
216 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
217 used if the router is not running as root, for example, when verifying at
218 ACL time, or when using -bh. The debugging output from this situation was
219 non-existent - all you got was a failure to exec. I have made two
222 (a) Failures to set uid/gid, the current directory, or a process leader
223 in a subprocess such as that created by queryprogram now generate
224 suitable debugging ouput when -d is set.
226 (b) The queryprogram router detects when it is not running as root,
227 outputs suitable debugging information if -d is set, and then runs
228 the subprocess without attempting to change uid/gid.
230 PH/15 Minor change to Makefile for building test_host (undocumented testing
233 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
234 additional section of a DNS packet that returns MX or SRV records.
235 Instead, it always explicitly searches for A/AAAA records. This avoids
236 major problems that occur when a DNS server includes only records of one
237 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
238 fixed another bug: if SRV records were looked up and the corresponding
239 address records were *not* found in the additional section, the port
240 values from the SRV records were lost.
242 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
243 using the correct key (the original address) when searching the retry
244 rules in order to find which one to use for generating the retry hint.
246 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
247 adding the default one. Similarly, if it contains a Reply-To: header, the
248 errors_reply_to option, if set, is not used.
250 PH/19 When calculating a retry time, Exim used to measure the "time since
251 failure" by looking at the "first failed" field in the retry record. Now
252 it does not use this if it is later than than the arrival time of the
253 message. Instead it uses the arrival time. This makes for better
254 behaviour in cases where some deliveries succeed, thus re-setting the
255 "first failed" field. An example is a quota failure for a huge message
256 when small messages continue to be delivered. Without this change, the
257 "time since failure" will always be short, possible causing more frequent
258 delivery attempts for the huge message than are intended.
259 [Note: This change was subsequently modified - see PH/04 for 4.62.]
261 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
262 $1, $2, $3) because the numerical variables can be reset during some
263 expansion items (e.g. "match"), thereby losing the authentication data.
265 PH/21 Make -bV show the size of off_t variables so that the test suite can
266 decide whether to run tests for quotas > 2G.
268 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
269 mailbox_size, and mailbox_filecount in the appendfile transport. If a
270 filecount value is greater than 2G or if a quota value is greater than 2G
271 on a system where the size of off_t is not greater than 4, a panic error
274 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
275 never match. The debug and -bh output now contains an explicit error
276 message indicating a malformed IPv4 address or mask.
278 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
279 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
282 PH/25 Do not write to syslog when running in the test harness. The only
283 occasion when this arises is a failure to open the main or panic logs
284 (for which there is an explicit test).
286 PH/26 Added the /no_tell option to "control=freeze".
288 PH/27 If a host name lookup failed very early in a connection, for example, if
289 the IP address matched host_lookup and the reverse lookup yielded a name
290 that did not have a forward lookup, an error message of the form "no IP
291 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
292 could be logged. Now it outputs the IP address instead of "NULL".
294 PH/28 An enabling patch from MH: add new function child_open_exim2() which
295 allows the sender and the authenticated sender to be set when
296 submitting a message from within Exim. Since child_open_exim() is
297 documented for local_scan(), the new function should be too.
299 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
300 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
301 results in an empty string is now treated as unset.
303 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
305 PH/31 Added sender_verify_fail logging option.
307 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
308 needed by RFC 822 but not by RFC 2822 was commented out. I have now
309 tidied the source and removed it altogether.
311 PH/33 When a queue run was abandoned because the load average was too high, a
312 log line was always written; now it is written only if the queue_run log
313 selector is set. In addition, the log line for abandonment now contains
314 information about the queue run such as the pid. This is always present
315 in "start" and "stop" lines but was omitted from the "abandon" line.
317 PH/34 Omit spaces between a header name and the colon in the error message that
318 is given when verify = headers_syntax fails (if there are lots of them,
319 the message gets confusing).
321 PH/35 Change the default for dns_check_names_pattern to allow slashes within
322 names, as there are now some PTR records that contain slashes. This check
323 is only to protect against broken name servers that fall over on strange
324 characters, so the fact that it applies to all lookups doesn't matter.
326 PH/36 Now that the new test suite is complete, we can remove some of the
327 special code in Exim that was needed for the old test suite. For example,
328 sorting DNS records because real resolvers return them in an arbitrary
329 order. The new test suite's fake resolver always returns records in the
332 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
333 bounces) except when queue_only is set, to avoid logging races between
334 the different processes.
336 PH/38 Panic-die if .include specifies a non-absolute path.
338 PH/39 A tweak to the "H" retry rule from its user.
340 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
341 a label. They prevented compilation on older perls.
343 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
344 a warning to be raised on newish perls.
346 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
347 on queue. Changes to match documented behaviour of showing count of
348 messages matching specified criteria.
350 PH/40 Changed the default ident timeout from 30s to 5s.
352 PH/41 Added support for the use of login_cap features, on those BSD systems
353 that have them, for controlling the resources used by pipe deliveries.
355 PH/42 The content-scanning code uses fopen() to create files in which to put
356 message data. Previously it was not paying any attention to the mode of
357 the files. Exim runs with umask(0) because the rest of the code creates
358 files with open(), and sets the required mode explicitly. Thus, these
359 files were ending up world-writeable. This was not a big issue, because,
360 being within the spool directory, they were not world-accessible. I have
361 created a function called modefopen, which takes an additional mode
362 argument. It sets umask(777), creates the file, chmods it to the required
363 mode, then resets the umask. All the relevant calls to fopen() in the
364 content scanning code have been changed to use this function.
366 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
367 to 24 hours. This avoids potential overflow problems when processing G
368 and H retry rules. I suspect nobody ever tinkers with this value.
370 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
372 PH/45 When the plaintext authenticator is running as a client, the server's
373 challenges are checked to ensure they are valid base64 strings. By
374 default, the authentication attempt is cancelled if an invalid string is
375 received. Setting client_ignore_invalid_base64 true ignores these errors.
376 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
377 they are received. Thus, the responses can be made to depend on the
378 challenges. If an invalid string is ignored, an empty string is placed in
381 PH/46 Messages that are created by the autoreply transport now contains a
382 References: header, in accordance with RFCs 2822 and 3834.
384 PH/47 Added authenticated_sender_force to the smtp transport.
386 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
388 PH/49 Installed latest patch from the Sieve maintainer.
390 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
391 was also set, the appendfile transport was unnecessarily scanning a
392 directory of message files (e.g. for maildir delivery) to find the count
393 of files (along with the size), even though it did not need this
394 information. It now does the scan only if it needs to find either the
395 size of the count of files.
397 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
399 PH/52 Two bugs concerned with error handling when the smtp transport is
402 (i) Exim was not creating retry information for temporary errors given
403 for individual recipients after the DATA command when the smtp transport
404 was used in LMTP mode. This meant that they could be retried too
405 frequently, and not timed out correctly.
407 (ii) Exim was setting the flag that allows error details to be returned
408 for LMTP errors on RCPT commands, but not for LMTP errors for individual
409 recipients that were returned after the DATA command.
411 PH/53 This is related to PH/52, but is more general: for any failing address,
412 when detailed error information was permitted to be returned to the
413 sender, but the error was temporary, then after the final timeout, only
414 "retry timeout exceeded" was returned. Now it returns the full error as
415 well as "retry timeout exceeded".
417 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
418 do this, and (what is worse) MTAs that accept it.
420 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
421 will now be deprecated.
423 PH/56 New os.c-cygwin from the Cygwin maintainer.
425 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
426 formats (previously only available in exim formats via -bpr, -bpru,
427 and -bpra. Now also available in native and exiqgrep formats)
429 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
430 with very large, slow to parse queues
432 JJ/08 exipick: added ! as generic prefix to negate any criteria format
434 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
436 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
437 responses to authentication challenges, though it was showing the
438 challenges; (ii) I've removed the CR characters from the debug output for
441 PH/58 Allow for the insertion of a newline as well as a space when a string
442 is turned into more than one encoded-word during RFC 2047 encoding. The
443 Sieve code now uses this.
445 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
446 data_4xx, lost_connection, tls_required.
448 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
449 message was being sent as an SMTP response.
451 PH/61 Add -l and -k options to exicyclog.
453 PH/62 When verifying, if an address was redirected to one new address, so that
454 verification continued, and the new address failed or deferred after
455 having set something in $address_data, the value of $address_data was not
456 passed back to the ACL. This was different to the case when no
457 redirection occurred. The value is now passed back in both cases.
459 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
460 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
461 use login_cap.h, so on its own it isn't the distinguishing feature. The
462 new name refers directly to the setclassresources() function.
464 PH/65 Added configuration files for NetBSD3.
466 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
468 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
469 is preferred over IPv4.
471 PH/68 The bounce_return_message and bounce_return_body options were not being
472 honoured for bounces generated during the reception of non-SMTP messages.
473 In particular, this applied to messages rejected by the ACL. This bug has
474 been fixed. However, if bounce_return_message is true and bounce_return_
475 body is false, the headers that are returned for a non-SMTP message
476 include only those that have been read before the error was detected.
477 (In the case of an ACL rejection, they have all been read.)
479 PH/69 The HTML version of the specification is now built in a directory called
480 spec_html instead of spec.html, because the latter looks like a path with
481 a MIME-type, and this confuses some software.
483 PH/70 Catch two compiler warnings in sieve.c.
485 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
486 function verify_get_ident() calls ip_connect() to connect a socket, but
487 if the "connect()" function timed out, ip_connect() used to close the
488 socket. However, verify_get_ident() also closes the socket later, and in
489 between Exim writes to the log, which may get opened at this point. When
490 the socket was closed in ip_connect(), the log could get the same file
491 descriptor number as the socket. This naturally causes chaos. The fix is
492 not to close the socket in ip_connect(); the socket should be closed by
493 the function that creates it. There was only one place in the code where
494 this was missing, in the iplookup router, which I don't think anybody now
495 uses, but I've fixed it anyway.
497 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
498 well as to direct DNS lookups. Otherwise the handling of names in host
499 lists is inconsistent and therefore confusing.
505 PH/01 Two changes to the default runtime configuration:
507 (1) Move the checks for relay_from_hosts and authenticated clients from
508 after to before the (commented out) DNS black list checks.
510 (2) Add control=submission to the relay_from_hosts and authenticated
511 clients checks, on the grounds that messages accepted by these
512 statements are most likely to be submissions.
514 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
516 (1) Generate an error if the third argument for the ${prvs expansion is
519 (2) Treat a missing third argument of ${prvscheck as if it were an empty
522 (3) Reset the variables that are obtained from the first argument of
523 ${prvscheck and used in the second argument before leaving the code,
524 because their memory is reclaimed, so using them afterwards may do
527 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
528 one (it's much easier than Tom thought :-).
530 (5) Because of (4), we can now allow for the use of $prvscheck_result
531 inside the third argument.
533 PH/03 For some reason, the default setting of PATH when running a command from
534 a pipe transport was just "/usr/bin". I have changed it to
537 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
538 anything to be listed in the output from -bV.
540 PH/05 When a filter generated an autoreply, the entire To: header line was
541 quoted in the delivery log line, like this:
543 => >A.N.Other <ano@some.domain> <original@ddress> ...
545 This has been changed so that it extracts the operative address. There
546 may be more than one such address. If so, they are comma-separated, like
549 => >ano@some.domain,ona@other.domain <original@ddress> ...
551 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
552 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
553 not being looked up in the rDNS to get a host name, Exim was showing the
554 IP address twice in Received: lines, even though the IP addresses were
555 identical. For example:
557 Received: from [1.2.3.4] (helo=[1.2.3.4])
559 However, if the real host name was known, it was omitting the HELO data
560 if it matched the actual IP address. This has been tidied up so that it
561 doesn't show the same IP address twice.
563 PH/07 When both +timestamp and +memory debugging was on, the value given by
564 $tod_xxx expansions could be wrong, because the tod_stamp() function was
565 called by the debug printing, thereby overwriting the timestamp buffer.
566 Debugging no longer uses the tod_stamp() function when +timestamp is set.
568 PH/08 When the original message was included in an autoreply transport, it
569 always said "this is a copy of the message, including all the headers",
570 even if body_only or headers_only was set. It now gives an appropriate
573 PH/09 Applied a patch from the Sieve maintainer which:
575 o fixes some comments
576 o adds the (disabled) notify extension core
577 o adds some debug output for the result of if/elsif tests
578 o points to the current vacation draft in the documentation
579 and documents the missing references header update
583 o fixes a bug in processing the envelope test (when testing
584 multiple envelope elements, the last element determinted the
587 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
588 Electronic Mail") by including:
590 Auto-submitted: auto-generated
592 in the messages that it generates (bounce messages and others, such as
593 warnings). In the case of bounce messages for non-SMTP mesages, there was
594 also a typo: it was using "Auto_submitted" (underscore instead of
595 hyphen). Since every message generated by Exim is necessarily in response
596 to another message, thes have all been changed to:
598 Auto-Submitted: auto-replied
600 in accordance with these statements in the RFC:
602 The auto-replied keyword:
604 - SHOULD be used on messages sent in direct response to another
605 message by an automatic process,
607 - MUST NOT be used on manually-generated messages,
609 - MAY be used on Delivery Status Notifications (DSNs) and Message
610 Disposition Notifications (MDNs),
612 - MUST NOT be used on messages generated by automatic or periodic
613 processes, except for messages which are automatic responses to
616 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
617 to the default Received: header definition.
619 PH/12 Added log selector acl_warn_skipped (default on).
621 PH/13 After a successful wildlsearch lookup, discard the values of numeric
622 variables because (a) they are in the wrong storage pool and (b) even if
623 they were copied, it wouldn't work properly because of the caching.
625 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
626 checking when decoding. Apparently there are clients that generate
627 overlong encoded strings. Why am I not surprised?
629 PH/15 If the first argument of "${if match_address" was not empty, but did not
630 contain an "@" character, Exim crashed. Now it writes a panic log message
631 and treats the condition as false.
633 PH/16 In autoreply, treat an empty string for "once" the same as unset.
635 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
636 extension "envelope-auth". The code is finished and in agreement with
637 other implementations, but there is no documentation so far and in fact,
638 nobody wrote the draft yet. This extension is currently #undef'ed, thus
639 not changing the active code.
641 Print executed "if" and "elsif" statements when debugging is used. This
642 helps a great deal to understand what a filter does.
644 Document more things not specified clearly in RFC3028. I had all this
645 sorted out, when out of a sudden new issues came to my mind. Oops."
647 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
650 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
651 canonical form (as documented). However, after a host name lookup from
652 the IP address, check_host() was doing a simple string comparison with
653 addresses acquired from the DNS when checking that the found name did
654 have the original IP as one of its addresses. Since any found IPv6
655 addresses are likely to be in abbreviated form, the comparison could
656 fail. Luckily, there already exists a function for doing the comparison
657 by converting both addresses to binary, so now that is used instead of
660 PH/20 There was another similar case to PH/19, when a complete host name was
661 given in a host list; looking up its IP address could give an abbreviated
662 form, whereas the current host's name might or might not be abbreviated.
663 The same fix has been applied.
669 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
670 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
673 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
674 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
676 PH/03 Typo: missing ".o" in src/pcre/Makefile.
678 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
679 header line, restrict the check to what is listed in RFCs 2369 and 2929.
680 Also, for "Auto-Submitted", treat anything other than "no" as
681 non-personal, in accordance with RFC 3834. (Previously it treated
682 anything starting "auto-" as non-personal.)
684 TF/01 The control=submission/name=... option had a problem with syntax
685 errors if the name included a slash character. The /name= option
686 now slurps the rest of the string, so it can include any characters
687 but it must come last in the list of options (after /sender_retain
690 PH/05 Some modifications to the interface to the fake nameserver for the new
698 TK/01 Added the "success_on_redirect" address verification option. See
699 NewStuff for rationale and an example.
701 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
703 PH/02 Patch to exigrep to allow it to work on syslog lines.
705 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
706 fread() to skip over the body file's header line, because in Cygwin the
707 header line is locked and is inaccessible.
709 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
710 co-exist for some time) to make it clear that it is the Exim ID that is
711 referenced, not the Message-ID: header line.
713 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
714 string_format() function, because snprintf() does not exist on all
717 PH/06 The use of forbid_filter_existstest now also locks out the use of the
718 ${stat: expansion item.
720 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
721 protocol synchronization error", to keep the pedants happy.
723 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
724 well as for IRIX systems, when gcc is being used. See the host.c source
727 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
729 PH/10 Named domain lists were not working if used in a queue_smtp_domains
732 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
733 transport and to the smtp transport in LMTP mode.
735 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
737 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
738 run a filter in a subprocess. This could lead to confusion in subsequent
739 lookups in the parent process. There should also be a search_tidyup() at
740 the end of the subprocess.
742 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
743 only if the host matched helo_try_verify_hosts, which caused the
744 verification to occur when the EHLO/HELO command was issued. The ACL just
745 tested the remembered result. Now, if a previous verification attempt has
746 not happened, "verify = helo" does it there and then.
748 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
750 TK/03 Fix log output including CR from clamd.
752 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
753 exist provoked a memory error which could cause a segfault.
755 PH/15 Installed PCRE 6.2
757 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
759 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
760 of the problem. Specifically, suggested +O2 rather than +O1 for the
763 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
765 PH/20 If a delivery was routed to a non-standard port by means of an SRV
766 record, the port was not correctly logged when the outgoing_port log
767 selector was set (it logged the transort's default port).
769 PH/21 Added support for host-specific ports to manualroute, queryprogram,
770 fallback_hosts, and "hosts" in the smtp transport.
772 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
773 host errors such as "Connection refused".
775 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
776 authentication with radiusclient 0.4.9:
778 - Error returned from rc_read_config was caught wrongly
779 - Username/password not passed on to radius server due to wrong length.
781 The presumption is that some radiusclient API changes for 4.51/PH/17
782 were not taken care of correctly. The code is still untested by me (my
783 Linux distribution still has 0.3.2 of radiusclient), but it was
784 contributed by a Radius user.
786 PH/24 When doing a callout, the value of $domain wasn't set correctly when
787 expanding the "port" option of the smtp transport.
789 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
790 while reading a MIME header. Thanks to Tom Hughes for a patch.
792 PH/24 Include config.h inside local_scan.h so that configuration settings are
795 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
796 that in an ACL for RCPT (for example), you can examine exactly what was
799 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
800 commands, but it was not correctly comparing the address with the actual
801 client host address. Thus, it would show the EHLO address in Received:
802 header lines when this was not necessary.
804 PH/27 Added the % operator to ${eval:}.
806 PH/28 Exim tries to create and chdir to its spool directory when it starts;
807 it should be ignoring failures (because with -C, for example, it has lost
808 privilege). It wasn't ignoring creation failures other than "already
811 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
814 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
815 because an input line was too long, either on its own, or by virtue of
816 too many continuations, the temporary file was not being removed, and the
817 return code was incorrect.
819 PH/31 Missing "BOOL" in function definition in filtertest.c.
821 PH/32 Applied Sieve patches from the maintainer.
823 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
825 PH/33 Added "verify = not_blind".
827 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
828 Local/Makefile (with some defaults set). These are used in built scripts
829 such as exicyclog, but they have never been used in the exim_install
830 script (though there are many overriding facilities there). I have
831 arranged that the exim_install script now takes note of these two
834 PH/35 Installed configuration files for Dragonfly.
836 PH/36 When a locally submitted message by a trusted user did not contain a
837 From: header, and the sender address was obtained from -f or from an SMTP
838 MAIL command, and the trusted user did not use -F to supply a sender
839 name, $originator_name was incorrectly used when constructing a From:
840 header. Furthermore, $originator_name was used for submission mode
841 messages from external hosts without From: headers in a similar way,
842 which is clearly wrong.
844 PH/37 Added control=suppress_local_fixups.
846 PH/38 When log_selector = +received_sender was set, and the addition of the
847 sender made the log line's construction buffer exactly full, or one byte
848 less than full, an overflow happened when the terminating "\n" was
851 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
852 when the result of a list match is failure because a DNS lookup failed.
854 PH/40 RM_COMMAND is now used in the building process.
856 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
857 the "build-* directories that it finds.
859 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
860 address in a domain literal was a prefix of an interface address.
862 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
863 when verifying a sender address, unless rewrite_headers is false.
865 PH/44 Wrote a long comment about why errors_to addresses are verified as
866 recipients, not senders.
868 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
869 the ratelimit ACL was added.
871 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
873 PH/46 Added extra information about PostgreSQL errors to the error string.
875 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
876 avoiding the need to install special zones in a real server. This is
877 backwards compatible; if it can't find the fake resolver, it drops back.
878 Thus, both old and new test suites can be run.
880 TF/02 Added util/ratelimit.pl
882 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
883 clock is set back in time.
885 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
886 Candler <B.Candler@pobox.com>.
888 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
889 OK for addresses that are the result of redirections.
891 PH/48 A number of further additions for the benefit of the new test suite,
892 including a fake gethostbyname() that interfaces to the fake DNS resolver
895 TF/06 The fix for widen_domains has also been applied to qualify_single and
896 search_parents which are the other dnslookup options that can cause
899 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
902 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
904 TF/07 Exim produced the error message "an SRV record indicated no SMTP
905 service" if it encountered an MX record with an empty target hostname.
906 The message is now "an MX or SRV record indicated no SMTP service".
908 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
909 if the DNS of the sending site is misconfigured. This is quite a
910 common situation. This change restores the behaviour of treating a
911 helo verification defer as a failure.
913 PH/51 If self=fail was set on a router, the bounce message did not include the
914 actual error message.
920 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
922 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
923 command itself ended in error, the underlying message about the transport
924 filter timeout was being overwritten with the pipe command error. Now the
925 underlying error message should be appended to the second error message.
927 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
929 PH/02 Reduce the amount of output that "make" produces by default. Full output
930 can still be requested.
932 PH/03 The warning log line about a condition test deferring for a "warn" verb
933 was being output only once per connection, rather than after each
934 occurrence (because it was using the same function as for successful
935 "warn" verbs). This seems wrong, so I have changed it.
937 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
938 it should not have, which might have caused a crash in the right
939 circumstances, but probably never did.
941 PH/04 Installed a modified version of Tony Finch's patch to make submission
942 mode fix the return path as well as the Sender: header line, and to
943 add a /name= option so that you can make the user's friendly name appear
946 TF/03 Added the control = fakedefer ACL modifier.
948 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
949 Mark Lowes for thorough testing.
951 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
953 TK/03 Merged latest SRS patch from Miles Wilton.
955 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
956 with the definition in sysexits.h (which is #included earlier).
957 Fortunately, Exim does not actually use EX_OK. The code used to try to
958 preserve the sysexits.h value, by assumimg that macro definitions were
959 scanned for macro replacements. I have been disabused of this notion,
960 so now the code just undefines EX_OK before #including unistd.h.
962 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
963 in the smtp transport. When a block could not be written in a single
964 write() function, the timeout was being re-applied to each part-write.
965 This seems wrong - if the receiver was accepting one byte at a time it
966 would take for ever. The timeout is now adjusted when this happens. It
967 doesn't have to be particularly precise.
969 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
970 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
972 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
973 without a domain if the check to <postmaster@domain> fails.
975 SC/01 Eximstats: added -xls and the ability to specify output files
976 (patch written by Frank Heydlauf).
978 SC/02 Eximstats: use FileHandles for outputing results.
980 SC/03 Eximstats: allow any combination of xls, txt, and html output.
982 SC/04 Eximstats: fixed display of large numbers with -nvr option
984 SC/05 Eximstats: fixed merging of reports with empty tables.
986 SC/06 Eximstats: added the -include_original_destination flag
988 SC/07 Eximstats: removed tabs and trailing whitespace.
990 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
992 TK/06 MBOX spool code: Add real "From " MBOX separator line
993 so the .eml file is really in mbox format (even though
994 most programs do not really care). Patch from Alex Miller.
996 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
997 The latter is generated from $received_to and is only set if the
998 message has one envelope recipient. SA can use these headers,
999 obviously out-of-the-box. Patch from Alex Miller.
1001 PH/08 The ${def test on a variable was returning false if the variable's
1002 value was "0", contrary to what the specification has always said!
1003 The result should be true unless the variable is empty.
1005 PH/09 The syntax error of a character other than { following "${if
1006 def:variable_name" (after optional whitespace) was not being diagnosed.
1007 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1008 accidental colon was present, for example, could give incorrect results.
1010 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1011 result is used (not including appendfile, where other changes are about
1014 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1015 This involved changing a lot of size variables from int to off_t. It
1016 should work with maildirs and everything.
1018 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1019 spamd dying while we are connected to it.
1021 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1024 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1025 function for BDB 4.3.
1027 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1029 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1030 the sources for PCRE have been re-arranged and more files are now
1033 PH/15 The code I had for printing potentially long long variables in PH/11
1034 above was not the best (it lost precision). The length of off_t variables
1035 is now inspected at build time, and an appropriate printing format (%ld
1036 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1037 to be "long long int" or "long int". This is needed for the internal
1038 formatting function string_vformat().
1040 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1041 the configuration file to be ":syslog", then the script "guesses" where
1042 the logs files are, rather than using the compiled in default. In our
1043 case the guess is not the same as the compiled default, so the script
1044 suddenly stopped working when I started to use syslog. The patch checks
1045 to see if log_file_path is "". If so, it attempts to read it from exim
1046 with no configuration file to get the compiled in version, before it
1047 falls back to the previous guessing code."
1049 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1050 implementing BATV in an Exim configuration. See NewStuff for the gory
1053 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1054 Makefile that are specific to HP-UX.
1056 PH/18 If the "use_postmaster" option was set for a recipient callout together
1057 with the "random" option, the postmaster address was used as the MAIL
1058 FROM address for the random test, but not for the subsequent recipient
1059 test. It is now used for both.
1061 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1062 patch removes a few documentation additions to RFC 3028, because the
1063 latest draft now contains them. It adds the new en;ascii-case comparator
1064 and a new error check for 8bit text in MIME parts. Comparator and
1065 require names are now matched exactly. I enabled the subaddress
1066 extension, but it is not well tested yet (read: it works for me)."
1068 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1069 rework some of the code of TK/09 above to avoid the hardwired use of
1070 "%lld" and "long long". Replaced the call to snprintf() with a call to
1073 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1074 records point to non-existent hosts", "retry timeout exceeded", and
1075 "retry time not reached for any host after a long failure period".
1077 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1078 experimental DomainKeys support:
1080 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1081 (2) On an error such as an illegally used "control", the wrong name for
1082 the control was given.
1084 These problems did NOT occur unless DomainKeys support was compiled.
1086 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1088 PH/24 Added ${if match_ip condition.
1090 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1091 databases so that it will be absolutely obvious if a crash occurs in the
1092 DB library. This is a regular occurrence (often caused by mis-matched
1095 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1096 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1097 picked up on a user's system that detects such things. There doesn't seem
1098 to be a gcc warning option for this - only an attribute that has to be
1099 put on the function's prototype. It seems that in Fedora Core 4 they have
1100 set this on a number of new functions. No doubt there will be more in due
1103 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1104 specify a transport. However, if an address that was verified by such a
1105 router was the subject of a callout, Exim crashed because it tried to
1106 read the rcpt_include_affixes from the non-existent transport. Now it
1107 just assumes that the setting of that option is false. This bug was
1108 introduced by 4.51/PH/31.
1110 PH/28 Changed -d+all to exclude +memory, because that information is very
1111 rarely of interest, but it makes the output a lot bigger. People tend to
1112 do -d+all out of habit.
1114 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1115 code in os-type was giving problems when libc.so lives in lib64, like on
1118 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1119 aren't the modern standard, and it seems that some systems' include files
1120 don't always have them. Exim was already checking for some of the newer
1121 ones like T_AAAA, and defining it itself. I've added checks for all the
1122 record types that Exim uses.
1124 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1125 not automatically generating a new one, as it is supposed to. This
1126 prevented TLS from working. If the file did exist, but contained invalid
1127 data, a new version was generated, as expected. It was only the case of a
1128 non-existent file that was broken.
1130 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1131 with a change in libdomainkeys > 0.64.
1133 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1134 from DNS. If the selector record carries the flag, it now has
1135 precedence over the domain-wide flag.
1137 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1139 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1140 the use of a port name that isn't defined in /etc/services) occurred, the
1141 message was deferred as in a normal delivery, and thus remained on the
1142 spool, instead of being failed because of the mua_wrapper setting. This
1143 is now fixed, and I tidied up some of the mua_wrapper messages at the
1146 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1147 the messages in a hash of arrays rather than using individual hashes.
1148 This is a bit cleaner and results in dramatic memory savings, albeit
1149 at a slight CPU cost.
1151 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1152 as requested by Marc Sherman.
1154 SC/10 Eximstats: added histograms for user specified patterns as requested
1157 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1159 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1160 fopen() in the content-scanning modules that did not already have it.
1166 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1167 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1169 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1171 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1173 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1174 to test Sieve filters that use "vacation".
1176 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1177 that changes the way the GnuTLS parameters are stored in the cache file.
1178 The new format can be generated externally. For backward compatibility,
1179 if the data in the cache doesn't make sense, Exim assumes it has read an
1180 old-format file, and it generates new data and writes a new file. This
1181 means that you can't go back to an older release without removing the
1184 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1185 work if there are any delivery delays because "one_time" forces the
1186 parent to be marked "delivered", so its unseen clone is never tried
1187 again. For this reason, Exim now forbids the simultaneous setting of
1190 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1191 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1192 entry if you want to know the details. The fix, however, neglected to
1193 consider the case when local delivery batching is involved. The test for
1194 "previously delivered" was not happening when checking to see if an
1195 address could be batched with a previous (undelivered) one; under
1196 certain circumstances this could lead to multiple deliveries to the same
1199 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1200 in its include files, and this causes problems building Exim.
1202 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1203 header_syntax) but Exim was just ignoring anything given after a slash.
1204 In particular, this caused confusion with an attempt to use "verify =
1205 reverse_host_lookup/defer_ok". An error is now given when options are
1206 supplied for verify items that do not have them. (Maybe reverse_host_
1207 lookup should have a defer_ok option, but that's a different point.)
1209 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1210 defined by RFC 821) to 2048, because there were problems with some AUTH
1211 commands, and RFC 1869 says the size should be increased for extended
1212 SMTP commands that take arguments.
1214 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1217 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1218 "unknown" error; now it says that the functionality isn't in the binary.
1220 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1221 an address' error message when a string expansion fails (syntax or
1222 whatever). Otherwise the password may appear in the log. Following change
1223 PH/42 below, there is no longer a chance of it appearing in a bounce
1226 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1228 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1229 of its addresses was ever tried. (Bugzilla bug #2.)
1231 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1232 the result incorrectly in the debug output. (It correctly added a newline
1233 to what was transported.)
1235 TF/01 Added $received_time.
1237 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1238 commented out examples of how to interface to a virus scanner and to
1239 SpamAssassin. Also added commented examples of av_scanner and
1240 spamd_address settings.
1242 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1243 and controls are allowed in which ACLs. There were a couple of minor
1244 errors. Some of the entries in the conditions table (which is a table of
1245 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1246 negation of where the condition IS allowed.
1248 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1250 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1251 header file does not have a version number, so I've had to invent a new
1252 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1253 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1254 radiusclient), but it was contributed by a Radius user.
1256 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1257 files or directories, for OpenSSL.
1259 PH/19 When an Exim process that is running as root has to create an Exim log
1260 file, it does so in a subprocess that runs as exim:exim so as to get the
1261 ownership right at creation (otherwise, other Exim processes might see
1262 the file with the wrong ownership). There was no test for failure of this
1263 fork() call, which would lead to the process getting stuck as it waited
1264 for a non-existent subprocess. Forks do occasionally fail when resources
1265 run out. I reviewed all the other calls to fork(); they all seem to check
1268 PH/20 When checking for unexpected SMTP input at connect time (before writing
1269 the banner), Exim was not dealing correctly with a non-positive return
1270 from the read() function. If the client had disconnected by this time,
1271 the result was a log entry for a synchronization error with an empty
1272 string after "input=" when read() returned zero. If read() returned -1
1273 (an event I could not check), uninitialized data bytes were printed.
1274 There were reports of junk text (parts of files, etc) appearing after
1277 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1279 PH/22 Added support for macro redefinition, and (re)definition in between
1280 driver and ACL definitions.
1282 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1283 forgetting to use the resulting value; it was using the unexpanded value.
1285 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1286 hadn't been configured. The fix is from Juergen Kreileder, who
1287 understands it better than I do:
1289 "Here's what I see happening with three configured cyrus_sasl
1290 authenticators configured (plain, login, cram-md5):
1292 On startup auth_cyrus_sasl_init() gets called for each of these.
1293 This means three calls to sasl_listmech() without a specified mech_list.
1294 => SASL tests which mechs of all available mechs actually work
1295 => three warnings about OTP not working
1296 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1298 With the patch, sasl_listmech() also gets called three times. But now
1299 SASL's mech_list option is set to the server_mech specified in the the
1300 authenticator. Or in other words, the answer from sasl_listmech()
1301 gets limited to just the mech you're testing for (which is different
1303 => the return list contains just 'plain' or 'login', 'cram-md5' or
1304 nothing depending on the value of ob->server_mech.
1306 I've just tested the patch: Authentication still works fine,
1307 unavailable mechs specified in the exim configuration are still
1308 caught, and the auth.log warnings about OTP are gone."
1310 PH/25 When debugging is enabled, the contents of the command line are added
1311 to the debugging output, even when log_selector=+arguments is not
1314 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1315 answer is "GNU", and only if the return is "GNU/something" is the answer
1318 PH/27 $acl_verify_message is now set immediately after the failure of a
1319 verification in an ACL, and so is available in subsequent modifiers. In
1320 particular, the message can be preserved by coding like this:
1322 warn !verify = sender
1323 set acl_m0 = $acl_verify_message
1325 Previously, $acl_verify_message was set only while expanding "message"
1326 and "log_message" when a very denied access.
1328 PH/28 Modified OS/os.c-Linux with
1330 -#ifndef OS_LOAD_AVERAGE
1331 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1333 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1334 nomenclature these days.)
1336 PH/29 Installed patch from the Sieve maintainer that adds the options
1337 sieve_useraddress and sieve_subaddress to the redirect router.
1339 PH/30 In these circumstances:
1340 . Two addresses routed to the same list of hosts;
1341 . First host does not offer TLS;
1342 . First host accepts first address;
1343 . First host gives temporary error to second address;
1344 . Second host offers TLS and a TLS session is established;
1345 . Second host accepts second address.
1346 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1347 and peerdn, if requested) that were in fact used only for the second
1350 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1351 attention to any local part prefix or suffix that was matched by the
1352 router that accepted the address. It now behaves in the same way as it
1353 does for delivery: the affixes are removed from the local part unless
1354 rcpt_include_affixes is set on the transport.
1356 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1357 timeout during the DATA phase of an incoming message.
1359 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1360 applied a patch sanctioned by the Sieve maintainer.
1362 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1363 the uid or gid is negative. A case of a negative gid caused this to be
1364 noticed. The fix allows for either to be negative.
1366 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1367 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1370 PH/36 The change PH/12 above was broken. Fixed it.
1372 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1373 the grounds that routing the same address twice would always produce the
1374 same answer. This might have been true once, but it is certainly no
1375 longer true now. Routing a child address may depend on the previous
1376 routing that produced that child. Some complicated redirection strategies
1377 went wrong when messages had multiple recipients, and made Exim's
1378 behaviour dependent on the order in which the addresses were given.
1380 I have moved the duplicate checking until after the routing is complete.
1381 Exim scans the addresses that are assigned to local and remote
1382 transports, and removes any duplicates. This means that more work will be
1383 done, as duplicates will always all be routed, but duplicates are
1384 presumably rare, so I don't expect this is of any significance.
1386 For deliveries to pipes, files, and autoreplies, the duplicate checking
1387 still happens during the routing process, since they are not going to be
1390 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1391 It corrects a timeout issue with spamd. This is Ian's comment: "The
1392 background is that sometimes spamd either never reads data from a
1393 connection it has accepted, or it never writes response data. The exiscan
1394 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1395 blindly assumes that writes won't block so it may never time out."
1397 PH/39 Allow G after quota size as well as K and M.
1399 PH/40 The value set for $authenticated_id in an authenticator may not contain
1400 binary zeroes or newlines because the value is written to log lines and
1401 to spool files. There was no check on this. Now the value is run through
1402 the string_printing() function so that such characters are converted to
1403 printable escape sequences.
1405 PH/41 $message_linecount is a new variable that contains the total number of
1406 lines in the message. Compare $body_linecount, which is the count for the
1409 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1410 bounce and delay warning messages, except in certain special cases, which
1413 (a) An SMTP error message from a remote host;
1414 (b) A message specified in a :fail: redirection;
1415 (c) A message specified in a "fail" command in a system filter;
1416 (d) A message specified in a FAIL return from the queryprogram router;
1417 (e) A message specified by the cannot_route_message router option.
1419 In these cases only, Exim does include the error details in bounce and
1420 warning messages. There are also a few cases where bland messages such
1421 as "unrouteable address" or "local delivery error" are given.
1423 PH/43 $value is now also set for the "else" part of a ${run expansion.
1425 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1426 being worked on, but at least Exim now implements the latest version to
1429 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1430 process to complete was treated as a delivery failure, a timeout while
1431 writing the message to the pipe was logged, but erroneously treated as a
1432 successful delivery. Such timeouts include transport filter timeouts. For
1433 consistency with the overall process timeout, these timeouts are now
1434 treated as errors, giving rise to delivery failures by default. However,
1435 there is now a new Boolean option for the pipe transport called
1436 timeout_defer, which, if set TRUE, converts the failures into defers for
1437 both kinds of timeout. A transport filter timeout is now identified in
1440 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1441 systems where "make" and "gmake" are different, calling "gmake" at top
1442 level broke things. I've arranged for the value of $(MAKE) to be passed
1443 from the Makefile to this script so that it can call the same version of
1447 A note about Exim versions 4.44 and 4.50
1448 ----------------------------------------
1450 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1451 changes of various kinds. As a consequence, a big documentation update was
1452 needed. This delayed the release for rather longer than seemed good, especially
1453 in the light of a couple of (minor) security issues. Therefore, the changes
1454 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1455 release. So 4.44 and 4.50 are in effect two different branches that both start
1458 I have left the 4.50 change log unchanged; it contains all the changes since
1459 4.43. The change log for 4.44 is below; many of its items are identical to
1460 those for 4.50. This seems to be the most sensible way to preserve the
1461 historical information.
1467 1. Minor wording change to the doc/README.SIEVE file.
1469 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1470 computation of the current number of files was incorrect.
1472 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1473 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1476 4. Give more explanation in the error message when the command for a transport
1477 filter fails to execute.
1479 5. There are several places where Exim runs a non-Exim command in a
1480 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1481 was being done only for the command run by the queryprogram router. It is
1482 now done for all such subprocesses. The other cases are: ${run, transport
1483 filters, and the commands run by the lmtp and pipe transports.
1485 6. Added CONFIGURE_GROUP build-time option.
1487 7. Some older OS have a limit of 256 on the maximum number of file
1488 descriptors. Exim was using setrlimit() to set 1000 as a large value
1489 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1490 systems. I've change it so that if it can't get 1000, it tries for 256.
1492 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1493 was an oversight, and furthermore, ever since the addition of extra
1494 controls (e.g. 4.43/32), the checks on when to allow different forms of
1495 "control" were broken. There should now be diagnostics for all cases when a
1496 control that does not make sense is encountered.
1498 9. Added the /retain_sender option to "control=submission".
1500 10. $recipients is now available in the predata ACL (oversight).
1502 11. Tidy the search cache before the fork to do a delivery from a message
1503 received from the command line. Otherwise the child will trigger a lookup
1504 failure and thereby defer the delivery if it tries to use (for example) a
1505 cached ldap connection that the parent has called unbind on.
1507 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1508 of $address_data from the recipient verification was clobbered by the
1509 sender verification.
1511 13. The value of address_data from a sender verification is now available in
1512 $sender_address_data in subsequent conditions in the ACL statement.
1514 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1516 15. Added a new option "connect=<time>" to callout options, to set a different
1519 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1520 was its contents. (It was OK if the option was not defined at all.)
1522 17. A "Completed" log line is now written for messages that are removed from
1523 the spool by the -Mrm option.
1525 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1526 information about exactly what failed.
1528 19. Added -dd to debug only the daemon process.
1530 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1531 handles timeouts, both on the server side and network timeouts. Renamed the
1532 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1534 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1535 It is now set to "smtps".
1537 22. $host_address is now set to the target address during the checking of
1538 ignore_target_hosts.
1540 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1541 being passed; this would have caused $sender_host_name to have been used if
1542 matching the list had actually called for a host name (not very likely,
1543 since this list is usually IP addresses). A host name is now passed as
1546 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1547 code that specifies a non-restarting handler (typically sigaction() in
1548 modern systems) in an attempt to fix a rare and obscure crash bug.
1550 25. Narrowed the window for a race in the daemon that could cause it to ignore
1551 SIGCHLD signals. This is not a major problem, because they are used only to
1552 wake it up if nothing else does.
1554 26. A malformed maildirsize file could cause Exim to calculate negative values
1555 for the mailbox size or file count. Odd effects could occur as a result.
1556 The maildirsize information is now recalculated if the size or filecount
1559 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1560 support for a long time. Removed HAVE_SYS_VFS_H.
1562 28. Installed the latest version of exipick from John Jetmore.
1564 29. In an address list, if the pattern was not a regular expression, an empty
1565 subject address (from a bounce message) matched only if the pattern was an
1566 empty string. Non-empty patterns were not even tested. This was the wrong
1567 because it is perfectly reasonable to use an empty address as part of a
1568 database query. An empty address is now tested by patterns that are
1569 lookups. However, all the other forms of pattern expect the subject to
1570 contain a local part and a domain, and therefore, for them, an empty
1571 address still always fails if the pattern is not itself empty.
1573 30. Exim went into a mad DNS loop when attempting to do a callout where the
1574 host was specified on an smtp transport, and looking it up yielded more
1575 than one IP address.
1577 31. Re-factored the code for checking spool and log partition space into a
1578 function that finds that data and another that does the check. The former
1579 is then used to implement four new variables: $spool_space, $log_space,
1580 $spool_inodes, and $log_inodes.
1582 32. The RFC2047 encoding function was originally intended for short strings
1583 such as real names; it was not keeping to the 75-character limit for
1584 encoded words that the RFC imposes. It now respects the limit, and
1585 generates multiple encoded words if necessary. To be on the safe side, I
1586 have increased the buffer size for the ${rfc2047: expansion operator from
1589 33. It is now permitted to omit both strings after an "if" condition; if the
1590 condition is true, the result is "true". As before, when the second string
1591 is omitted, a false condition yields an empty string. This makes it less
1592 cumbersome to write custom ACL and router conditions.
1594 34. Failure to deliver a bounce message always caused it to be frozen, even if
1595 there was an errors_to setting on the router. The errors_to setting is now
1598 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1599 canonical form (fully expanded) before being placed in
1600 $sender_host_address.
1602 36. The table in the code that translates DNS record types into text (T_A to
1603 "A" for instance) was missing entries for NS and CNAME. It is just possible
1604 that this could have caused confusion if both these types were looked up
1605 for the same domain, because the text type is used as part of Exim's
1606 per-process caching. But the chance of anyone hitting this buglet seems
1609 37. The dnsdb lookup has been extended in a number of ways.
1611 (1) There is a new type, "zns", which walks up the domain tree until it
1612 finds some nameserver records. It should be used with care.
1614 (2) There is a new type, "mxh", which is like "mx" except that it returns
1615 just the host names, not the priorities.
1617 (3) It is now possible to give a list of domains (or IP addresses) to be
1618 looked up. The behaviour when one of the lookups defers can be
1619 controlled by a keyword.
1621 (4) It is now possible to specify the separator character for use when
1622 multiple records are returned.
1624 38. The dnslists ACL condition has been extended: it is now possible to supply
1625 a list of IP addresses and/or domains to be looked up in a particular DNS
1628 39. Added log_selector=+queue_time_overall.
1630 40. When running the queue in the test harness, wait just a tad after forking a
1631 delivery process, to get repeatability of debugging output.
1633 41. Include certificate and key file names in error message when GnuTLS fails
1634 to set them up, because the GnuTLS error message doesn't include the name
1635 of the failing file when there is a problem reading it.
1637 42. Allow both -bf and -bF in the same test run.
1639 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1641 44. The "Exiscan patch" is now merged into the mainline Exim source.
1643 45. Sometimes the final signoff response after QUIT could fail to get
1644 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1645 < 0 before doing a fflush(). This bug looks as though it goes back to the
1646 introduction of TLS in release 3.20, but "sometimes" must have been rare
1647 because the tests only now provoked it.
1649 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1650 (this can affect the format of dates).
1652 47. exim_tidydb, when checking for the continued existence of a message for
1653 which it has found a message-specific retry record, was not finding
1654 messages that were in split spool directories. Consequently, it was
1655 deleting retry records that should have stayed in existence.
1657 48. Steve fixed some bugs in eximstats.
1659 49. The SPA authentication driver was not abandoning authentication and moving
1660 on to the next authenticator when an expansion was forced to fail,
1661 contradicting the general specification for all authenticators. Instead it
1662 was generating a temporary error. It now behaves as specified.
1664 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1665 (the order specifies the preference for clients). The order is now AES256,
1666 AES128, 3DES, ARCFOUR128.
1668 51. Small patch to Sieve code - explicitly set From: when generating an
1671 52. Exim crashed if a remote delivery caused a very long error message to be
1672 recorded - for instance if somebody sent an entire SpamAssassin report back
1673 as a large number of 550 error lines. This bug was coincidentally fixed by
1674 increasing the size of one of Exim's internal buffers (big_buffer) that
1675 happened as part of the Exiscan merge. However, to be on the safe side, I
1676 have made the code more robust (and fixed the comments that describe what
1679 53. Now that there can be additional text after "Completed" in log lines (if
1680 the queue_time_overall log selector is set), a one-byte patch to exigrep
1681 was needed to allow it to recognize "Completed" as not the last thing in
1684 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1685 patch that reportedly fixes this has been added. I am not expert enough to
1686 create a test for it. This is what the patch creator wrote:
1688 "I found a little strange behaviour of ldap code when working with
1689 Windows 2003 AD Domain, where users was placed in more than one
1690 Organization Units. When I tried to give exim partial DN, the exit code
1691 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1692 But simultaneously result of request was absolutely normal ldap result,
1693 so I produce this patch..."
1695 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1696 REFERENCE, so I have modified the code to exclude the patch when that macro
1699 55. Some experimental protocols are using DNS PTR records for new purposes. The
1700 keys for these records are domain names, not reversed IP addresses. The
1701 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1702 leaves it alone. Component reversal etc. now happens only for IP addresses.
1705 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1707 57. Double the size of the debug message buffer (to 2048) so that more of very
1708 long debug lines gets shown.
1710 58. The exicyclog utility now does better if the number of log files to keep
1711 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1713 59. Two changes related to the smtp_active_hostname option:
1715 (1) $smtp_active_hostname is now available as a variable.
1716 (2) The default for smtp_banner uses $smtp_active_hostname instead
1717 of $primary_hostname.
1719 60. The host_aton() function is supposed to be passed a string that is known
1720 to be a valid IP address. However, in the case of IPv6 addresses, it was
1721 not checking this. This is a hostage to fortune. Exim now panics and dies
1722 if the condition is not met. A case was found where this could be provoked
1723 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1724 components; fortuitously, this particular loophole had already been fixed
1725 by change 4.50/55 above.
1727 If there are any other similar loopholes, the new check in host_aton()
1728 itself should stop them being exploited. The report I received stated that
1729 data on the command line could provoke the exploit when Exim was running as
1730 exim, but did not say which command line option was involved. All I could
1731 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1732 running as the user.
1735 61. There was a buffer overflow vulnerability in the SPA authentication code
1736 (which came originally from the Samba project). I have added a test to the
1737 spa_base64_to_bits() function which I hope fixes it.
1740 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1741 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1743 63. The daemon start-up calls getloadavg() while still root for those OS that
1744 need the first call to be done as root, but it missed one case: when
1745 deliver_queue_load_max is set with deliver_drop_privilege. This is
1746 necessary for the benefit of the queue runner, because there is no re-exec
1747 when deliver_drop_privilege is set.
1749 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1750 This has been fixed.
1752 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1753 was in use, was not putting the data itself into the right store pool;
1754 consequently, it could be overwritten for a subsequent message in the same
1755 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1758 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1760 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1761 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1762 the function were treating the return as a boolean value, which happened to
1763 work because 0=false and not-0=true, but is not correct code.
1765 68. The host_aton() function was not handling scoped IPv6 addresses (those
1766 with, for example, "%eth0" on the end) correctly.
1768 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1769 negated items (that is, ~something) in unsigned ints. Some compilers
1770 apparently mutter when there is no cast.
1772 70. If an address verification called from an ACL failed, and did not produce a
1773 user-specific message (i.e. there was only a "system" message), nothing was
1774 put in $acl_verify_message. In this situation, it now puts the system
1777 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1778 session; change 4.31/43 added the unwanted input to the log line - except
1779 that it did not do this in the start of session case. It now does.
1781 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1782 This is wrong and can cause the other end to generate a synchronization
1783 error if it is another Exim or anything else that does the synchronization
1784 check. A QUIT command is no longer sent after a timeout.
1786 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1787 during host lookups.
1789 74. The defer_ok option of callout verification was not working if it was used
1790 when verifying addresses in header lines, that is, for this case:
1792 verify = header_sender/callout=defer_ok
1794 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1795 those file descriptors could be used for SMTP connections. If anything
1796 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1797 could be sent to the SMTP client, causing chaos. The daemon now opens
1798 stdin, stdout, and stderr to /dev/null when it puts itself into the
1801 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1802 log by default. The user can override this with suitable Perl magic.
1804 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1805 the log message when discard triggers, was not working for the DATA ACL or
1806 for the non-SMTP ACL.
1808 78. Error message wording change in sieve.c.
1810 79. If smtp_accept_max_per_host was set, the number of connections could be
1811 restricted to fewer than expected, because the daemon was trying to set up
1812 a new connection before checking whether the processes handling previous
1813 connections had finished. The check for completed processes is now done
1814 earlier. On busy systems, this bug wouldn't be noticed because something
1815 else would have woken the daemon, and it would have reaped the completed
1818 80. If a message was submitted locally by a user whose login name contained one
1819 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1820 It caused a spool format error. I have fixed the spool reading code. A
1821 related problem was that the "from" clause in the Received: line became
1822 illegal because of the space(s). It is now covered by ${quote_local_part.
1824 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1827 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1829 83. Patch from Sieve maintainer to fix unterminated string problem in
1830 "vacation" handling.
1832 84. Some minor changes to the Linux configuration files to help with other
1833 OS variants using glibc.
1835 85. One more patch for Sieve to update vacation handling to latest spec.
1838 ----------------------------------------------------
1839 See the note above about the 4.44 and 4.50 releases.
1840 ----------------------------------------------------
1846 1. Change 4.43/35 introduced a bug that caused file counts to be
1847 incorrectly computed when quota_filecount was set in an appendfile
1850 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1851 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1854 3. Give more explanation in the error message when the command for a transport
1855 filter fails to execute.
1857 4. There are several places where Exim runs a non-Exim command in a
1858 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1859 was being done only for the command run by the queryprogram router. It is
1860 now done for all such subprocesses. The other cases are: ${run, transport
1861 filters, and the commands run by the lmtp and pipe transports.
1863 5. Some older OS have a limit of 256 on the maximum number of file
1864 descriptors. Exim was using setrlimit() to set 1000 as a large value
1865 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1866 systems. I've change it so that if it can't get 1000, it tries for 256.
1868 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1869 was an oversight, and furthermore, ever since the addition of extra
1870 controls (e.g. 4.43/32), the checks on when to allow different forms of
1871 "control" were broken. There should now be diagnostics for all cases when a
1872 control that does not make sense is encountered.
1874 7. $recipients is now available in the predata ACL (oversight).
1876 8. Tidy the search cache before the fork to do a delivery from a message
1877 received from the command line. Otherwise the child will trigger a lookup
1878 failure and thereby defer the delivery if it tries to use (for example) a
1879 cached ldap connection that the parent has called unbind on.
1881 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1882 of $address_data from the recipient verification was clobbered by the
1883 sender verification.
1885 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1886 was its contents. (It was OK if the option was not defined at all.)
1888 11. A "Completed" log line is now written for messages that are removed from
1889 the spool by the -Mrm option.
1891 12. $host_address is now set to the target address during the checking of
1892 ignore_target_hosts.
1894 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1895 being passed; this would have caused $sender_host_name to have been used if
1896 matching the list had actually called for a host name (not very likely,
1897 since this list is usually IP addresses). A host name is now passed as
1900 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1901 code that specifies a non-restarting handler (typically sigaction() in
1902 modern systems) in an attempt to fix a rare and obscure crash bug.
1904 15. Narrowed the window for a race in the daemon that could cause it to ignore
1905 SIGCHLD signals. This is not a major problem, because they are used only to
1906 wake it up if nothing else does.
1908 16. A malformed maildirsize file could cause Exim to calculate negative values
1909 for the mailbox size or file count. Odd effects could occur as a result.
1910 The maildirsize information is now recalculated if the size or filecount
1913 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1914 support for a long time. Removed HAVE_SYS_VFS_H.
1916 18. Updated exipick to current release from John Jetmore.
1918 19. Allow an empty sender to be matched against a lookup in an address list.
1919 Previously the only cases considered were a regular expression, or an
1922 20. Exim went into a mad DNS lookup loop when doing a callout where the
1923 host was specified on the transport, if the DNS lookup yielded more than
1926 21. The RFC2047 encoding function was originally intended for short strings
1927 such as real names; it was not keeping to the 75-character limit for
1928 encoded words that the RFC imposes. It now respects the limit, and
1929 generates multiple encoded words if necessary. To be on the safe side, I
1930 have increased the buffer size for the ${rfc2047: expansion operator from
1933 22. Failure to deliver a bounce message always caused it to be frozen, even if
1934 there was an errors_to setting on the router. The errors_to setting is now
1937 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1938 canonical form (fully expanded) before being placed in
1939 $sender_host_address.
1941 24. Updated eximstats to version 1.33
1943 25. Include certificate and key file names in error message when GnuTLS fails
1944 to set them up, because the GnuTLS error message doesn't include the name
1945 of the failing file when there is a problem reading it.
1947 26. Expand error message when OpenSSL has problems setting up cert/key files.
1950 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1951 (this can affect the format of dates).
1953 28. exim_tidydb, when checking for the continued existence of a message for
1954 which it has found a message-specific retry record, was not finding
1955 messages that were in split spool directories. Consequently, it was
1956 deleting retry records that should have stayed in existence.
1958 29. eximstats updated to version 1.35
1959 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1960 1.35 - bugfix such that pie charts by volume are generated correctly
1962 30. The SPA authentication driver was not abandoning authentication and moving
1963 on to the next authenticator when an expansion was forced to fail,
1964 contradicting the general specification for all authenticators. Instead it
1965 was generating a temporary error. It now behaves as specified.
1967 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1968 (the order specifies the preference for clients). The order is now AES256,
1969 AES128, 3DES, ARCFOUR128.
1971 31. Small patch to Sieve code - explicitly set From: when generating an
1974 32. Exim crashed if a remote delivery caused a very long error message to be
1975 recorded - for instance if somebody sent an entire SpamAssassin report back
1976 as a large number of 550 error lines. This bug was coincidentally fixed by
1977 increasing the size of one of Exim's internal buffers (big_buffer) that
1978 happened as part of the Exiscan merge. However, to be on the safe side, I
1979 have made the code more robust (and fixed the comments that describe what
1982 33. Some experimental protocols are using DNS PTR records for new purposes. The
1983 keys for these records are domain names, not reversed IP addresses. The
1984 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1985 leaves it alone. Component reversal etc. now happens only for IP addresses.
1988 34. The host_aton() function is supposed to be passed a string that is known
1989 to be a valid IP address. However, in the case of IPv6 addresses, it was
1990 not checking this. This is a hostage to fortune. Exim now panics and dies
1991 if the condition is not met. A case was found where this could be provoked
1992 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1993 components; fortuitously, this particular loophole had already been fixed
1994 by change 4.50/55 or 4.44/33 above.
1996 If there are any other similar loopholes, the new check in host_aton()
1997 itself should stop them being exploited. The report I received stated that
1998 data on the command line could provoke the exploit when Exim was running as
1999 exim, but did not say which command line option was involved. All I could
2000 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2001 running as the user.
2004 35. There was a buffer overflow vulnerability in the SPA authentication code
2005 (which came originally from the Samba project). I have added a test to the
2006 spa_base64_to_bits() function which I hope fixes it.
2009 36. The daemon start-up calls getloadavg() while still root for those OS that
2010 need the first call to be done as root, but it missed one case: when
2011 deliver_queue_load_max is set with deliver_drop_privilege. This is
2012 necessary for the benefit of the queue runner, because there is no re-exec
2013 when deliver_drop_privilege is set.
2015 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2016 was in use, was not putting the data itself into the right store pool;
2017 consequently, it could be overwritten for a subsequent message in the same
2018 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2021 38. Sometimes the final signoff response after QUIT could fail to get
2022 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2023 < 0 before doing a fflush(). This bug looks as though it goes back to the
2024 introduction of TLS in release 3.20, but "sometimes" must have been rare
2025 because the tests only now provoked it.
2031 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2032 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2033 Now it returns an integer. A number of calls were still expecting a T/F
2034 return. Fortuitously, in all cases, the tests worked in OK situations,
2035 which is the norm. However, things would have gone wrong on any write
2036 failures on the smtp file descriptor. This function is used when sending
2037 messages over SMTP and also when doing verify callouts.
2039 2. When Exim is called to do synchronous delivery of a locally submitted
2040 message (the -odf or -odi options), it no longer closes stderr before doing
2043 3. Implemented the mua_wrapper option.
2045 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2047 5. Implemented the functions header_remove(), header_testname(),
2048 header_add_at_position(), and receive_remove_recipient(), and exported them
2051 6. If an ACL "warn" statement specified the addition of headers, Exim already
2052 inserted X-ACL-Warn: at the start if there was no header name. However, it
2053 was not making this test for the second and subsequent header lines if
2054 there were newlines in the string. This meant that an invalid header could
2055 be inserted if Exim was badly configured.
2057 7. Allow an ACL "warn" statement to add header lines at the start or after all
2058 the Received: headers, as well as at the end.
2060 8. Added the rcpt_4xx retry error code.
2062 9. Added postmaster_mailfrom=xxx to callout verification option.
2064 10. Added mailfrom=xxxx to the callout verification option, for verify=
2067 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2068 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2070 12. Inserted some casts to stop certain compilers warning when using pointer
2071 differences as field lengths or precisions in printf-type calls (mostly
2072 affecting debugging statements).
2074 13. Added optional readline() support for -be (dynamically loaded).
2076 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2077 same clock tick as a message's arrival, so that its received time was the
2078 same as the "first fail" time on the retry record, and that message
2079 remained on the queue past the ultimate address timeout, every queue runner
2080 would try a delivery (because it was past the ultimate address timeout) but
2081 after another failure, the ultimate address timeout, which should have then
2082 bounced the address, did not kick in. This was a "< instead of <=" error;
2083 in most cases the first failure would have been in the next clock tick
2084 after the received time, and all would be well.
2086 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2087 being recognized when the domain list was tested by the match_domain
2088 condition in an expansion string.
2090 16. Added the ${str2b64: operator.
2092 17. Exim was always calling setrlimit() to set a large limit for the number of
2093 processes, without checking whether the existing limit was already
2094 adequate. (It did check for the limit on file descriptors.) Furthermore,
2095 errors from getrlimit() and setrlimit() were being ignored. Now they are
2096 logged to the main and panic logs, but Exim does carry on, to try to do its
2097 job under whatever limits there are.
2099 18. Imported PCRE 5.0.
2101 19. Trivial typo in log message " temporarily refused connection" (the leading
2104 20. If the log selector return_path_on_delivery was set and an address was
2105 redirected to /dev/null, the delivery process crashed because it assumed
2106 that a return path would always be set for a "successful" delivery. In this
2107 case, the whole delivery is bypassed as an optimization, and therefore no
2110 21. Internal re-arrangement: the function for sending a challenge and reading
2111 a response while authentication was assuming a zero-terminated challenge
2112 string. It's now changed to take a pointer and a length, to allow for
2113 binary data in such strings.
2115 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2117 23. Exim was not respecting finduser_retries when seeking the login of the
2118 uid under which it was called; it was always trying 10 times. (The default
2119 setting of finduser_retries is zero.) Also, it was sleeping after the final
2120 failure, which is pointless.
2122 24. Implemented tls_on_connect_ports.
2124 25. Implemented acl_smtp_predata.
2126 26. If the domain in control=submission is set empty, Exim assumes that the
2127 authenticated id is a complete email address when it generates From: or
2128 Sender: header lines.
2130 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2131 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2132 chown and chgrp in /bin and hostname in /usr/bin.
2134 28. Exim was keeping the "process log" file open after each use, just as it
2135 does for the main log. This opens the possibility of it remaining open for
2136 long periods when the USR1 signal hits a daemon. Occasional processlog
2137 errors were reported, that could have been caused by this. Anyway, it seems
2138 much more sensible not to leave this file open at all, so that is what now
2141 29. The long-running daemon process does not normally write to the log once it
2142 has entered its main loop, and it closes the log before doing so. This is
2143 so that log files can straightforwardly be renamed and moved. However,
2144 there are a couple of unusual error situations where the daemon does write
2145 log entries, and I had neglected to close the log afterwards.
2147 30. The text of an SMTP error response that was received during a remote
2148 delivery was being truncated at 512 bytes. This is too short for some of
2149 the long messages that one sometimes sees. I've increased the limit to
2152 31. It is now possible to make retry rules that apply only when a message has a
2153 specific sender, in particular, an empty sender.
2155 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2156 it possible to be selective about when SMTP synchronization is enforced.
2158 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2160 32. Implemented hosts_connection_nolog.
2162 33. Added an ACL for QUIT.
2164 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2167 35. Added mailbox_size and mailbox_filecount to appendfile.
2169 36. Added control = no_multiline_responses to ACLs.
2171 37. There was a bug in the logic of the code that waits for the clock to tick
2172 in the case where the clock went backwards by a substantial amount such
2173 that the microsecond fraction of "now" was more than the microsecond
2174 fraction of "then" (but the whole seconds number was less).
2176 38. Added support for the libradius Radius client library this is found on
2177 FreeBSD (previously only the radiusclient library was supported).
2183 1. When certain lookups returned multiple values in the form name=value, the
2184 quoting of the values was not always being done properly. Specifically:
2185 (a) If the value started with a double quote, but contained no whitespace,
2187 (b) If the value contained whitespace other than a space character (i.e.
2188 tabs or newlines or carriage returns) it was not quoted.
2189 This fix has been applied to the mysql and pgsql lookups by writing a
2190 separate quoting function and calling it from the lookup code. The fix
2191 should probably also be applied to nisplus, ibase and oracle lookups, but
2192 since I cannot test any of those, I have not disturbed their existing code.
2194 2. A hit in the callout cache for a specific address caused a log line with no
2195 reason for rejecting RCPT. Now it says "Previous (cached) callout
2196 verification failure".
2198 3. There was an off-by-one bug in the queryprogram router. An over-long
2199 return line was truncated at 256 instead of 255 characters, thereby
2200 overflowing its buffer with the terminating zero. As well as fixing this, I
2201 have increased the buffer size to 1024 (and made a note to document this).
2203 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2204 when Exim is waiting for an SMTP response from a remote server, Exim
2205 restarts its select() call on the socket, thereby resetting its timeout.
2206 This is not a problem when such interrupts are rare. Somebody set up a cron
2207 job to run exiwhat every 2 minutes, which is less than the normal select()
2208 timeout (5 or 10 minutes). This meant that the select() timeout never
2209 kicked in because it was always reset. I have fixed this by comparing the
2210 time when an interrupt arrives with the time at the start of the first call
2211 to select(). If more time than the timeout has elapsed, the interrupt is
2212 treated as a timeout.
2214 5. Some internal re-factoring in preparation for the addition of Sieve
2215 extensions (by MH). In particular, the "personal" test is moved to a
2216 separate function, and given an option for scanning Cc: and Bcc: (which is
2217 not set for Exim filters).
2219 6. When Exim created an email address using the login of the caller as the
2220 local part (e.g. when creating a From: or Sender: header line), it was not
2221 quoting the local part when it contained special characters such as @.
2223 7. Installed new OpenBSD configuration files.
2225 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2226 try to make them clearer.
2228 9. Callout options, other than the timeout value, were being ignored when
2229 verifying sender addresses in header lines. For example, when using
2231 verify = header_sender/callout=no_cache
2233 the cache was (incorrectly) being used.
2235 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2236 only the Cygwin environment.
2238 11. When return_path_on_delivery was set as a log selector, if different remote
2239 addresses in the same message used different return paths and parallel
2240 remote delivery occurred, the wrong values would sometimes be logged.
2241 (Whenever a remote delivery process finished, the return path value from
2242 the most recently started remote delivery process was logged.)
2244 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2245 lines when AUTH and/or TLS are in use. This is the "received protocol"
2246 field. Exim used to use "asmtp" for authenticated SMTP, without any
2247 indication (in the protocol name) for TLS use. Now it follows the RFC and
2248 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2249 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2250 names appear in log lines as well as in Received: header lines.
2252 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2253 extensions, and comparison tests, and to fix some bugs.
2255 14. Changes to the "personal" filter test:
2257 (1) The test was buggy in that it was just doing the equivalent of
2258 "contains" tests on header lines. For example, if a user's address was
2259 anne@some.where, the "personal" test would incorrectly be true for
2261 To: susanne@some.where
2263 This test is now done by extracting each address from the header in turn,
2264 and checking the entire address. Other tests that are part of "personal"
2265 are now done using regular expressions (for example, to check local parts
2266 of addresses in From: header lines).
2268 (2) The list of non-personal local parts in From: addresses has been
2269 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2270 taken from the Sieve specification recommendations.
2272 (3) If the message contains any header line starting with "List-" it is
2273 treated as non-personal.
2275 (4) The test for "circular" in the Subject: header line has been removed
2276 because it now seems ill-conceived.
2278 15. Minor typos in src/EDITME comments corrected.
2280 16. Installed latest exipick from John Jetmore.
2282 17. If headers_add on a router specified a text string that was too long for
2283 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2284 of string_sprintf() is now avoided.
2286 18. $message_body_size was not set (it was always zero) when running the DATA
2287 ACL and the local_scan() function.
2289 19. For the "mail" command in an Exim filter, no default was being set for
2290 the once_repeat time, causing a random time value to be used if "once" was
2291 specified. (If the value happened to be <= 0, no repeat happened.) The
2292 default is now 0s, meaning "never repeat". The "vacation" command was OK
2293 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2294 (I found it when inspecting the code).
2296 20. There is now an overall timeout for performing a callout verification. It
2297 defaults to 4 times the callout timeout, which applies to individual SMTP
2298 commands during the callout. The overall timeout applies when there is more
2299 than one host that can be tried. The timeout is checked before trying the
2300 next host. This prevents very long delays if there are a large number of
2301 hosts and all are timing out (e.g. when the network connections are timing
2302 out). The value of the overall timeout can be changed by specifying an
2303 additional sub-option for "callout", called "maxwait". For example:
2305 verify = sender/callout=5s,maxwait=20s
2307 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2308 to the end before writing, but this should make it even safer).
2310 22. Exim was forgetting that it had advertised PIPELINING for the second and
2311 subsequent messages on an SMTP connection. It was also not resetting its
2312 memory on STARTTLS and an internal HELO.
2314 23. When Exim logs an SMTP synchronization error within a session, it now
2315 records whether PIPELINING has been advertised or not.
2317 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2318 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2319 rather than long int.
2321 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2323 26. Added the never_mail option to autoreply.
2329 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2330 crash if the getsockname() call failed; this can happen if a connection is
2331 closed very soon after it is established. The problem was simply in the
2332 order in which certain operations were done, causing Exim to try to write
2333 to the SMTP stream before it had set up the file descriptor. The bug has
2334 been fixed by making things happen in the correct order.
2340 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2341 before the connection was closed, thus losing the rejection response.
2343 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2344 some early Solaris releases, but causes trouble in current releases where
2345 socklen_t is defined.
2347 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2350 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2353 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2354 that releasing the top of it at the end releases what was used for sub-
2355 expansions (unless the block got too big). However, discard this block if
2356 the first thing is a variable or header, so that we can use its block when
2357 it is dynamic (useful for very large $message_headers, for example).
2359 6. Lookups now cache *every* query, not just the most recent. A new, separate
2360 store pool is used for this. It can be recovered when all lookup caches are
2361 flushed. Lookups now release memory at the end of their result strings.
2362 This has involved some general refactoring of the lookup sources.
2364 7. Some code has been added to the store_xxx() functions to reduce the amount
2365 of flapping under certain conditions.
2367 8. log_incoming_interface used to affect only the <= reception log lines. Now
2368 it causes the local interface and port to be added to several more SMTP log
2369 lines, for example "SMTP connection from", and rejection lines.
2371 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2373 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2375 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2376 could be overwritten at the end of the current message (or the start of a
2377 new message if it was set in a HELO ACL). The value is now preserved for
2378 the duration of the SMTP connection.
2380 12. If a transport had a headers_rewrite setting, and a matching header line
2381 contained an unqualified address, that address was qualified, even if it
2382 did not match any rewriting rules. The underlying bug was that the values
2383 of the flags that permit the existence of unqualified sender and recipient
2384 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2385 non-local messages, and by -bnq for local messages) were not being
2386 preserved with the message after it was received.
2388 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2389 "next input=" as part of the text comprising the host identity instead of
2390 the correct text. The code was using the same buffer for two different
2391 strings. However, depending on which order the printing function evaluated
2392 its arguments, the bug did not always show up. Under Linux, for example, my
2393 test suite worked just fine.
2395 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2396 doesn't work with some older versions of Perl. It has been changed to "my",
2397 which in any case is probably the better facility to use.
2399 15. A really picky compiler found some instances of statements for creating
2400 error messages that either had too many or two few arguments for the format
2403 16. The size of the buffer for calls to the DNS resolver has been increased
2404 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2405 for addresses that have a lot of PTR records. This alleviates a problem; it
2406 does not fully solve it.
2408 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2409 buffer now truncates the list and logs the incident, which is the same
2410 action as happens when Exim is looking up a host name and its aliases.
2411 Previously in this situation something unpredictable would happen;
2412 sometimes it was "internal error: store_reset failed".
2414 18. If a server dropped the connection unexpectedly when an Exim client was
2415 using GnuTLS and trying to read a response, the client delivery process
2416 crashed while trying to generate an error log message.
2418 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2419 string, for example:
2421 warn message = H1: something\nH2: something
2423 the text was added as a single header line from Exim's point of view
2424 though it ended up OK in the delivered message. However, searching for the
2425 second and subsequent header lines using $h_h2: did not work. This has been
2426 fixed. Similarly, if a system filter added multiple headers in this way,
2427 the routers could not see them.
2429 20. Expanded the error message when iplsearch is called with an invalid key to
2430 suggest using net-iplsearch in a host list.
2432 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2433 ACLs are no longer actually imposed (and a message to that effect is
2436 22. If a "gecos" field in a passwd entry contained escaped characters, in
2437 particular, if it contained a \" sequence, Exim got it wrong when building
2438 a From: or a Sender: header from that name. A second bug also caused
2439 incorrect handling when an unquoted " was present following a character
2440 that needed quoting.
2442 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2443 was not being matched caselessly.
2445 24. Arranged for all hyphens in the exim.8 source to be escaped with
2448 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2449 properly. Recipient callout cache records were still being keyed to include
2450 the sender, even when use_sender was set false. This led to far more
2451 callouts that were necessary. The sender is no longer included in the key
2452 when use_sender is false.
2454 26. Added "control = submission" modifier to ACLs.
2456 27. Added the ${base62d: operator to decode base 62 numbers.
2458 28. dnsdb lookups can now access SRV records.
2460 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2461 the configuration file.
2463 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2464 (-v) mode. This makes the output for a verbose queue run more intelligible.
2466 31. Added a use_postmaster feature to recipient callouts.
2468 32. Added the $body_zerocount variable, containing the number of binary zero
2469 bytes in the message body.
2471 33. The time of last modification of the "new" subdirectory is now used as the
2472 "mailbox time last read" when there is a quota error for a maildir
2475 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2477 35. Added +ignore_unknown as a special item in host lists.
2479 36. Code for decoding IPv6 addresses in host lists is now included, even if
2480 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2481 address was recognized as an IP address, but was then not correctly decoded
2482 into binary, causing unexpected and incorrect effects when compared with
2489 1. Very minor rewording of debugging text in manualroute to say "list of
2490 hosts" instead of "hostlist".
2492 2. If verify=header_syntax was set, and a header line with an unqualified
2493 address (no domain) and a large number of spaces between the end of the
2494 name and the colon was received, the reception process suffered a buffer
2495 overflow, and (when I tested it) crashed. This was caused by some obsolete
2496 code that should have been removed. The fix is to remove it!
2498 3. When running in the test harness, delay a bit after writing a bounce
2499 message to get a bit more predictability in the log output.
2501 4. Added a call to search_tidyup() just before forking a reception process. In
2502 theory, someone could use a lookup in the expansion of smtp_accept_max_
2503 per_host which, without the tidyup, could leave open a database connection.
2505 5. Added the variables $recipient_data and $sender_data which get set from a
2506 lookup success in an ACL "recipients" or "senders" condition, or a router
2507 "senders" option, similar to $domain_data and $local_part_data.
2509 6. Moved the writing of debug_print from before to after the "senders" test
2512 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2513 problems for message scanning, either using a data ACL, or using
2514 local_scan() because the Received: header was not generated till after they
2515 were called (in order to set the time as the time of reception completion).
2516 I have revised the way this works. The header is now generated after the
2517 body is received, but before the ACL or local_scan() are called. After they
2518 are run, the timestamp in the header is updated.
2524 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2525 before starting a queue runner without re-exec. This happened only when
2526 deliver_drop_privilege was set or when the Exim user was set to root. The
2527 effect of the bug was that timeouts during subsequent deliveries caused
2528 crashes instead of being properly handled. The handler is now left at its
2529 default (and expected) setting.
2531 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2532 message, again when deliver_drop_privilege is set or Exim is run as root.
2533 The bug described in (1) was not present in this case, but the tidying up
2534 of the other signals was missing. I have made the two cases consistent.
2536 3. The ignore_target_hosts setting on a manualroute router was being ignored
2537 for hosts that were looked up using the /MX notation.
2539 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2542 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2543 operated on the sender address. After changing the $sender_address to <>
2544 for the sender address verify, Exim was re-instated it as the original
2545 (before rewriting) address, but remembering that it had rewritten it, so it
2546 wasn't rewriting it again. This bug also had the effect of breaking the
2547 sender address verification caching when the sender address was rewritten.
2549 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2550 This has been changed so that if the ip literal address matches
2551 ignore_target_hosts, the router declines.
2553 7. Added expansion conditions match_domain, match_address, and match_local_
2554 part (NOT match_host).
2556 8. The placeholder for the Received: header didn't have a length field set.
2558 9. Added code to Exim itself and to exim_lock to test for a specific race
2559 condition that could lead to file corruption when using MBX delivery. The
2560 issue is with the lockfile that is created in /tmp. If this file is removed
2561 after a process has opened it but before that process has acquired a lock,
2562 there is the potential for a second process to recreate the file and also
2563 acquire a lock. This could lead to two Exim processes writing to the file
2564 at the same time. The added code performs the same test as UW imapd; it
2565 checks after acquiring the lock that its file descriptor still refers to
2566 the same named file.
2568 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2569 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2572 11. Added the smtp_active_hostname option. If used, this will typically be made
2573 to depend on the incoming interface address. Because $interface_address is
2574 not set up until the daemon has forked a reception process, error responses
2575 that can happen earlier (such as "too many connections") no longer contain
2578 12. If an expansion in a condition on a "warn" statement fails because a lookup
2579 defers, the "warn" statement is abandoned, and the next ACL statement is
2580 processed. Previously this caused the whole ACL to be aborted.
2582 13. Added the iplsearch lookup type.
2584 14. Added ident_timeout as a log selector.
2586 15. Added tls_certificate_verified as a log selector.
2588 16. Added a global option tls_require_ciphers (compare the smtp transport
2589 option of the same name). This controls incoming TLS connections.
2591 17. I finally figured out how to make tls_require_ciphers do a similar thing
2592 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2593 before starting the TLS session.
2595 18. Tabs are now shown as \t in -bP output.
2597 19. If the log selector return_path_on_delivery was set, Exim crashed when
2598 bouncing a message because it had too many Received: header lines.
2600 20. If two routers both had headers_remove settings, and the first one included
2601 a superfluous trailing colon, the final name in the first list and the
2602 first name in the second list were incorrectly joined into one item (with a
2603 colon in the middle).
2609 1. Added -C and -D options to the exinext utility, mainly to make it easier
2610 to include in the automated testing, but these could be helpful when
2611 multiple configurations are in use.
2613 2. The exinext utility was not formatting the output nicely when there was
2614 an alternate port involved in the retry record key, nor when there was a
2615 message id as well (for retries that were specific to a specific message
2616 and a specific host). It was also confused by IPv6 addresses, because of
2617 the additional colons they contain. I have fixed the IPv4 problem, and
2618 patched it up to do a reasonable job for IPv6.
2620 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2621 delivery, the log line now contains "pipelined" if PIPELINING was used.
2623 4. An SMTP transport process used to panic and die if the bind() call to set
2624 an explicit outgoing interface failed. This has been changed; it is now
2625 treated in the same way as a connect() failure.
2627 5. A reference to $sender_host_name in the part of a conditional expansion
2628 that was being skipped was still causing a DNS lookup. This no longer
2631 6. The def: expansion condition was not recognizing references to header lines
2632 that used bh_ and bheader_.
2634 7. Added the _cache feature to named lists.
2636 8. The code for checking quota_filecount in the appendfile transport was
2637 allowing one more file than it should have been.
2639 9. For compatibility with Sendmail, the command line option
2647 and sets the incoming protocol and host name (for trusted callers). The
2648 host name and its colon can be omitted when only the protocol is to be set.
2649 Note the Exim already has two private options, -pd and -ps, that refer to
2650 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2651 "s", but I don't think that's a major issue.
2653 10. A number of refactoring changes to the code, none of which should affect
2656 (a) The number of logging options was getting close to filling up the
2657 32-bit word that was used as a bit map. I have split them into two classes:
2658 those that are passed in the argument to log_write(), and those that are
2659 only ever tested independently outside of that function. These are now in
2660 separate 32-bit words, so there is plenty of room for expansion again.
2661 There is no change in the user interface or the logging behaviour.
2663 (b) When building, for example, log lines, the code previously used a
2664 macro that called string_cat() twice, in order to add two strings. This is
2665 not really sufficiently general. Furthermore, there was one instance where
2666 it was actually wrong because one of the argument was used twice, and in
2667 one call a function was used. (As it happened, calling the function twice
2668 did not affect the overall behaviour.) The macro has been replaced by a
2669 function that can join an arbitrary number of extra strings onto a growing
2672 (c) The code for expansion conditions now uses a table and a binary chop
2673 instead of a serial search (which was left over from when there were very
2674 few conditions). Also, it now recognizes conditions like "pam" even when
2675 the relevant support is not compiled in: a suitably worded error message is
2676 given if an attempt is made to use such a condition.
2678 11. Added ${time_interval:xxxxx}.
2680 12. A bug was causing one of the ddress fields not to be passed back correctly
2681 from remote delivery subprocesses. The field in question was not being
2682 subsequently used, so this caused to problems in practice.
2684 13. Added new log selectors queue_time and deliver_time.
2686 14. Might have fixed a bug in maildirsizefile handling that threw up
2687 "unexpected character" debug warnings, and recalculated the data
2688 unnecessarily. In any case, I expanded the warning message to give more
2691 15. Added the message "Restricted characters in address" to the statements in
2692 the default ACL that block characters like @ and % in local parts.
2694 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2695 Three changes have been made:
2697 (a) There was a serious bug; a negative response to MAIL caused the whole
2698 recipient domain to be cached as invalid, thereby blocking all messages
2699 to all local parts at the same domain, from all senders. This bug has
2700 been fixed. The domain is no longer cached after a negative response to
2701 MAIL if the sender used is not empty.
2703 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2706 (c) A new callout option, "use_sender" has been added for people who want
2707 the modified behaviour.
2713 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2716 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2717 indeed breaks things for older releases.
2719 3. Added additional logging to the case where there is a problem reading data
2720 from a filter that is running in a subprocess using a pipe, in order to
2721 try to track down a specific problem.
2723 4. Testing facility fudge: when running in the test harness and attempting
2724 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2725 getting "No route to host". Convert this to a timeout.
2727 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2730 6. Some OS don't have socklen_t but use size_t instead. This affects the
2731 fifth argument of getsockopt() amongst other things. This is now
2732 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2733 can be set for individual OS. I have set it for SunOS5, OSF1, and
2734 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2735 some earlier ones do not.
2737 7. Change 4.30/15 was not doing the test caselessly.
2739 8. The standard form for an IPv6 address literal was being rejected by address
2740 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2741 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2742 this, as well as the form without the "IPv6" on the front (but only when
2743 address literals are enabled, of course).
2745 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2747 10. Exim crashed if a message with an empty sender address specified by -f
2748 encountered a router with an errors_to setting. This could be provoked only
2749 by a command such as
2753 where an empty string was supplied; "<>" did not hit this bug.
2755 11. Installed PCRE release 4.5.
2757 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2758 remained set. It is now erased.
2760 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2761 times from message ids (which are base 36 rather than the normal 62).
2763 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2764 were being counted as actual protocol errors, and logged if the log
2765 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2766 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2767 and DATA following a set of rejected RCPTs do not count as protocol errors.
2768 In other words, Exim assumes they were pipelined, though this may not
2769 actually be the case. Of course, in all cases the client gets an
2770 appropriate error code.
2772 15. If a lookup fails in an ACL condition, a message about the failure may
2773 be available; it is used if testing the ACL cannot continue, because most
2774 such messages specify what the cause of the deferral is. However, some
2775 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2776 that caused an old message to be retained and used if a later statement
2777 caused a defer, replacing the real cause of the deferral.
2779 16. If an IP address had so many PTR records that the DNS lookup buffer
2780 was not large enough to hold them, Exim could crash while trying to process
2781 the truncated data. It now detects and logs this case.
2783 17. Further to 4.21/58, another change has been made: if (and only if) the
2784 first line of a message (the first header line) ends with CRLF, a bare LF
2785 in a subsequent header line has a space inserted after it, so as not to
2786 terminate the header.
2788 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2789 unnecessarily, used atoi() instead of strtol(), and didn't check the
2790 termination when getting file sizes from file names by regex.
2792 19. Completely re-implemented the support for maildirsize files, in the light
2793 of a number of problems with the previous contributed implementation
2794 (4.30/29). In particular:
2796 . If the quota is zero, the maildirsize file is maintained, but no quota is
2799 . If the maildir directory does not exist, it is created before any attempt
2800 to write a maildirsize file.
2802 . The quota value in the file is just a cache; if the quota is changed in
2803 the transport, the new value overrides.
2805 . A regular expression is available for excluding directories from the
2808 20. The autoreply transport checks the characters in options that define the
2809 message's headers; it allows continued headers, but it was checking with
2810 isspace() after an embedded newline instead of explicitly looking for a
2813 21. If all the "regular" hosts to which an address was routed had passed their
2814 expiry times, and had not reached their retry times, the address was
2815 bounced, even if fallback hosts were defined. Now Exim should go on to try
2818 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2819 equivalent code in the SMTP transport. Some hosts send humungous responses
2820 to HELO/EHLO, more than 1024 it seems.
2822 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2823 gives compiler warnings in some environments. I've now done it "properly",
2826 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2827 (because of problems with the built-in one) was declared to return uschar *
2828 instead of char *, causing compiler failure.
2830 25. Fixed a file descriptor leak when processing alias/forward files.
2832 26. Fixed a minor format string issue in dbfn.c.
2834 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2836 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2837 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2840 29. When a sender address is verified, it is cached, to save repeating the test
2841 when there is more than one recipient in a message. However, when the
2842 verification involves a callout, it is possible for different callout
2843 options to be set for different recipients. It is too complicated to keep
2844 track of this in the cache, so now Exim always runs a verification when a
2845 callout is required, relying on the callout cache for the optimization.
2846 The overhead is duplication of the address routing, but this should not be
2849 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2850 to be verified with callout=postmaster, and the main callout worked but the
2851 postmaster check failed, the verification correctly failed. However, if a
2852 subsequent RCPT command asked for sender verification *without* the
2853 postmaster check, incorrect caching caused this verification also to fail,
2856 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2857 it was not caching the DNS options (qualify_single, search_parents) that
2858 were used when the lookup failed. A subsequent lookup with different
2859 options therefore always gave the same answer, though there were cases
2860 where it should not have. (Example: a "domains = !$mx_any" option on a
2861 dnslookup router: the "domains" option is always processed without any
2862 widening, but the router might have qualify_single set.) Now Exim uses the
2863 cached value only when the same options are set.
2865 32. Added John Jetmore's "exipick" utility to the distribution.
2867 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2868 than a timeout (e.g. a certificate is required, and is not provided), an
2869 Exim server now closes the connection immediately. Previously it waited for
2870 the client to close - but if the client is SSL, it seems that they each
2871 wait for each other, leading to a delay before one of them times out.
2873 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2874 maintained 0.8.x compatibility because I don't think many are using it, and
2875 it is clearly obsolete.
2877 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2880 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2881 client certificate was expired. A simple patch fixes this, though I don't
2882 understand the full logic of why the verify callback is called multiple
2885 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2886 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2887 which causes problems with some clients (such as the Certicom SSL Plus
2888 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2889 disables the coutermeasure allowing Eudora to connect."
2891 38. Exim was not checking that a write() to a log file succeeded. This could
2892 lead to Bad Things if a log got too big, in particular if it hit a file
2893 size limit. Exim now panics and dies if it cannot write to a log file, just
2894 as it does if it cannot open a log file.
2896 39. Modified OS/Makefile-Linux so that it now contains
2898 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2900 The two -D definitions ensure that Exim is compiled with large file
2901 support, which makes it possible to handle log files that are bigger than
2904 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2905 instance) a domain was checked against a named list that involved a lookup,
2906 causing $domain_data to be set, then another domain was checked against the
2907 same list, then the first domain was re-checked, the value of $domain_data
2908 after the final check could be wrong. In particular, if the second check
2909 failed, it could be set empty. This bug probably also applied to
2912 41. The strip_trailing_dot option was not being applied to the address given
2913 with the -f command-line option.
2915 42. The code for reading a message's header from the spool was incrementing
2916 $received_count, but never initializing it. This meant that the value was
2917 incorrect (doubled) while delivering a message in the same process in which
2918 it was received. In the most common configuration of Exim, this never
2919 happens - a fresh exec is done - but it can happen when
2920 deliver_drop_privilege is set.
2922 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2923 it now includes up to 150 characters of the unexpected data in the log
2926 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2927 and building data strings. The size of both of these buffers was 10 000
2928 bytes - far larger than anybody would *ever* want, thought I. Needless to
2929 say, somebody hit the limit. I have increased the maximum line length to
2930 20 000 and the maximum data length of concatenated lines to 100 000. I have
2931 also fixed two bugs, because there was no checking on these buffers. Tsk,
2932 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2933 buffer is too small.
2935 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2936 lsearch lookups. Now it does.
2938 46. When parsing a route_list item in a manualroute router, a fixed-length
2939 buffer was used for the list of hosts. I made this 1024 bytes long,
2940 thinking that nobody would ever have a list of hosts that long. Wrong.
2941 Somebody had a whole pile of complicated expansion conditions, and the
2942 string was silently truncated, leading to an expansion error. It turns out
2943 that it is easier to change to an unlimited length (owing to other changes
2944 that have happened since this code was originally written) than to build
2945 structure for giving a limitation error. The length of the item that
2946 expands into the list of hosts is now unlimited.
2948 47. The lsearch lookup could not handle data where the length of text line was
2949 more than 4095 characters. Such lines were truncated, leading to shortened
2950 data being returned. It should now handle lines of any length.
2952 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2953 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2956 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2957 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2958 no longer happens in comment lines. A list of replacements is now placed
2959 at the head of all of the source files, except those whose only change is
2960 to replace PERL_COMMAND in the very first #! line.
2962 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2963 messages on the queue, with a bottom-up merge sort, using code contributed
2964 by Michael Haardt. This should make operations like -bp somewhat faster on
2965 large queues. It won't affect queue runners, except when queue_run_in_order
2968 51. Installed eximstats 1.31 in the distribution.
2970 52. Added support for SRV lookups to the dnslookup router.
2972 53. If an ACL referred to $message_body or $message_body_end, the value was not
2973 reset for any messages that followed in the same SMTP session.
2975 54. The store-handling optimization for building very long strings was not
2976 differentiating between the different store pools. I don't think this
2977 actually made any difference in practice, but I've tidied it.
2979 55. While running the routers to verify a sender address, $sender_address
2980 was still set to the sender address. This is wrong, because when routing to
2981 send a bounce to the sender, it would be empty. Therefore, I have changed
2982 it so that, while verifying a sender address, $sender_address is set to <>.
2983 (There is no change to what happens when verifying a recipient address.)
2985 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2986 target A or AAAA records (if not already returned) without resetting the
2987 qualify_single or search_parents options of the DNS resolver. These are
2988 inappropriate in this case because the targets of MX and SRV records must
2989 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2990 target that, when qualified, matched something in the local domain. These
2991 two options are now turned off when doing these lookups.
2993 57. It seems that at least some releases of Reiserfs (which does not have the
2994 concept of a fixed number of inodes) returns zero and not -1 for the
2995 number of available inodes. This interacted badly with check_spool_inodes,
2996 which assumed that -1 was the "no such thing" setting. What I have done is
2997 to check that the total number of inodes is greater than zero before doing
2998 the test of how many are available.
3000 58. When a "warn" ACL statement has a log_message modifier, the message is
3001 remembered, and not repeated. This is to avoid a lot of repetition when a
3002 message has many recipients that cause the same warning to be written.
3003 Howewer, Exim was preserving the list of already written lines for an
3004 entire SMTP session, which doesn't seem right. The memory is now reset if a
3005 new message is started.
3007 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3008 debugging output unless log_rewrite was also set.
3010 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3011 of (int)(handle) when we know that handle contains (void *)(-1).
3013 61. The Exim daemon panic-logs an error return when it closes the incoming
3014 connection. However "connection reset by peer" seems to be common, and
3015 isn't really an error worthy of noting specially, so that particular error
3018 62. When Exim is trying to find all the local interfaces, it used to panic and
3019 die if the ioctl to get the interface flags failed. However, it seems that
3020 on at least one OS (Solaris 9) it is possible to have an interface that is
3021 included in the list of interfaces, but for which you get a failure error
3022 for this call. This happens when the interface is not "plumbed" into a
3023 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3024 failure of the "get flags" call assumes that the interface is down.
3026 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3027 makes life easier for people who are doing arithmetic on fields extracted
3028 from dates, where you often get leading zeros that should not be
3029 interpreted as octal.
3031 64. Added qualify_domain to the redirect router, to override the global
3034 65. If a pathologically long header line contained very many addresses (the
3035 report of this problem mentioned 10 000) and each of them was rewritten,
3036 Exim could use up a very large amount of memory. (It kept on making new
3037 copies of the header line as it rewrote, and never released the old ones.)
3038 At the expense of a bit more processing, the header rewriting function has
3039 been changed so that it no longer eats memory in this way.
3041 66. The generation of the Received: header has been moved from the time that a
3042 message starts to be received, to the time that it finishes. The timestamp
3043 in the Received: header should now be very close to that of the <= log
3044 line. There are two side-effects of this change:
3046 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3047 logged header lines no longer include the local Received: line, because
3048 it has not yet been created. The same applies to a copy of the message
3049 that is returned to a non-SMTP sender when a message is rejected.
3051 (b) When a filter file is tested using -bf, no additional Received: header
3052 is added to the test message. After some thought, I decided that this
3055 This change does not affect the value of $received_for. It is still set
3056 after address rewriting, but before local_scan() is called.
3058 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3060 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3061 gave an unhelpful panic error message, and a defer error. I have managed to
3062 change this behaviour so that it now rejects any supplied certificate,
3063 which seems right, as the list of acceptable certificates is empty.
3065 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3066 gave an unhelpful defer error. I have not managed to make this reject any
3067 supplied certificates, but the error message it gives is "no certificate
3068 supplied", which is not helpful.
3070 70. exigrep's output now also includes lines that are not associated with any
3071 message, but which match the given pattern. Implemented by a patch from
3072 Martin Sluka, which also tidied up the Perl a bit.
3074 71. Recipient callout verification, like sender verification, was using <> in
3075 the MAIL FROM command. This isn't really the right thing, since the actual
3076 sender may affect whether the remote host accepts the recipient or not. I
3077 have changed it to use the actual sender in the callout; this means that
3078 the cache record is now keyed on a recipient/sender pair, not just the
3079 recipient address. There doesn't seem to be a real danger of callout loops,
3080 since a callout by the remote host to check the sender would use <>.
3081 [SEE ABOVE: changed after hitting problems.]
3083 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3084 temporary errors. However, in the case of such a code being given after
3085 the end of a data transmission (i.e. after ".") Exim was failing to write
3086 a retry record for the message. (Yes, there was some broken host that was
3087 actually sending 8xx at this point.)
3089 73. An unknown lookup type in a host list could cause Exim to panic-die when
3090 the list was checked. (An example that provoked this was putting <; in the
3091 middle of a list instead of at the start.) If this happened during a DATA
3092 ACL check, a -D file could be left lying around. This kind of configuration
3093 error no longer causes Exim to die; instead it causes a defer errror. The
3094 incident is still logged to the main and panic logs.
3096 74. Buglet left over from Exim 3 conversion. The message "too many messages
3097 in one connection" was written to the rejectlog but not the mainlog, except
3098 when address rewriting (yes!) was being logged.
3100 75. Added write_rejectlog option.
3102 76. When a system filter was run not as root (that is, when system_filter_user
3103 was set), the values of the $n variables were not being returned to the
3104 main process; thus, they were not subsequently available in the $sn
3107 77. Added +return_path_on_delivery log selector.
3109 78. A connection timeout was being treated differently from recipients deferred
3110 when testing hosts_max_try with a message that was older than the host's
3111 retry timeout. (The host should not be counted, thus allowing all hosts to
3112 be tried at least once before bouncing.) This may have been the cause of an
3113 occasionally reported bug whereby a message would remain on the queue
3114 longer than the retry timeout, but would be bounced if a delivery was
3115 forced. I say "may" because I never totally pinned down the problem;
3116 setting up timeout/retry tests is difficult. See also the next item.
3118 79. The ultimate address timeout was not being applied to errors that involved
3119 a combination of host plus message (for example, a timeout on a MAIL
3120 command). When an address resolved to a number of possible hosts, and they
3121 were not all tried for each delivery (e.g. because of hosts_max_try), a
3122 message could remain on the queue longer than the retry timeout.
3124 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3127 81. Fixed an obscure SMTP outgoing bug which required at least the following
3128 conditions: (a) there was another message waiting for the same server;
3129 (b) the server returned 5xx to all RCPT commands in the first message so
3130 that the message was not completed; (c) the server dropped the connection
3131 or gave a negative response to the RSET that Exim sends to abort the
3132 transaction. The observed case was a dropped connection after DATA that had
3133 been sent in pipelining mode. That is, the server had advertised PIPELINING
3134 but was not implementing it correctly. The effect of the bug was incorrect
3135 behaviour, such as trying another host, and this could lead to a crash.
3141 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3142 and daemon.c were passed as pointers to ints; they should have been
3143 pointers to socklen_t variables (which are typically unsigned ints).
3145 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3148 3. Fixed a really odd bug that affected only the testing scheme; patching a
3149 certain fixed string in the binary changed the value of another string that
3150 happened to be identical to the end of the original first string.
3152 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3153 name", it returns that address as the IP address. On some operating
3154 systems (e.g. Solaris), it also passes back the IP address string as the
3155 "host name". However, on others (e.g. Linux), it passes back an empty
3156 string. Exim wasn't checking for this, and was changing the host name to an
3157 empty string, assuming it had been canonicized.
3159 5. Although rare, it is permitted to have more than one PTR record for a given
3160 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3161 all the names associated with an address, because they do in Solaris.
3162 However, it seems that they do not in Linux for data that comes from the
3163 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3164 I found this out when I moved to a new Linux workstation and tried to run
3165 the Exim test suite.
3167 To get round this problem I have changed the code so that it now does its
3168 own call to the DNS to look up PTR records when searching for a host name.
3169 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3170 addresses that are only in /etc/hosts are still found.
3172 This behaviour is, however, controlled by an option called host_lookup_
3173 order, which defaults to "bydns:byaddr". If people want to use the other
3174 order, or indeed, just use one or the other means of lookup, they can
3175 specify it in this variable.
3177 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3178 some operating systems, this comes back from gethostbyaddr() as an empty
3179 string, and this is what Exim used to test for. However, it seems that in
3180 other systems, "." is yielded. Exim now tests for this case too.
3182 7. The values of check_spool_space and check_log_space are now held internally
3183 as a number of kilobytes instead of an absolute number of bytes. If a
3184 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3185 kilobyte. This means that much larger values can be stored.
3187 8. Exim monitor: an attempt to get the action menu when not actually pointing
3188 at a message produces an empty menu entitled "No message selected". This
3189 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3190 no entries in it ("Shell widget menu has zero width and/or height"). So I
3191 have added a single, blank menu entry in this case.
3193 9. Added ${quote_local_part.
3195 10. MIME decoding is now applied to the contents of Subject: header lines when
3198 11. Now that a reference to $sender_host_address automatically causes a reverse
3199 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3200 host lookup before query-style lookups in lists that might use this
3201 variable. This has therefore been abolished, and the "net-" prefix is no
3202 longer necessary for query-style lookups.
3204 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3205 have been a typo for LFLAGS, so it has been changed.
3207 13. The install script calls Exim with "-C /dev/null" in order to find the
3208 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3209 to be output. Howeve, since Exim outputs its version number before the
3210 error, it didn't break the script. It just looked ugly. I fixed this by
3211 always allowing "-C /dev/null" if the caller is root.
3213 14. Ignore overlarge ACL variable number when reading spool file - insurance
3214 against a later release with more variables having written the file.
3216 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3217 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3218 this, as well as the form without the "IPv6" on the front.
3220 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3221 OS/Makefile-Darwin file.
3223 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3224 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3226 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3227 variable. However, some OS use ints for inodes. I've added cast to long int
3228 to get rid of the compiler warning.
3230 19. I had forgotten to lock out "/../" in configuration file names when
3231 ALT_CONFIG_PREFIX was set.
3233 20. Routers used for verification do not need to specify transports. However,
3234 if such a router generated a host list, and callout was configured, Exim
3235 crashed, because it could not find a port number from the (non-existent)
3236 transport. It now assumes port 25 in this circumstance.
3238 21. Added the -t option to exigrep.
3240 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3241 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3242 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3243 them is set, LOOKUP_LSEARCH is forced.
3245 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3247 24. Added sender and host information to the "rejected by local_scan()" log
3248 line; previously there was no indication of these.
3250 25. Added .include_if_exists.
3252 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3253 Linux. It turns out that not all file systems support this. Apparently some
3254 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3255 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3256 not supported on the file descriptor, is now ignored when Exim is trying to
3257 sync a directory. This applies only to Linux.
3259 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3261 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3262 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3265 29. Added support for maildirsize files from supplied patch (modified a bit).
3267 30. The use of :fail: followed by an empty string could lead Exim to respond to
3268 sender verification failures with (e.g.):
3270 550 Verification failed for <xxx>
3271 550 Sender verify failed
3273 where the first response line was missing the '-' that indicates it is not
3274 the final line of the response.
3276 31. The loop for finding the name of the user that called Exim had a hardwired
3277 limit of 10; it now uses the value of finduser_retries, which is used for
3278 all other user lookups.
3280 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3283 33. Exim was neglecting to zero errno before one call of strtol() when
3284 expanding a string and expecting an integer value. On some systems this
3285 resulted in spurious "integer overflow" errors. Also, it was casting the
3286 result into an int without checking.
3288 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3289 did not work. The code looks as if it has *never* worked, though it appears
3290 to have been documented since at least releast 1.62. I have made it work.
3292 35. The "timeout_DNS" error in retry rules, also documented since at least
3293 1.62, also never worked. As it isn't clear exactly what this means, and
3294 clearly it isn't a major issue, I have abolished the feature by treating it
3295 as "timeout", and writing a warning to the main and panic logs.
3297 36. The display of retry rules for -brt wasn't always showing the error code
3300 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3301 timeout_connect_A, timeout_connect_MX.
3303 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3304 to the empty sender.
3306 39. The daemon was not analysing the content of -oX till after it had closed
3307 stderr and disconnected from the controlling terminal. This meant that any
3308 syntax errors were only noted on the panic log, and the return code from
3309 the command was 0. By re-arranging the code a little, I've made the
3310 decoding happen first, so such errors now appear on stderr, and the return
3311 code is 1. However, the actual setting up of the sockets still happens in
3312 the disconnected process, so errors there are still only recorded on the
3315 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3316 connections (as happens on some IP stacks) was logged at start up time as
3317 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3318 it from "IPv6 and IPv4", which means that two separate sockets are being
3321 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3322 says whether AF_INET or AF_INET6 was passed as an argument.
3324 42. Exiwhat output was messed up when time zones were included in log
3327 43. Exiwhat now gives more information about the daemon's listening ports,
3328 and whether -tls-on-connect was used.
3330 44. The "port" option of the smtp transport is now expanded.
3332 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3333 silently ignored. Now an error message is written to the main and panic
3336 46. There's a new ACL modifier called "logwrite" which writes to a log file
3337 as soon as it is encountered.
3339 47. Added $local_user_uid and $local_user_gid at routing time.
3341 48. Exim crashed when trying to verify a sender address that was being
3344 49. Exim was recognizing only a space character after ".include". It now also
3345 recognizes a tab character.
3347 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3348 extracting the relevant information from the specification. The man page no
3349 longer contains scrambled data for the -d option, and I've added a section
3350 at the front about calling Exim under different names.
3352 51. Added "extra_headers" argument to the "mail" command in filter files.
3354 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3357 53. Installed eximstats 1.29.
3359 54. Added transport_filter_timeout as a generic transport option.
3361 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3362 Cc: header lines. This was required by RFC 822, but it not required by RFC
3365 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3366 incoming messages that did not have them. Now it does so only if the
3367 message originates locally, that is, if there is no associated remote host
3368 address. When Resent- header lines are present, this applies to the Resent-
3369 lines rather than the non-Resent- lines.
3371 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3372 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3374 58. Messages for configuration errors now include the name of the main
3375 configuration file - useful now that there may be more than one file in a
3376 list (.included file names were always shown).
3378 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3379 for those rare installations that do not start the daemon as root or run it
3380 setuid root. I've cut out the call to initgroups() if the daemon is not
3383 60. The Exim user and group can now be bound into the binary as text strings
3384 that are looked up at the start of Exim's processing.
3386 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3388 62. Added $mailstore_basename variable.
3390 63. Installed patch to sieve.c from Michael Haardt.
3392 64. When Exim failed to open the panic log after failing to open the main log,
3393 the original message it was trying to log was written to stderr and debug
3394 output, but if they were not available (the usual case in production), it
3395 was lost. Now it is written to syslog before the two lines that record the
3396 failures to open the logs.
3398 65. Users' Exim filters run in subprocesses under the user's uid. It is
3399 possible for a "deliver" command or an alias in a "personal" command to
3400 provoke an address rewrite. If logging of address rewriting is configured,
3401 this fails because the process is not running as root or exim. There may be
3402 a better way of dealing with this, but for the moment (because 4.30 needs
3403 to be released), I have disabled address rewrite logging when running a
3404 filter in a non-root, non-exim process.
3410 1. The buildconfig auxiliary program wasn't quoting the value set for
3411 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3412 not defined. This bug was masked in 4.22 by the effect that was fixed in
3415 2. Some messages that were rejected after a message id was allocated were
3416 shown as "incomplete" by exigrep. It no longer does this for messages that
3417 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3419 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3420 have allow_domain_literals set, the ID did not get logged in the <= line.
3421 Domain literals are now always recognized in Message-ID: header lines.
3423 4. The first argument for a ${extract expansion item is the key name or field
3424 number. Leading and trailing spaces in this item were not being ignored,
3425 causing some misleading effects.
3427 5. When deliver_drop_privilege was set, single queue runner processes started
3428 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3429 same command in the process it spins off) were not dropping privilege.
3431 6. When the daemon running as "exim" started a queue runner, it always
3432 re-executed Exim in the spun-off process. This is a waste of effort when
3433 deliver_drop_privilege is set. The new process now just calls the
3434 queue-runner function directly.
3440 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3443 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3444 ignored. Though the use of -oP was forcing the writing of a pid file, it
3445 was always written to the default place.
3447 3. If the message "no IP address found for host xxxx" is generated during
3448 incoming verification, it is now followed by identification of the incoming
3449 connection (so you can more easily find what provoked it).
3451 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3453 5. Added some features to "harden" Exim a bit more against certain attacks:
3455 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3456 be put in Local/Makefile. This is like the never_users runtime option,
3457 but it cannot be overridden. The default setting is "root".
3459 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3460 prefix string with which any file named in a -C command line option
3463 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3464 is retained for -C and -D only if the caller of Exim is root. Without
3465 it, the exim user may also use -C and -D and retain privilege.
3467 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3468 command line option is disabled.
3470 6. Macro names set by the -D option must start with an upper case letter, just
3471 like macro names defined in the configuration file.
3473 7. Added "dereference=" facility to LDAP.
3475 8. Two instances of the typo "uknown" in the source files are fixed.
3477 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3478 the Configure-Makefile script screwed up while processing it.
3480 10. Incorporated PCRE 4.4.
3482 11. The SMTP synchronization check was not operating right at the start of an
3483 SMTP session. For example, it could not catch a HELO sent before the client
3484 waited for the greeting. There is now a check for outstanding input at the
3485 point when the greeting is written. Because of the duplex, asynchronous
3486 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3487 way, but not yet received, when the check is performed.
3489 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3490 on TCP/IP sockets, because this apparently causes some broken clients to
3493 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3494 unchanged) from the Cygwin maintainer.
3496 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3497 was supported instead of testing for mbx. Effectively a typo.
3499 15. The spa authenticator server code was not checking that the input it
3500 received was valid base64.
3502 16. The debug output line for the "set" modifier in ACLs was not showing the
3503 name of the variable that was being set.
3505 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3507 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3508 lookup on its own. Something else was needed to trigger the lookup. For
3509 example, a match in host_lookup or the need for a host name in a host list.
3510 Now, if $sender_host_name is referenced and the host name has not yet been
3511 looked up, a lookup is performed. If the lookup fails, the variable remains
3512 empty, and $host_lookup_failed is set to "1".
3514 19. Added "eqi" as a case-independent comparison operator.
3516 20. The saslauthd authentication condition could segfault if neither service
3517 nor realm was specified.
3519 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3520 error message that was logged was misleading, and incoming SMTP
3521 connections were dropped. The message is now more accurate, and temporary
3522 errors are given to SMTP connections.
3524 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3525 (except RSET) with a 421 error, until QUIT is received. However, it was
3526 failing to send a response to QUIT.
3528 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3529 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3530 ACL is now run after the helo_try_verify_hosts code.
3532 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3533 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3534 case-independent, so other case variants are also recognized). Apparently
3535 some systems use these upper case variants.
3537 25. If more than two messages were waiting for the same host, and a transport
3538 filter was specified for the transport, Exim sent two messages over the
3539 same TCP/IP connection, and then failed with "socket operation on non-
3540 socket" when it tried to send the third.
3542 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3544 27. The extern definition of crypt16() in expand.c was not being excluded when
3545 the OS had its own crypt16() function.
3547 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3548 as a preferred synonym for return_size_limit, both as an option and as an
3551 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3553 30. Changed the default configuration ACL to relax the local part checking rule
3554 for addresses that are not in any local domains. For these addresses,
3555 slashes and pipe symbols are allowed within local parts, but the sequence
3556 /../ is explicitly forbidden.
3558 31. SPA server authentication was not clearing the challenge buffer before
3561 32. log_message in a "warn" ACL statement was writing to the reject log as
3562 well as to the main log, which contradicts the documentation and doesn't
3563 seem right (because no rejection is happening). So I have stopped it.
3565 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3566 However, I am unable to do any testing of this.
3568 34. Fixed an infelicity in the appendfile transport. When checking directories
3569 for a mailbox, to see if any needed to be created, it was accidentally
3570 using path names with one or more superfluous leading slashes; tracing
3571 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3573 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3574 contents are added to the log line that is written for every discarded
3575 recipient. (Previously a log_message setting was ignored.)
3577 36. The ${quote: operator now quotes the string if it is empty.
3579 37. The install script runs exim in order to find its version number. If for
3580 some reason other than non-existence or emptiness, which it checks, it
3581 could not run './exim', it was installing it with an empty version number,
3582 i.e. as "exim-". This error state is now caught, and the installation is
3585 38. An argument was missing from the function that creates an error message
3586 when Exim fails to connect to the socket for saslauthd authentication.
3587 This could cause Exim to crash, or give a corrupted message.
3589 39. Added isip, isip4, and isip6 to ${if conditions.
3591 40. The ACL variables $acl_xx are now saved with the message, and can be
3592 accessed later in routers, transports, and filters.
3594 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3595 strings in the file are not string-expanded.
3597 42. If a MAIL command specified a SIZE value that was too large to fit into an
3598 int variable, the check against message_size_limit failed. Such values are
3599 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3600 day this will have to be increased, but I don't think I want to be around
3601 when emails are that large.
3608 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3609 iconv() is not standard in FreeBSD.
3611 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3612 IPv6 enabled. The observed symptom was a segmentation fault on return from
3613 the function os_common_find_running_interfaces() in src/os.c.
3615 3. In the check_special_case() function in daemon.c I had used "errno" as an
3616 argument name, which causes warnings on some systems. This was basically a
3617 typo, since it was named "eno" in the comments!
3619 4. The code that waits for the clock to tick (at a resolution of some fraction
3620 of a second) so as to ensure message-id uniqueness was always waiting for
3621 at least one whole tick, when it could have waited for less. [This is
3622 almost certainly not relevant at current processor speeds, where it is
3623 unlikely to ever wait at all. But we try to future-proof.]
3625 5. The function that sleeps for a time interval that includes fractions of a
3626 second contained a race. It did not block SIGALRM between setting the
3627 timer, and suspending (a couple of lines later). If the interval was short
3628 and the sigsuspend() was delayed until after it had expired, the suspension
3629 never ended. On busy systems this could lead to processes getting stuck for
3632 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3633 process, before it forks any delivery processes. The open lookup caching
3634 mechanism meant that the open file or database connection was passed into
3635 the delivery process. The problem was that delivery processes always tidy
3636 up cached lookup data. This could cause a problem for the next delivery
3637 process started by the queue runner, because the external queue runner
3638 process does not know about the closure. So the next delivery process
3639 still has data in the lookup cache. In the case of a file lookup, there was
3640 no problem because closing a file descriptor in a subprocess doesn't affect
3641 the parent. However, if the lookup was caching a connection to a database,
3642 the connection was closed, and the second delivery process was likely to
3643 see errors such as "PGSQL: query failed: server closed the connection
3644 unexpectedly". The problem has been fixed by closing all cached lookups
3645 in a queue runner before running a delivery process.
3647 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3648 seem to have the "const" qualifier which it has on other OS. I've
3651 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3652 *of the same type* (client or server) with the same public name that an
3653 error should be diagnosed.
3655 9. When Exim looked up a host name for an IP address, but failed to find the
3656 original IP address when looking up the host name (a safety check), it
3657 output the message "<ip address> does not match any IP for NULL", which was
3658 confusing, to say the least. The bug was that the host name should have
3659 appeared instead of "NULL".
3661 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3662 user that is built into the binary, and the -C or -D options is used, root
3663 privilege is dropped before the configuration file is read. In addition,
3664 logging is switched to stderr instead of the normal log files. If the
3665 configuration then re-defines the Exim user, the unprivileged environment
3666 is probably not what is expected, so Exim logs a panic warning message (but
3669 However, if deliver_drop_privilege is set, the unprivileged state may well
3670 be exactly what is intended, so the warning has been cut out in that case,
3671 and Exim is allowed to try to write to its normal log files.
3677 1. smtp_return_error_details was not giving details for temporary sender
3678 or receiver verification errors.
3680 2. Diagnose a configuration error if two authenticators have the same public
3683 3. Exim used not to create the message log file for a message until the first
3684 delivery attempt. This could be confusing when incoming messages were held
3685 for policy or load reasons. The message log file is now created at the time
3686 the message is received, and an initial "Received" line is written to it.
3688 4. The automatically generated man page for command line options had a minor
3689 bug that caused no ill effects; however, a more serious problem was that
3690 the procedure for building the man page automatically didn't always
3691 operate. Consequently, release 4.20 contains an out-of-date version. This
3692 shouldn't happen again.
3694 5. When building Exim with embedded Perl support, the script that builds the
3695 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3696 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3698 6. The freeze_tell option was not being used for messages that were frozen on
3699 arrival, either by an ACL or by local_scan().
3701 7. Added the smtp_incomplete_transaction log selector.
3703 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3704 was accepting AUTH without a new EHLO.
3706 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3707 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3710 10. Logging of TCP/IP connections (when configured) now happens in the main
3711 daemon process instead of the child process, so that the TCP/IP connection
3712 count is more accurate (but it can never be perfect).
3714 11. The use of "drop" in a nested ACL was not being handled correctly in the
3715 outer ACL. Now, if condition failure induced by the nested "drop" causes
3716 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3717 or "require"), the connection is dropped.
3719 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3720 that yield "discard" can now be used with an "accept" or a "discard" verb,
3721 but an error is generated for any others (because I can't see a useful way
3722 to define what should happen).
3724 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3725 are now processed in the same way as lines in the Exim configuration file.
3726 In particular, continuation lines are supported.
3728 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3730 15. Added -ti meaning -t -i.
3732 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3733 domains, and warn by logging if others are found.
3735 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3736 returned from GIFCONF when Exim is trying to find the list of interfaces on
3737 a host. The code in os.c has been modified to copy each ifreq to an aligned
3738 structure in all cases.
3740 Also, in some cases, the returned ifreq's were being copied to a 'struct
3741 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3742 means the last couple of bytes of an IPv6 address could be chopped if the
3743 ifreq contained only a normal sockaddr (14 bytes storage).
3745 18. Named domain lists were not supported in the hosts_treat_as_local option.
3746 An entry such as +xxxx was not recognized, and was treated as a literal
3749 19. Ensure that header lines added by a DATA ACL are included in the reject log
3750 if the ACL subsequently rejects the message.
3752 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3753 MD5 (which is deprecated).
3755 21. When testing a filter file using -bf, Exim was writing a message when it
3756 took the sender from a "From " line in the message, but it was not doing so
3757 when it took $return_path from a Return-Path: header line. It now does.
3759 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3760 with a valid header line field name (a series of printing characters
3761 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3763 23. Changed "disc" in the source to "disk" to conform to the documentation and
3764 the book and for uniformity.
3766 24. Ignore Sendmail's -Ooption=value command line item.
3768 25. When execve() failed while trying to run a command in a pipe transport,
3769 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3770 could be confused with a return value of 69 from the command itself. This
3771 has been changed to 127, the value the shell returns if it is asked to run
3772 a non-existent command. The wording for the related log line suggests a
3773 non-existent command as the problem.
3775 26. If received_header_text expands to an empty string, do not add a Received:
3776 header line to the message. (Well, it adds a token one on the spool, but
3777 marks it "old" so that it doesn't get used or transmitted.)
3779 27. Installed eximstats 1.28 (addition of -nt option).
3781 28. There was no check for failure on the call to getsockname() in the daemon
3782 code. This can fail if there is a shortage of resources on the system, with
3783 ENOMEM, for example. A temporary error is now given on failure.
3785 29. Contrary to the C standard, it seems that in some environments, the
3786 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3787 program. Exim now does this explicitly; it affects the formatting of
3788 timestamps using strftime().
3790 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3791 complaints. I've now initialized all the variables, to avoid this.
3793 32. Header lines added by a system filter were not being "seen" during
3794 transport-time rewrites.
3796 33. The info_callback() function passed to OpenSSL is set up with type void
3797 (*)(SSL *, int, int), as described somewhere. However, when calling the
3798 function (actually a macro) that sets it up, the type void(*)() is
3799 expected. I've put in a cast to prevent warnings from picky compilers.
3801 34. If a DNS black list lookup found a CNAME record, but there were no A
3802 records associated with the domain it pointed at, Exim crashed.
3804 35. If a DNS black list lookup returned more than one A record, Exim ignored
3805 all but the first. It now scans all returned addresses if a particular IP
3806 value is being sought. In this situation, the contents of the
3807 $dnslist_value variable are a list of all the addresses, separated by a
3810 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3811 to accept a host name and all its aliases if the forward lookup for any of
3812 them yielded the IP address of the incoming connection. Now it accepts only
3813 those names whose forward lookup yields the correct IP address. Any other
3814 names are discarded. This closes a loophole whereby a rogue DNS
3815 administrator could create reverse DNS records to break through a
3816 wildcarded host restriction in an ACL.
3818 37. If a user filter or a system filter that ran in a subprocess used any of
3819 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3820 the wrong values were passed to the pipe command ($thisaddress had the
3821 value of $0, $0 had the value of $1, etc). This bug was introduced by
3822 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3824 38. Improved the line breaking for long SMTP error messages from ACLs.
3825 Previously, if there was no break point between 40 and 75 characters, Exim
3826 left the rest of the message alone. Two changes have been made: (a) I've
3827 reduced the minimum length to 35 characters; (b) if it can't find a break
3828 point between 35 and 75 characters, it looks ahead and uses the first one
3829 that it finds. This may give the occasional overlong line, but at least the
3830 remaining text gets split now.
3832 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3833 file descriptors might be low, and that setting 1000 would always raise it.
3834 It turns out that in some environments, the limit is already over 1000 and
3835 that lowering it causes trouble. So now Exim takes care not to decrease it.
3837 40. When delivering a message, the value of $return_path is set to $sender_
3838 address at the start of routing (routers may change the value). By an
3839 oversight, this default was not being set up when an address was tested by
3840 -bt or -bv, which affected the outcome if any router or filter referred to
3843 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3844 log only when "message" or "log_message" are set. However, if one of the
3845 conditions was an address verification, or a call to a nested ACL, the
3846 messages generated by the underlying test were being passed through. This
3847 no longer happens. The underlying message is available in $acl_verify_
3848 message for both "message" and "log_message" expansions, so it can be
3849 passed through if needed.
3851 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3852 new expansion $bh_ to give the encoded byte string without charset
3853 translation. Translation happens only if iconv() is available; HAVE_ICONV
3854 indicates this at build time. HEADERS_CHARSET gives the charset to
3855 translate to; headers_charset can change it in the configuration, and
3856 "headers charset" can change it in an individual filter file.
3858 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3859 that creates RFC 2047 encoded "words" labels them as that charset instead
3860 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3861 expansion operator; (ii) when Exim creates a From: line for a local
3862 message; (iii) when a header line is rewritten to include a "phrase" part.
3864 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3865 buggy, causing it to skip the first lines of messages whose message ID
3866 ended in 'D'. This would not have bitten before Exim release 4.14, because
3867 message IDs were unlikely to end in 'D' before then. The effect was to have
3868 incorrect size information for certain domains.
3870 45. #include "config.h" was missing at the start of the crypt16.c module. This
3871 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3874 46. If there was a timeout during a "random" callout check, Exim treated it as
3875 a failure of the random address, and carried on sending RSET and the real
3876 address. If the delay was just some slowness somewhere, the response to the
3877 original RCPT would be taken as a response to RSET and so on, causing
3878 mayhem of various kinds.
3880 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3881 when I implemented it. It didn't allow for the fact that some option values
3882 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3883 the right test for positive values.
3885 48. Domain names in DNS records are case-independent. Exim always looks them up
3886 in lower case. Some resolvers return domain names in exactly the case they
3887 appear in the zone file, that is, they may contain uppercase letters. Not
3888 all resolvers do this - some return always lower case. Exim was treating a
3889 change of case by a resolver as a change of domain, similar to a widening
3890 of a domain abbreviation. This triggered its re-routing code and so it was
3891 trying to route what was effectively the same domain again. This normally
3892 caused routing to fail (because the router wouldn't handle the domain
3893 twice). Now Exim checks for this case specially, and just changes the
3894 casing of the domain that it ultimately uses when it transmits the message
3897 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3900 50. If a filter generated a file delivery with a non-absolute name (possible if
3901 no home directory exists for the router), the forbid_file option was not
3904 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3905 the existing equality matching.
3907 52. Exim was using ints instead of ino_t variables in some places where it was
3908 dealing with inode numbers.
3910 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3911 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3912 called TMPDIR, and if it finds it is different, it changes its value.
3914 54. The smtp_printf() function is now made available to local_scan() so
3915 additional output lines can be written before returning. There is also an
3916 smtp_fflush() function to enable the detection of a dropped connection.
3917 The variables smtp_input and smtp_batched_input are exported to
3920 55. Changed the default runtime configuration: the message "Unknown user"
3921 has been removed from the ACL, and instead placed on the localuser router,
3922 using the cannot_route_message feature. This means that any verification
3923 failures that generate their own messages won't get overridden. Similarly,
3924 the "Unrouteable address" message that was in the ACL for unverifiable
3925 relay addresses has also been removed.
3927 56. Added hosts_avoid_esmtp to the smtp transport.
3929 57. The exicyclog script was not checking for the esoteric option
3930 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3931 will work only if exicyclog is run under the appropriate euid.
3933 58. Following a discussion on the list, the rules by which Exim recognises line
3934 endings on incoming messages have been changed. The -dropcr and drop_cr
3935 options are now no-ops, retained only for backwards compatibility. The
3936 following line terminators are recognized: LF CRLF CR. However, special
3937 processing applies to CR:
3939 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3940 nor a local message in the state where . is a terminator.
3942 (ii) If a bare CR is encountered in a header line, an extra space is added
3943 after the line terminator so as not to end the header. The reasoning
3944 behind this is that bare CRs in header lines are most likely either
3945 to be mistakes, or people trying to play silly games.
3947 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3948 was being incorrectly given as 18 bytes larger than it should have been.
3949 This is a VOB (very old bug).
3951 60. This may never have affected anything current, but just in case it has:
3952 When the local host is found other than at the start of a list of hosts,
3953 the local host, those with the same MX, and any that follow, are discarded.
3954 When the list in question was part of a longer list of hosts, the following
3955 hosts (not currently being processed) were also being discarded. This no
3956 longer happens. I'm not sure if this situation could ever has previously
3959 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3962 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3963 header. This is something that is recommended in a new Internet Draft, and
3964 is something that is documented as being done by Sendmail. There are two
3965 possible values. For messages generated by the autoreply transport, Exim
3968 Auto-Submitted: auto-replied
3970 whereas for all other generated messages (e.g. bounces) it adds
3972 Auto-Submitted: auto-generated
3974 63. The "personal" condition in filters now includes a test for the
3975 Auto-Submitted: header. If it contains the string "auto-" the message it
3976 not considered personal.
3978 64. Added rcpt_include_affixes as a generic transport option.
3980 65. Added queue_only_override (default true).
3982 66. Added the syslog_duplication option.
3984 67. If what should have been the first header line of a message consisted of
3985 a space followed by a colon, Exim was mis-interpreting it as a header line.
3986 It isn't of course - it is syntactically invalid and should therefore be
3987 treated as the start of the message body. The misbehaviour could have
3988 caused a number of strange effects, including loss of data in subsequent
3989 header lines, and spool format errors.
3991 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3992 client host had authenticated. This control can now be exercised by an ACL
3993 for more flexibility.
3995 69. By default, callouts do not happen when testing with -bh. There is now a
3996 variant, -bhc, which does actually run the callout code, including
3997 consulting and updating the callout cache.
3999 70. Added support for saslauthd authentication, courtesy of Alexander
4002 71. If statvfs() failed on the spool or log directories while checking their
4003 size for availability, Exim confusingly gave the error "space shortage".
4004 Furthermore, in debugging mode it crashed with a floating point exception.
4005 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4006 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4007 serious problem, Exim now writes to the main and panic logs when this
4008 happens, with details of the failure. It then refuses to accept the
4009 incoming message, giving the message "spool directory problem" or "log
4010 directory problem" with a 421 code for SMTP messages.
4012 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4013 0, 1, and 2 exist, because some OS complain for execs without them (see
4014 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4015 descriptors. However, the code omitted to check that the open succeeded,
4016 causing mysterious errors if for some reason the permissions on /dev/null
4017 got screwed. Now Exim writes a message to the main and panic logs, and
4018 bombs out if it can't open /dev/null.
4020 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4021 interact so that it is all more flexible. It is supposed to remain
4022 backwards compatible. Also added extra_local_interfaces.
4024 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4025 to bomb out with an assertion failure - to the client this appears as a
4026 connection drop. This problem occurs in the part of the code that was taken
4027 from the Samba project. Fortunately, the assertion is in a very simple
4028 function, so I have fixed this by reproducing the function inline in the
4029 one place where it is called, and arranging for authentication to fail
4030 instead of killing the process with assert().
4032 75. The SPA client code was not working when the server requested OEM rather
4033 than Unicode encoding.
4035 76. Added code to make require_files with a specific uid setting more usable in
4036 the case where statting the file as root fails - usually a non-root-mounted
4037 NFS file system. When this happens and the failure is EACCES, Exim now
4038 forks a subprocess and does the per-uid checking as the relevant uid.
4040 77. Added process_log_path.
4042 78. If log_file_path was not explicitly set, a setting of check_log_space or
4043 check_log_inodes was ignored.
4045 79. If a space check for the spool or log partitions fails, the incident is now
4046 logged. Of course, in the latter case the data may get lost...
4048 80. Added the %p formatting code to string_format() so that it can be used to
4049 print addresses in debug_print(). Adjusted all the address printing in the
4050 debugging in store.c to use %p rather than %d.
4052 81. There was a concern that a line of code in smtp_in.c could overflow a
4053 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4054 initially expressed, the concern was not well-founded, because trailing
4055 spaces are removed early. However, if the trailing spaces were followed by
4056 a NULL, they did not get removed, so the overflow was possible. Two fixes
4059 (a) I re-wrote the offending code in a cleaner fashion.
4060 (b) If an incoming SMTP command contains a NULL character, it is rejected
4063 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4064 runs initgroups(), so that if the Exim user is in any additional groups,
4065 they will be used during message reception.
4071 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob