1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.366 2006/07/03 15:19:44 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.
77 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
78 other effects) broke the use of negated acl sub-conditions.
80 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
83 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
84 "Deny" causes Exim to reject the incoming connection with a 554 error.
85 Unfortunately, if there is a major crisis, such as a disk failure,
86 tcp-wrappers gives "deny", whereas what one would like would be some
87 kind of temporary error. A kludge has been added to help with this.
88 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
89 554 error is used if errno is still zero or contains ENOENT (which occurs
90 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
93 PH/03 Add -lutil to the default FreeBSD LIBS setting.
95 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
96 errors. Otherwise a message that provokes a temporary error (when other
97 messages do not) can cause a whole host to time out.
99 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
100 addresses were routed directly to files or pipes from a redirect router.
101 File deliveries just didn't batch; pipe deliveries might have suffered
104 PH/06 A failure to get a lock for a hints database would erroneously always say
105 "Failed to get write lock", even when it was really a read lock.
107 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
108 of 0600. This has been changed to use the value of the lockfile_mode
109 option (which defaults to 0600).
111 PH/08 Applied small patch from the Sieve maintainer.
113 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
114 folder from quota calculations, a direct delivery into this folder messed
115 up the contents of the maildirsize file. This was because the regex was
116 used only to exclude .Trash (or whatever) when the size of the mailbox
117 was calculated. There was no check that a delivery was happening into an
118 excluded directory. This bug has been fixed by ignoring all quota
119 processing for deliveries into excluded directories.
121 PH/10 Added the maildirfolder_create_regex option to appendfile.
127 PH/01 The code for finding all the local interface addresses on a FreeBSD
128 system running IPv6 was broken. This may well have applied to all BSD
129 systems, as well as to others that have similar system calls. The broken
130 code found IPv4 interfaces correctly, but gave incorrect values for the
131 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
132 that it would not match correctly against @[] and not recognize the IPv6
135 PH/02 The ipliteral router was not recognizing addresses of the form user@
136 [ipv6:....] because it didn't know about the "ipv6:" prefix.
138 PH/03 Added disable_ipv6.
140 PH/04 Changed $reply_address to use the raw form of the headers instead of the
141 decoded form, because it is most often used to construct To: headers
142 lines in autoreplies, and the decoded form may well be syntactically
143 invalid. However, $reply_address has leading white space removed, and all
144 newlines turned into spaces so that the autoreply transport does not
147 PH/05 If group was specified without a user on a router, and no group or user
148 was specified on a transport, the group from the router was ignored.
150 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
151 for visible compile-time settings that can be used to change these
152 numbers, for those that want even more. Backwards compatibility with old
153 spool files has been maintained. However, going back to a previous Exim
154 release will lost any variables that are in spool files.
156 PH/07 Two small changes when running in the test harness: increase delay when
157 passing a TCP/IP connection to a new process, in case the original
158 process has to generate a bounce, and remove special handling of
159 127.0.0.2 (sic), which is no longer necessary.
161 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
162 be the same on different OS.
164 PH/09 Moved a debug statement in filter processing to avoid a race problem when
167 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
168 whether --show-vars was specified or not
170 JJ/02 exipick: Added support for new ACL variable spool format introduced
173 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
174 syntactically invalid From: or Reply-to: line, and a filter used this to
175 generate an autoreply, and therefore failed to obtain an address for the
176 autoreply, Exim could try to deliver to a non-existent relative file
177 name, causing unrelated and misleading errors. What now happens is that
178 it logs this as a hard delivery error, but does not attempt to create a
181 PH/11 The exinext utility has a -C option for testing purposes, but although
182 the given file was scanned by exinext itself; it wasn't being passed on
183 when Exim was called.
185 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
186 an end-of-file indication when reading a command response.
188 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
189 compiled. In many other places in Exim, IPv6 addresses are always
190 recognized, so I have changed this. It also means that IPv4 domain
191 literals of the form [IPV4:n.n.n.n] are now always recognized.
193 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
194 used if the router is not running as root, for example, when verifying at
195 ACL time, or when using -bh. The debugging output from this situation was
196 non-existent - all you got was a failure to exec. I have made two
199 (a) Failures to set uid/gid, the current directory, or a process leader
200 in a subprocess such as that created by queryprogram now generate
201 suitable debugging ouput when -d is set.
203 (b) The queryprogram router detects when it is not running as root,
204 outputs suitable debugging information if -d is set, and then runs
205 the subprocess without attempting to change uid/gid.
207 PH/15 Minor change to Makefile for building test_host (undocumented testing
210 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
211 additional section of a DNS packet that returns MX or SRV records.
212 Instead, it always explicitly searches for A/AAAA records. This avoids
213 major problems that occur when a DNS server includes only records of one
214 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
215 fixed another bug: if SRV records were looked up and the corresponding
216 address records were *not* found in the additional section, the port
217 values from the SRV records were lost.
219 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
220 using the correct key (the original address) when searching the retry
221 rules in order to find which one to use for generating the retry hint.
223 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
224 adding the default one. Similarly, if it contains a Reply-To: header, the
225 errors_reply_to option, if set, is not used.
227 PH/19 When calculating a retry time, Exim used to measure the "time since
228 failure" by looking at the "first failed" field in the retry record. Now
229 it does not use this if it is later than than the arrival time of the
230 message. Instead it uses the arrival time. This makes for better
231 behaviour in cases where some deliveries succeed, thus re-setting the
232 "first failed" field. An example is a quota failure for a huge message
233 when small messages continue to be delivered. Without this change, the
234 "time since failure" will always be short, possible causing more frequent
235 delivery attempts for the huge message than are intended.
236 [Note: This change was subsequently modified - see PH/04 for 4.62.]
238 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
239 $1, $2, $3) because the numerical variables can be reset during some
240 expansion items (e.g. "match"), thereby losing the authentication data.
242 PH/21 Make -bV show the size of off_t variables so that the test suite can
243 decide whether to run tests for quotas > 2G.
245 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
246 mailbox_size, and mailbox_filecount in the appendfile transport. If a
247 filecount value is greater than 2G or if a quota value is greater than 2G
248 on a system where the size of off_t is not greater than 4, a panic error
251 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
252 never match. The debug and -bh output now contains an explicit error
253 message indicating a malformed IPv4 address or mask.
255 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
256 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
259 PH/25 Do not write to syslog when running in the test harness. The only
260 occasion when this arises is a failure to open the main or panic logs
261 (for which there is an explicit test).
263 PH/26 Added the /no_tell option to "control=freeze".
265 PH/27 If a host name lookup failed very early in a connection, for example, if
266 the IP address matched host_lookup and the reverse lookup yielded a name
267 that did not have a forward lookup, an error message of the form "no IP
268 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
269 could be logged. Now it outputs the IP address instead of "NULL".
271 PH/28 An enabling patch from MH: add new function child_open_exim2() which
272 allows the sender and the authenticated sender to be set when
273 submitting a message from within Exim. Since child_open_exim() is
274 documented for local_scan(), the new function should be too.
276 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
277 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
278 results in an empty string is now treated as unset.
280 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
282 PH/31 Added sender_verify_fail logging option.
284 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
285 needed by RFC 822 but not by RFC 2822 was commented out. I have now
286 tidied the source and removed it altogether.
288 PH/33 When a queue run was abandoned because the load average was too high, a
289 log line was always written; now it is written only if the queue_run log
290 selector is set. In addition, the log line for abandonment now contains
291 information about the queue run such as the pid. This is always present
292 in "start" and "stop" lines but was omitted from the "abandon" line.
294 PH/34 Omit spaces between a header name and the colon in the error message that
295 is given when verify = headers_syntax fails (if there are lots of them,
296 the message gets confusing).
298 PH/35 Change the default for dns_check_names_pattern to allow slashes within
299 names, as there are now some PTR records that contain slashes. This check
300 is only to protect against broken name servers that fall over on strange
301 characters, so the fact that it applies to all lookups doesn't matter.
303 PH/36 Now that the new test suite is complete, we can remove some of the
304 special code in Exim that was needed for the old test suite. For example,
305 sorting DNS records because real resolvers return them in an arbitrary
306 order. The new test suite's fake resolver always returns records in the
309 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
310 bounces) except when queue_only is set, to avoid logging races between
311 the different processes.
313 PH/38 Panic-die if .include specifies a non-absolute path.
315 PH/39 A tweak to the "H" retry rule from its user.
317 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
318 a label. They prevented compilation on older perls.
320 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
321 a warning to be raised on newish perls.
323 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
324 on queue. Changes to match documented behaviour of showing count of
325 messages matching specified criteria.
327 PH/40 Changed the default ident timeout from 30s to 5s.
329 PH/41 Added support for the use of login_cap features, on those BSD systems
330 that have them, for controlling the resources used by pipe deliveries.
332 PH/42 The content-scanning code uses fopen() to create files in which to put
333 message data. Previously it was not paying any attention to the mode of
334 the files. Exim runs with umask(0) because the rest of the code creates
335 files with open(), and sets the required mode explicitly. Thus, these
336 files were ending up world-writeable. This was not a big issue, because,
337 being within the spool directory, they were not world-accessible. I have
338 created a function called modefopen, which takes an additional mode
339 argument. It sets umask(777), creates the file, chmods it to the required
340 mode, then resets the umask. All the relevant calls to fopen() in the
341 content scanning code have been changed to use this function.
343 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
344 to 24 hours. This avoids potential overflow problems when processing G
345 and H retry rules. I suspect nobody ever tinkers with this value.
347 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
349 PH/45 When the plaintext authenticator is running as a client, the server's
350 challenges are checked to ensure they are valid base64 strings. By
351 default, the authentication attempt is cancelled if an invalid string is
352 received. Setting client_ignore_invalid_base64 true ignores these errors.
353 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
354 they are received. Thus, the responses can be made to depend on the
355 challenges. If an invalid string is ignored, an empty string is placed in
358 PH/46 Messages that are created by the autoreply transport now contains a
359 References: header, in accordance with RFCs 2822 and 3834.
361 PH/47 Added authenticated_sender_force to the smtp transport.
363 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
365 PH/49 Installed latest patch from the Sieve maintainer.
367 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
368 was also set, the appendfile transport was unnecessarily scanning a
369 directory of message files (e.g. for maildir delivery) to find the count
370 of files (along with the size), even though it did not need this
371 information. It now does the scan only if it needs to find either the
372 size of the count of files.
374 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
376 PH/52 Two bugs concerned with error handling when the smtp transport is
379 (i) Exim was not creating retry information for temporary errors given
380 for individual recipients after the DATA command when the smtp transport
381 was used in LMTP mode. This meant that they could be retried too
382 frequently, and not timed out correctly.
384 (ii) Exim was setting the flag that allows error details to be returned
385 for LMTP errors on RCPT commands, but not for LMTP errors for individual
386 recipients that were returned after the DATA command.
388 PH/53 This is related to PH/52, but is more general: for any failing address,
389 when detailed error information was permitted to be returned to the
390 sender, but the error was temporary, then after the final timeout, only
391 "retry timeout exceeded" was returned. Now it returns the full error as
392 well as "retry timeout exceeded".
394 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
395 do this, and (what is worse) MTAs that accept it.
397 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
398 will now be deprecated.
400 PH/56 New os.c-cygwin from the Cygwin maintainer.
402 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
403 formats (previously only available in exim formats via -bpr, -bpru,
404 and -bpra. Now also available in native and exiqgrep formats)
406 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
407 with very large, slow to parse queues
409 JJ/08 exipick: added ! as generic prefix to negate any criteria format
411 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
413 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
414 responses to authentication challenges, though it was showing the
415 challenges; (ii) I've removed the CR characters from the debug output for
418 PH/58 Allow for the insertion of a newline as well as a space when a string
419 is turned into more than one encoded-word during RFC 2047 encoding. The
420 Sieve code now uses this.
422 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
423 data_4xx, lost_connection, tls_required.
425 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
426 message was being sent as an SMTP response.
428 PH/61 Add -l and -k options to exicyclog.
430 PH/62 When verifying, if an address was redirected to one new address, so that
431 verification continued, and the new address failed or deferred after
432 having set something in $address_data, the value of $address_data was not
433 passed back to the ACL. This was different to the case when no
434 redirection occurred. The value is now passed back in both cases.
436 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
437 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
438 use login_cap.h, so on its own it isn't the distinguishing feature. The
439 new name refers directly to the setclassresources() function.
441 PH/65 Added configuration files for NetBSD3.
443 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
445 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
446 is preferred over IPv4.
448 PH/68 The bounce_return_message and bounce_return_body options were not being
449 honoured for bounces generated during the reception of non-SMTP messages.
450 In particular, this applied to messages rejected by the ACL. This bug has
451 been fixed. However, if bounce_return_message is true and bounce_return_
452 body is false, the headers that are returned for a non-SMTP message
453 include only those that have been read before the error was detected.
454 (In the case of an ACL rejection, they have all been read.)
456 PH/69 The HTML version of the specification is now built in a directory called
457 spec_html instead of spec.html, because the latter looks like a path with
458 a MIME-type, and this confuses some software.
460 PH/70 Catch two compiler warnings in sieve.c.
462 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
463 function verify_get_ident() calls ip_connect() to connect a socket, but
464 if the "connect()" function timed out, ip_connect() used to close the
465 socket. However, verify_get_ident() also closes the socket later, and in
466 between Exim writes to the log, which may get opened at this point. When
467 the socket was closed in ip_connect(), the log could get the same file
468 descriptor number as the socket. This naturally causes chaos. The fix is
469 not to close the socket in ip_connect(); the socket should be closed by
470 the function that creates it. There was only one place in the code where
471 this was missing, in the iplookup router, which I don't think anybody now
472 uses, but I've fixed it anyway.
474 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
475 well as to direct DNS lookups. Otherwise the handling of names in host
476 lists is inconsistent and therefore confusing.
482 PH/01 Two changes to the default runtime configuration:
484 (1) Move the checks for relay_from_hosts and authenticated clients from
485 after to before the (commented out) DNS black list checks.
487 (2) Add control=submission to the relay_from_hosts and authenticated
488 clients checks, on the grounds that messages accepted by these
489 statements are most likely to be submissions.
491 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
493 (1) Generate an error if the third argument for the ${prvs expansion is
496 (2) Treat a missing third argument of ${prvscheck as if it were an empty
499 (3) Reset the variables that are obtained from the first argument of
500 ${prvscheck and used in the second argument before leaving the code,
501 because their memory is reclaimed, so using them afterwards may do
504 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
505 one (it's much easier than Tom thought :-).
507 (5) Because of (4), we can now allow for the use of $prvscheck_result
508 inside the third argument.
510 PH/03 For some reason, the default setting of PATH when running a command from
511 a pipe transport was just "/usr/bin". I have changed it to
514 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
515 anything to be listed in the output from -bV.
517 PH/05 When a filter generated an autoreply, the entire To: header line was
518 quoted in the delivery log line, like this:
520 => >A.N.Other <ano@some.domain> <original@ddress> ...
522 This has been changed so that it extracts the operative address. There
523 may be more than one such address. If so, they are comma-separated, like
526 => >ano@some.domain,ona@other.domain <original@ddress> ...
528 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
529 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
530 not being looked up in the rDNS to get a host name, Exim was showing the
531 IP address twice in Received: lines, even though the IP addresses were
532 identical. For example:
534 Received: from [1.2.3.4] (helo=[1.2.3.4])
536 However, if the real host name was known, it was omitting the HELO data
537 if it matched the actual IP address. This has been tidied up so that it
538 doesn't show the same IP address twice.
540 PH/07 When both +timestamp and +memory debugging was on, the value given by
541 $tod_xxx expansions could be wrong, because the tod_stamp() function was
542 called by the debug printing, thereby overwriting the timestamp buffer.
543 Debugging no longer uses the tod_stamp() function when +timestamp is set.
545 PH/08 When the original message was included in an autoreply transport, it
546 always said "this is a copy of the message, including all the headers",
547 even if body_only or headers_only was set. It now gives an appropriate
550 PH/09 Applied a patch from the Sieve maintainer which:
552 o fixes some comments
553 o adds the (disabled) notify extension core
554 o adds some debug output for the result of if/elsif tests
555 o points to the current vacation draft in the documentation
556 and documents the missing references header update
560 o fixes a bug in processing the envelope test (when testing
561 multiple envelope elements, the last element determinted the
564 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
565 Electronic Mail") by including:
567 Auto-submitted: auto-generated
569 in the messages that it generates (bounce messages and others, such as
570 warnings). In the case of bounce messages for non-SMTP mesages, there was
571 also a typo: it was using "Auto_submitted" (underscore instead of
572 hyphen). Since every message generated by Exim is necessarily in response
573 to another message, thes have all been changed to:
575 Auto-Submitted: auto-replied
577 in accordance with these statements in the RFC:
579 The auto-replied keyword:
581 - SHOULD be used on messages sent in direct response to another
582 message by an automatic process,
584 - MUST NOT be used on manually-generated messages,
586 - MAY be used on Delivery Status Notifications (DSNs) and Message
587 Disposition Notifications (MDNs),
589 - MUST NOT be used on messages generated by automatic or periodic
590 processes, except for messages which are automatic responses to
593 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
594 to the default Received: header definition.
596 PH/12 Added log selector acl_warn_skipped (default on).
598 PH/13 After a successful wildlsearch lookup, discard the values of numeric
599 variables because (a) they are in the wrong storage pool and (b) even if
600 they were copied, it wouldn't work properly because of the caching.
602 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
603 checking when decoding. Apparently there are clients that generate
604 overlong encoded strings. Why am I not surprised?
606 PH/15 If the first argument of "${if match_address" was not empty, but did not
607 contain an "@" character, Exim crashed. Now it writes a panic log message
608 and treats the condition as false.
610 PH/16 In autoreply, treat an empty string for "once" the same as unset.
612 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
613 extension "envelope-auth". The code is finished and in agreement with
614 other implementations, but there is no documentation so far and in fact,
615 nobody wrote the draft yet. This extension is currently #undef'ed, thus
616 not changing the active code.
618 Print executed "if" and "elsif" statements when debugging is used. This
619 helps a great deal to understand what a filter does.
621 Document more things not specified clearly in RFC3028. I had all this
622 sorted out, when out of a sudden new issues came to my mind. Oops."
624 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
627 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
628 canonical form (as documented). However, after a host name lookup from
629 the IP address, check_host() was doing a simple string comparison with
630 addresses acquired from the DNS when checking that the found name did
631 have the original IP as one of its addresses. Since any found IPv6
632 addresses are likely to be in abbreviated form, the comparison could
633 fail. Luckily, there already exists a function for doing the comparison
634 by converting both addresses to binary, so now that is used instead of
637 PH/20 There was another similar case to PH/19, when a complete host name was
638 given in a host list; looking up its IP address could give an abbreviated
639 form, whereas the current host's name might or might not be abbreviated.
640 The same fix has been applied.
646 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
647 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
650 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
651 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
653 PH/03 Typo: missing ".o" in src/pcre/Makefile.
655 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
656 header line, restrict the check to what is listed in RFCs 2369 and 2929.
657 Also, for "Auto-Submitted", treat anything other than "no" as
658 non-personal, in accordance with RFC 3834. (Previously it treated
659 anything starting "auto-" as non-personal.)
661 TF/01 The control=submission/name=... option had a problem with syntax
662 errors if the name included a slash character. The /name= option
663 now slurps the rest of the string, so it can include any characters
664 but it must come last in the list of options (after /sender_retain
667 PH/05 Some modifications to the interface to the fake nameserver for the new
675 TK/01 Added the "success_on_redirect" address verification option. See
676 NewStuff for rationale and an example.
678 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
680 PH/02 Patch to exigrep to allow it to work on syslog lines.
682 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
683 fread() to skip over the body file's header line, because in Cygwin the
684 header line is locked and is inaccessible.
686 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
687 co-exist for some time) to make it clear that it is the Exim ID that is
688 referenced, not the Message-ID: header line.
690 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
691 string_format() function, because snprintf() does not exist on all
694 PH/06 The use of forbid_filter_existstest now also locks out the use of the
695 ${stat: expansion item.
697 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
698 protocol synchronization error", to keep the pedants happy.
700 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
701 well as for IRIX systems, when gcc is being used. See the host.c source
704 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
706 PH/10 Named domain lists were not working if used in a queue_smtp_domains
709 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
710 transport and to the smtp transport in LMTP mode.
712 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
714 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
715 run a filter in a subprocess. This could lead to confusion in subsequent
716 lookups in the parent process. There should also be a search_tidyup() at
717 the end of the subprocess.
719 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
720 only if the host matched helo_try_verify_hosts, which caused the
721 verification to occur when the EHLO/HELO command was issued. The ACL just
722 tested the remembered result. Now, if a previous verification attempt has
723 not happened, "verify = helo" does it there and then.
725 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
727 TK/03 Fix log output including CR from clamd.
729 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
730 exist provoked a memory error which could cause a segfault.
732 PH/15 Installed PCRE 6.2
734 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
736 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
737 of the problem. Specifically, suggested +O2 rather than +O1 for the
740 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
742 PH/20 If a delivery was routed to a non-standard port by means of an SRV
743 record, the port was not correctly logged when the outgoing_port log
744 selector was set (it logged the transort's default port).
746 PH/21 Added support for host-specific ports to manualroute, queryprogram,
747 fallback_hosts, and "hosts" in the smtp transport.
749 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
750 host errors such as "Connection refused".
752 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
753 authentication with radiusclient 0.4.9:
755 - Error returned from rc_read_config was caught wrongly
756 - Username/password not passed on to radius server due to wrong length.
758 The presumption is that some radiusclient API changes for 4.51/PH/17
759 were not taken care of correctly. The code is still untested by me (my
760 Linux distribution still has 0.3.2 of radiusclient), but it was
761 contributed by a Radius user.
763 PH/24 When doing a callout, the value of $domain wasn't set correctly when
764 expanding the "port" option of the smtp transport.
766 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
767 while reading a MIME header. Thanks to Tom Hughes for a patch.
769 PH/24 Include config.h inside local_scan.h so that configuration settings are
772 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
773 that in an ACL for RCPT (for example), you can examine exactly what was
776 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
777 commands, but it was not correctly comparing the address with the actual
778 client host address. Thus, it would show the EHLO address in Received:
779 header lines when this was not necessary.
781 PH/27 Added the % operator to ${eval:}.
783 PH/28 Exim tries to create and chdir to its spool directory when it starts;
784 it should be ignoring failures (because with -C, for example, it has lost
785 privilege). It wasn't ignoring creation failures other than "already
788 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
791 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
792 because an input line was too long, either on its own, or by virtue of
793 too many continuations, the temporary file was not being removed, and the
794 return code was incorrect.
796 PH/31 Missing "BOOL" in function definition in filtertest.c.
798 PH/32 Applied Sieve patches from the maintainer.
800 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
802 PH/33 Added "verify = not_blind".
804 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
805 Local/Makefile (with some defaults set). These are used in built scripts
806 such as exicyclog, but they have never been used in the exim_install
807 script (though there are many overriding facilities there). I have
808 arranged that the exim_install script now takes note of these two
811 PH/35 Installed configuration files for Dragonfly.
813 PH/36 When a locally submitted message by a trusted user did not contain a
814 From: header, and the sender address was obtained from -f or from an SMTP
815 MAIL command, and the trusted user did not use -F to supply a sender
816 name, $originator_name was incorrectly used when constructing a From:
817 header. Furthermore, $originator_name was used for submission mode
818 messages from external hosts without From: headers in a similar way,
819 which is clearly wrong.
821 PH/37 Added control=suppress_local_fixups.
823 PH/38 When log_selector = +received_sender was set, and the addition of the
824 sender made the log line's construction buffer exactly full, or one byte
825 less than full, an overflow happened when the terminating "\n" was
828 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
829 when the result of a list match is failure because a DNS lookup failed.
831 PH/40 RM_COMMAND is now used in the building process.
833 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
834 the "build-* directories that it finds.
836 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
837 address in a domain literal was a prefix of an interface address.
839 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
840 when verifying a sender address, unless rewrite_headers is false.
842 PH/44 Wrote a long comment about why errors_to addresses are verified as
843 recipients, not senders.
845 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
846 the ratelimit ACL was added.
848 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
850 PH/46 Added extra information about PostgreSQL errors to the error string.
852 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
853 avoiding the need to install special zones in a real server. This is
854 backwards compatible; if it can't find the fake resolver, it drops back.
855 Thus, both old and new test suites can be run.
857 TF/02 Added util/ratelimit.pl
859 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
860 clock is set back in time.
862 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
863 Candler <B.Candler@pobox.com>.
865 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
866 OK for addresses that are the result of redirections.
868 PH/48 A number of further additions for the benefit of the new test suite,
869 including a fake gethostbyname() that interfaces to the fake DNS resolver
872 TF/06 The fix for widen_domains has also been applied to qualify_single and
873 search_parents which are the other dnslookup options that can cause
876 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
879 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
881 TF/07 Exim produced the error message "an SRV record indicated no SMTP
882 service" if it encountered an MX record with an empty target hostname.
883 The message is now "an MX or SRV record indicated no SMTP service".
885 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
886 if the DNS of the sending site is misconfigured. This is quite a
887 common situation. This change restores the behaviour of treating a
888 helo verification defer as a failure.
890 PH/51 If self=fail was set on a router, the bounce message did not include the
891 actual error message.
897 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
899 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
900 command itself ended in error, the underlying message about the transport
901 filter timeout was being overwritten with the pipe command error. Now the
902 underlying error message should be appended to the second error message.
904 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
906 PH/02 Reduce the amount of output that "make" produces by default. Full output
907 can still be requested.
909 PH/03 The warning log line about a condition test deferring for a "warn" verb
910 was being output only once per connection, rather than after each
911 occurrence (because it was using the same function as for successful
912 "warn" verbs). This seems wrong, so I have changed it.
914 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
915 it should not have, which might have caused a crash in the right
916 circumstances, but probably never did.
918 PH/04 Installed a modified version of Tony Finch's patch to make submission
919 mode fix the return path as well as the Sender: header line, and to
920 add a /name= option so that you can make the user's friendly name appear
923 TF/03 Added the control = fakedefer ACL modifier.
925 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
926 Mark Lowes for thorough testing.
928 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
930 TK/03 Merged latest SRS patch from Miles Wilton.
932 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
933 with the definition in sysexits.h (which is #included earlier).
934 Fortunately, Exim does not actually use EX_OK. The code used to try to
935 preserve the sysexits.h value, by assumimg that macro definitions were
936 scanned for macro replacements. I have been disabused of this notion,
937 so now the code just undefines EX_OK before #including unistd.h.
939 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
940 in the smtp transport. When a block could not be written in a single
941 write() function, the timeout was being re-applied to each part-write.
942 This seems wrong - if the receiver was accepting one byte at a time it
943 would take for ever. The timeout is now adjusted when this happens. It
944 doesn't have to be particularly precise.
946 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
947 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
949 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
950 without a domain if the check to <postmaster@domain> fails.
952 SC/01 Eximstats: added -xls and the ability to specify output files
953 (patch written by Frank Heydlauf).
955 SC/02 Eximstats: use FileHandles for outputing results.
957 SC/03 Eximstats: allow any combination of xls, txt, and html output.
959 SC/04 Eximstats: fixed display of large numbers with -nvr option
961 SC/05 Eximstats: fixed merging of reports with empty tables.
963 SC/06 Eximstats: added the -include_original_destination flag
965 SC/07 Eximstats: removed tabs and trailing whitespace.
967 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
969 TK/06 MBOX spool code: Add real "From " MBOX separator line
970 so the .eml file is really in mbox format (even though
971 most programs do not really care). Patch from Alex Miller.
973 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
974 The latter is generated from $received_to and is only set if the
975 message has one envelope recipient. SA can use these headers,
976 obviously out-of-the-box. Patch from Alex Miller.
978 PH/08 The ${def test on a variable was returning false if the variable's
979 value was "0", contrary to what the specification has always said!
980 The result should be true unless the variable is empty.
982 PH/09 The syntax error of a character other than { following "${if
983 def:variable_name" (after optional whitespace) was not being diagnosed.
984 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
985 accidental colon was present, for example, could give incorrect results.
987 PH/10 Tidied the code in a number of places where the st_size field of a stat()
988 result is used (not including appendfile, where other changes are about
991 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
992 This involved changing a lot of size variables from int to off_t. It
993 should work with maildirs and everything.
995 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
996 spamd dying while we are connected to it.
998 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1001 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1002 function for BDB 4.3.
1004 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1006 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1007 the sources for PCRE have been re-arranged and more files are now
1010 PH/15 The code I had for printing potentially long long variables in PH/11
1011 above was not the best (it lost precision). The length of off_t variables
1012 is now inspected at build time, and an appropriate printing format (%ld
1013 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1014 to be "long long int" or "long int". This is needed for the internal
1015 formatting function string_vformat().
1017 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1018 the configuration file to be ":syslog", then the script "guesses" where
1019 the logs files are, rather than using the compiled in default. In our
1020 case the guess is not the same as the compiled default, so the script
1021 suddenly stopped working when I started to use syslog. The patch checks
1022 to see if log_file_path is "". If so, it attempts to read it from exim
1023 with no configuration file to get the compiled in version, before it
1024 falls back to the previous guessing code."
1026 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1027 implementing BATV in an Exim configuration. See NewStuff for the gory
1030 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1031 Makefile that are specific to HP-UX.
1033 PH/18 If the "use_postmaster" option was set for a recipient callout together
1034 with the "random" option, the postmaster address was used as the MAIL
1035 FROM address for the random test, but not for the subsequent recipient
1036 test. It is now used for both.
1038 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1039 patch removes a few documentation additions to RFC 3028, because the
1040 latest draft now contains them. It adds the new en;ascii-case comparator
1041 and a new error check for 8bit text in MIME parts. Comparator and
1042 require names are now matched exactly. I enabled the subaddress
1043 extension, but it is not well tested yet (read: it works for me)."
1045 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1046 rework some of the code of TK/09 above to avoid the hardwired use of
1047 "%lld" and "long long". Replaced the call to snprintf() with a call to
1050 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1051 records point to non-existent hosts", "retry timeout exceeded", and
1052 "retry time not reached for any host after a long failure period".
1054 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1055 experimental DomainKeys support:
1057 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1058 (2) On an error such as an illegally used "control", the wrong name for
1059 the control was given.
1061 These problems did NOT occur unless DomainKeys support was compiled.
1063 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1065 PH/24 Added ${if match_ip condition.
1067 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1068 databases so that it will be absolutely obvious if a crash occurs in the
1069 DB library. This is a regular occurrence (often caused by mis-matched
1072 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1073 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1074 picked up on a user's system that detects such things. There doesn't seem
1075 to be a gcc warning option for this - only an attribute that has to be
1076 put on the function's prototype. It seems that in Fedora Core 4 they have
1077 set this on a number of new functions. No doubt there will be more in due
1080 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1081 specify a transport. However, if an address that was verified by such a
1082 router was the subject of a callout, Exim crashed because it tried to
1083 read the rcpt_include_affixes from the non-existent transport. Now it
1084 just assumes that the setting of that option is false. This bug was
1085 introduced by 4.51/PH/31.
1087 PH/28 Changed -d+all to exclude +memory, because that information is very
1088 rarely of interest, but it makes the output a lot bigger. People tend to
1089 do -d+all out of habit.
1091 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1092 code in os-type was giving problems when libc.so lives in lib64, like on
1095 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1096 aren't the modern standard, and it seems that some systems' include files
1097 don't always have them. Exim was already checking for some of the newer
1098 ones like T_AAAA, and defining it itself. I've added checks for all the
1099 record types that Exim uses.
1101 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1102 not automatically generating a new one, as it is supposed to. This
1103 prevented TLS from working. If the file did exist, but contained invalid
1104 data, a new version was generated, as expected. It was only the case of a
1105 non-existent file that was broken.
1107 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1108 with a change in libdomainkeys > 0.64.
1110 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1111 from DNS. If the selector record carries the flag, it now has
1112 precedence over the domain-wide flag.
1114 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1116 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1117 the use of a port name that isn't defined in /etc/services) occurred, the
1118 message was deferred as in a normal delivery, and thus remained on the
1119 spool, instead of being failed because of the mua_wrapper setting. This
1120 is now fixed, and I tidied up some of the mua_wrapper messages at the
1123 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1124 the messages in a hash of arrays rather than using individual hashes.
1125 This is a bit cleaner and results in dramatic memory savings, albeit
1126 at a slight CPU cost.
1128 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1129 as requested by Marc Sherman.
1131 SC/10 Eximstats: added histograms for user specified patterns as requested
1134 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1136 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1137 fopen() in the content-scanning modules that did not already have it.
1143 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1144 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1146 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1148 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1150 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1151 to test Sieve filters that use "vacation".
1153 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1154 that changes the way the GnuTLS parameters are stored in the cache file.
1155 The new format can be generated externally. For backward compatibility,
1156 if the data in the cache doesn't make sense, Exim assumes it has read an
1157 old-format file, and it generates new data and writes a new file. This
1158 means that you can't go back to an older release without removing the
1161 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1162 work if there are any delivery delays because "one_time" forces the
1163 parent to be marked "delivered", so its unseen clone is never tried
1164 again. For this reason, Exim now forbids the simultaneous setting of
1167 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1168 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1169 entry if you want to know the details. The fix, however, neglected to
1170 consider the case when local delivery batching is involved. The test for
1171 "previously delivered" was not happening when checking to see if an
1172 address could be batched with a previous (undelivered) one; under
1173 certain circumstances this could lead to multiple deliveries to the same
1176 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1177 in its include files, and this causes problems building Exim.
1179 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1180 header_syntax) but Exim was just ignoring anything given after a slash.
1181 In particular, this caused confusion with an attempt to use "verify =
1182 reverse_host_lookup/defer_ok". An error is now given when options are
1183 supplied for verify items that do not have them. (Maybe reverse_host_
1184 lookup should have a defer_ok option, but that's a different point.)
1186 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1187 defined by RFC 821) to 2048, because there were problems with some AUTH
1188 commands, and RFC 1869 says the size should be increased for extended
1189 SMTP commands that take arguments.
1191 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1194 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1195 "unknown" error; now it says that the functionality isn't in the binary.
1197 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1198 an address' error message when a string expansion fails (syntax or
1199 whatever). Otherwise the password may appear in the log. Following change
1200 PH/42 below, there is no longer a chance of it appearing in a bounce
1203 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1205 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1206 of its addresses was ever tried. (Bugzilla bug #2.)
1208 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1209 the result incorrectly in the debug output. (It correctly added a newline
1210 to what was transported.)
1212 TF/01 Added $received_time.
1214 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1215 commented out examples of how to interface to a virus scanner and to
1216 SpamAssassin. Also added commented examples of av_scanner and
1217 spamd_address settings.
1219 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1220 and controls are allowed in which ACLs. There were a couple of minor
1221 errors. Some of the entries in the conditions table (which is a table of
1222 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1223 negation of where the condition IS allowed.
1225 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1227 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1228 header file does not have a version number, so I've had to invent a new
1229 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1230 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1231 radiusclient), but it was contributed by a Radius user.
1233 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1234 files or directories, for OpenSSL.
1236 PH/19 When an Exim process that is running as root has to create an Exim log
1237 file, it does so in a subprocess that runs as exim:exim so as to get the
1238 ownership right at creation (otherwise, other Exim processes might see
1239 the file with the wrong ownership). There was no test for failure of this
1240 fork() call, which would lead to the process getting stuck as it waited
1241 for a non-existent subprocess. Forks do occasionally fail when resources
1242 run out. I reviewed all the other calls to fork(); they all seem to check
1245 PH/20 When checking for unexpected SMTP input at connect time (before writing
1246 the banner), Exim was not dealing correctly with a non-positive return
1247 from the read() function. If the client had disconnected by this time,
1248 the result was a log entry for a synchronization error with an empty
1249 string after "input=" when read() returned zero. If read() returned -1
1250 (an event I could not check), uninitialized data bytes were printed.
1251 There were reports of junk text (parts of files, etc) appearing after
1254 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1256 PH/22 Added support for macro redefinition, and (re)definition in between
1257 driver and ACL definitions.
1259 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1260 forgetting to use the resulting value; it was using the unexpanded value.
1262 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1263 hadn't been configured. The fix is from Juergen Kreileder, who
1264 understands it better than I do:
1266 "Here's what I see happening with three configured cyrus_sasl
1267 authenticators configured (plain, login, cram-md5):
1269 On startup auth_cyrus_sasl_init() gets called for each of these.
1270 This means three calls to sasl_listmech() without a specified mech_list.
1271 => SASL tests which mechs of all available mechs actually work
1272 => three warnings about OTP not working
1273 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1275 With the patch, sasl_listmech() also gets called three times. But now
1276 SASL's mech_list option is set to the server_mech specified in the the
1277 authenticator. Or in other words, the answer from sasl_listmech()
1278 gets limited to just the mech you're testing for (which is different
1280 => the return list contains just 'plain' or 'login', 'cram-md5' or
1281 nothing depending on the value of ob->server_mech.
1283 I've just tested the patch: Authentication still works fine,
1284 unavailable mechs specified in the exim configuration are still
1285 caught, and the auth.log warnings about OTP are gone."
1287 PH/25 When debugging is enabled, the contents of the command line are added
1288 to the debugging output, even when log_selector=+arguments is not
1291 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1292 answer is "GNU", and only if the return is "GNU/something" is the answer
1295 PH/27 $acl_verify_message is now set immediately after the failure of a
1296 verification in an ACL, and so is available in subsequent modifiers. In
1297 particular, the message can be preserved by coding like this:
1299 warn !verify = sender
1300 set acl_m0 = $acl_verify_message
1302 Previously, $acl_verify_message was set only while expanding "message"
1303 and "log_message" when a very denied access.
1305 PH/28 Modified OS/os.c-Linux with
1307 -#ifndef OS_LOAD_AVERAGE
1308 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1310 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1311 nomenclature these days.)
1313 PH/29 Installed patch from the Sieve maintainer that adds the options
1314 sieve_useraddress and sieve_subaddress to the redirect router.
1316 PH/30 In these circumstances:
1317 . Two addresses routed to the same list of hosts;
1318 . First host does not offer TLS;
1319 . First host accepts first address;
1320 . First host gives temporary error to second address;
1321 . Second host offers TLS and a TLS session is established;
1322 . Second host accepts second address.
1323 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1324 and peerdn, if requested) that were in fact used only for the second
1327 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1328 attention to any local part prefix or suffix that was matched by the
1329 router that accepted the address. It now behaves in the same way as it
1330 does for delivery: the affixes are removed from the local part unless
1331 rcpt_include_affixes is set on the transport.
1333 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1334 timeout during the DATA phase of an incoming message.
1336 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1337 applied a patch sanctioned by the Sieve maintainer.
1339 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1340 the uid or gid is negative. A case of a negative gid caused this to be
1341 noticed. The fix allows for either to be negative.
1343 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1344 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1347 PH/36 The change PH/12 above was broken. Fixed it.
1349 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1350 the grounds that routing the same address twice would always produce the
1351 same answer. This might have been true once, but it is certainly no
1352 longer true now. Routing a child address may depend on the previous
1353 routing that produced that child. Some complicated redirection strategies
1354 went wrong when messages had multiple recipients, and made Exim's
1355 behaviour dependent on the order in which the addresses were given.
1357 I have moved the duplicate checking until after the routing is complete.
1358 Exim scans the addresses that are assigned to local and remote
1359 transports, and removes any duplicates. This means that more work will be
1360 done, as duplicates will always all be routed, but duplicates are
1361 presumably rare, so I don't expect this is of any significance.
1363 For deliveries to pipes, files, and autoreplies, the duplicate checking
1364 still happens during the routing process, since they are not going to be
1367 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1368 It corrects a timeout issue with spamd. This is Ian's comment: "The
1369 background is that sometimes spamd either never reads data from a
1370 connection it has accepted, or it never writes response data. The exiscan
1371 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1372 blindly assumes that writes won't block so it may never time out."
1374 PH/39 Allow G after quota size as well as K and M.
1376 PH/40 The value set for $authenticated_id in an authenticator may not contain
1377 binary zeroes or newlines because the value is written to log lines and
1378 to spool files. There was no check on this. Now the value is run through
1379 the string_printing() function so that such characters are converted to
1380 printable escape sequences.
1382 PH/41 $message_linecount is a new variable that contains the total number of
1383 lines in the message. Compare $body_linecount, which is the count for the
1386 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1387 bounce and delay warning messages, except in certain special cases, which
1390 (a) An SMTP error message from a remote host;
1391 (b) A message specified in a :fail: redirection;
1392 (c) A message specified in a "fail" command in a system filter;
1393 (d) A message specified in a FAIL return from the queryprogram router;
1394 (e) A message specified by the cannot_route_message router option.
1396 In these cases only, Exim does include the error details in bounce and
1397 warning messages. There are also a few cases where bland messages such
1398 as "unrouteable address" or "local delivery error" are given.
1400 PH/43 $value is now also set for the "else" part of a ${run expansion.
1402 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1403 being worked on, but at least Exim now implements the latest version to
1406 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1407 process to complete was treated as a delivery failure, a timeout while
1408 writing the message to the pipe was logged, but erroneously treated as a
1409 successful delivery. Such timeouts include transport filter timeouts. For
1410 consistency with the overall process timeout, these timeouts are now
1411 treated as errors, giving rise to delivery failures by default. However,
1412 there is now a new Boolean option for the pipe transport called
1413 timeout_defer, which, if set TRUE, converts the failures into defers for
1414 both kinds of timeout. A transport filter timeout is now identified in
1417 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1418 systems where "make" and "gmake" are different, calling "gmake" at top
1419 level broke things. I've arranged for the value of $(MAKE) to be passed
1420 from the Makefile to this script so that it can call the same version of
1424 A note about Exim versions 4.44 and 4.50
1425 ----------------------------------------
1427 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1428 changes of various kinds. As a consequence, a big documentation update was
1429 needed. This delayed the release for rather longer than seemed good, especially
1430 in the light of a couple of (minor) security issues. Therefore, the changes
1431 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1432 release. So 4.44 and 4.50 are in effect two different branches that both start
1435 I have left the 4.50 change log unchanged; it contains all the changes since
1436 4.43. The change log for 4.44 is below; many of its items are identical to
1437 those for 4.50. This seems to be the most sensible way to preserve the
1438 historical information.
1444 1. Minor wording change to the doc/README.SIEVE file.
1446 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1447 computation of the current number of files was incorrect.
1449 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1450 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1453 4. Give more explanation in the error message when the command for a transport
1454 filter fails to execute.
1456 5. There are several places where Exim runs a non-Exim command in a
1457 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1458 was being done only for the command run by the queryprogram router. It is
1459 now done for all such subprocesses. The other cases are: ${run, transport
1460 filters, and the commands run by the lmtp and pipe transports.
1462 6. Added CONFIGURE_GROUP build-time option.
1464 7. Some older OS have a limit of 256 on the maximum number of file
1465 descriptors. Exim was using setrlimit() to set 1000 as a large value
1466 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1467 systems. I've change it so that if it can't get 1000, it tries for 256.
1469 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1470 was an oversight, and furthermore, ever since the addition of extra
1471 controls (e.g. 4.43/32), the checks on when to allow different forms of
1472 "control" were broken. There should now be diagnostics for all cases when a
1473 control that does not make sense is encountered.
1475 9. Added the /retain_sender option to "control=submission".
1477 10. $recipients is now available in the predata ACL (oversight).
1479 11. Tidy the search cache before the fork to do a delivery from a message
1480 received from the command line. Otherwise the child will trigger a lookup
1481 failure and thereby defer the delivery if it tries to use (for example) a
1482 cached ldap connection that the parent has called unbind on.
1484 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1485 of $address_data from the recipient verification was clobbered by the
1486 sender verification.
1488 13. The value of address_data from a sender verification is now available in
1489 $sender_address_data in subsequent conditions in the ACL statement.
1491 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1493 15. Added a new option "connect=<time>" to callout options, to set a different
1496 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1497 was its contents. (It was OK if the option was not defined at all.)
1499 17. A "Completed" log line is now written for messages that are removed from
1500 the spool by the -Mrm option.
1502 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1503 information about exactly what failed.
1505 19. Added -dd to debug only the daemon process.
1507 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1508 handles timeouts, both on the server side and network timeouts. Renamed the
1509 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1511 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1512 It is now set to "smtps".
1514 22. $host_address is now set to the target address during the checking of
1515 ignore_target_hosts.
1517 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1518 being passed; this would have caused $sender_host_name to have been used if
1519 matching the list had actually called for a host name (not very likely,
1520 since this list is usually IP addresses). A host name is now passed as
1523 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1524 code that specifies a non-restarting handler (typically sigaction() in
1525 modern systems) in an attempt to fix a rare and obscure crash bug.
1527 25. Narrowed the window for a race in the daemon that could cause it to ignore
1528 SIGCHLD signals. This is not a major problem, because they are used only to
1529 wake it up if nothing else does.
1531 26. A malformed maildirsize file could cause Exim to calculate negative values
1532 for the mailbox size or file count. Odd effects could occur as a result.
1533 The maildirsize information is now recalculated if the size or filecount
1536 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1537 support for a long time. Removed HAVE_SYS_VFS_H.
1539 28. Installed the latest version of exipick from John Jetmore.
1541 29. In an address list, if the pattern was not a regular expression, an empty
1542 subject address (from a bounce message) matched only if the pattern was an
1543 empty string. Non-empty patterns were not even tested. This was the wrong
1544 because it is perfectly reasonable to use an empty address as part of a
1545 database query. An empty address is now tested by patterns that are
1546 lookups. However, all the other forms of pattern expect the subject to
1547 contain a local part and a domain, and therefore, for them, an empty
1548 address still always fails if the pattern is not itself empty.
1550 30. Exim went into a mad DNS loop when attempting to do a callout where the
1551 host was specified on an smtp transport, and looking it up yielded more
1552 than one IP address.
1554 31. Re-factored the code for checking spool and log partition space into a
1555 function that finds that data and another that does the check. The former
1556 is then used to implement four new variables: $spool_space, $log_space,
1557 $spool_inodes, and $log_inodes.
1559 32. The RFC2047 encoding function was originally intended for short strings
1560 such as real names; it was not keeping to the 75-character limit for
1561 encoded words that the RFC imposes. It now respects the limit, and
1562 generates multiple encoded words if necessary. To be on the safe side, I
1563 have increased the buffer size for the ${rfc2047: expansion operator from
1566 33. It is now permitted to omit both strings after an "if" condition; if the
1567 condition is true, the result is "true". As before, when the second string
1568 is omitted, a false condition yields an empty string. This makes it less
1569 cumbersome to write custom ACL and router conditions.
1571 34. Failure to deliver a bounce message always caused it to be frozen, even if
1572 there was an errors_to setting on the router. The errors_to setting is now
1575 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1576 canonical form (fully expanded) before being placed in
1577 $sender_host_address.
1579 36. The table in the code that translates DNS record types into text (T_A to
1580 "A" for instance) was missing entries for NS and CNAME. It is just possible
1581 that this could have caused confusion if both these types were looked up
1582 for the same domain, because the text type is used as part of Exim's
1583 per-process caching. But the chance of anyone hitting this buglet seems
1586 37. The dnsdb lookup has been extended in a number of ways.
1588 (1) There is a new type, "zns", which walks up the domain tree until it
1589 finds some nameserver records. It should be used with care.
1591 (2) There is a new type, "mxh", which is like "mx" except that it returns
1592 just the host names, not the priorities.
1594 (3) It is now possible to give a list of domains (or IP addresses) to be
1595 looked up. The behaviour when one of the lookups defers can be
1596 controlled by a keyword.
1598 (4) It is now possible to specify the separator character for use when
1599 multiple records are returned.
1601 38. The dnslists ACL condition has been extended: it is now possible to supply
1602 a list of IP addresses and/or domains to be looked up in a particular DNS
1605 39. Added log_selector=+queue_time_overall.
1607 40. When running the queue in the test harness, wait just a tad after forking a
1608 delivery process, to get repeatability of debugging output.
1610 41. Include certificate and key file names in error message when GnuTLS fails
1611 to set them up, because the GnuTLS error message doesn't include the name
1612 of the failing file when there is a problem reading it.
1614 42. Allow both -bf and -bF in the same test run.
1616 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1618 44. The "Exiscan patch" is now merged into the mainline Exim source.
1620 45. Sometimes the final signoff response after QUIT could fail to get
1621 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1622 < 0 before doing a fflush(). This bug looks as though it goes back to the
1623 introduction of TLS in release 3.20, but "sometimes" must have been rare
1624 because the tests only now provoked it.
1626 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1627 (this can affect the format of dates).
1629 47. exim_tidydb, when checking for the continued existence of a message for
1630 which it has found a message-specific retry record, was not finding
1631 messages that were in split spool directories. Consequently, it was
1632 deleting retry records that should have stayed in existence.
1634 48. Steve fixed some bugs in eximstats.
1636 49. The SPA authentication driver was not abandoning authentication and moving
1637 on to the next authenticator when an expansion was forced to fail,
1638 contradicting the general specification for all authenticators. Instead it
1639 was generating a temporary error. It now behaves as specified.
1641 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1642 (the order specifies the preference for clients). The order is now AES256,
1643 AES128, 3DES, ARCFOUR128.
1645 51. Small patch to Sieve code - explicitly set From: when generating an
1648 52. Exim crashed if a remote delivery caused a very long error message to be
1649 recorded - for instance if somebody sent an entire SpamAssassin report back
1650 as a large number of 550 error lines. This bug was coincidentally fixed by
1651 increasing the size of one of Exim's internal buffers (big_buffer) that
1652 happened as part of the Exiscan merge. However, to be on the safe side, I
1653 have made the code more robust (and fixed the comments that describe what
1656 53. Now that there can be additional text after "Completed" in log lines (if
1657 the queue_time_overall log selector is set), a one-byte patch to exigrep
1658 was needed to allow it to recognize "Completed" as not the last thing in
1661 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1662 patch that reportedly fixes this has been added. I am not expert enough to
1663 create a test for it. This is what the patch creator wrote:
1665 "I found a little strange behaviour of ldap code when working with
1666 Windows 2003 AD Domain, where users was placed in more than one
1667 Organization Units. When I tried to give exim partial DN, the exit code
1668 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1669 But simultaneously result of request was absolutely normal ldap result,
1670 so I produce this patch..."
1672 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1673 REFERENCE, so I have modified the code to exclude the patch when that macro
1676 55. Some experimental protocols are using DNS PTR records for new purposes. The
1677 keys for these records are domain names, not reversed IP addresses. The
1678 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1679 leaves it alone. Component reversal etc. now happens only for IP addresses.
1682 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1684 57. Double the size of the debug message buffer (to 2048) so that more of very
1685 long debug lines gets shown.
1687 58. The exicyclog utility now does better if the number of log files to keep
1688 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1690 59. Two changes related to the smtp_active_hostname option:
1692 (1) $smtp_active_hostname is now available as a variable.
1693 (2) The default for smtp_banner uses $smtp_active_hostname instead
1694 of $primary_hostname.
1696 60. The host_aton() function is supposed to be passed a string that is known
1697 to be a valid IP address. However, in the case of IPv6 addresses, it was
1698 not checking this. This is a hostage to fortune. Exim now panics and dies
1699 if the condition is not met. A case was found where this could be provoked
1700 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1701 components; fortuitously, this particular loophole had already been fixed
1702 by change 4.50/55 above.
1704 If there are any other similar loopholes, the new check in host_aton()
1705 itself should stop them being exploited. The report I received stated that
1706 data on the command line could provoke the exploit when Exim was running as
1707 exim, but did not say which command line option was involved. All I could
1708 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1709 running as the user.
1712 61. There was a buffer overflow vulnerability in the SPA authentication code
1713 (which came originally from the Samba project). I have added a test to the
1714 spa_base64_to_bits() function which I hope fixes it.
1717 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1718 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1720 63. The daemon start-up calls getloadavg() while still root for those OS that
1721 need the first call to be done as root, but it missed one case: when
1722 deliver_queue_load_max is set with deliver_drop_privilege. This is
1723 necessary for the benefit of the queue runner, because there is no re-exec
1724 when deliver_drop_privilege is set.
1726 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1727 This has been fixed.
1729 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1730 was in use, was not putting the data itself into the right store pool;
1731 consequently, it could be overwritten for a subsequent message in the same
1732 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1735 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1737 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1738 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1739 the function were treating the return as a boolean value, which happened to
1740 work because 0=false and not-0=true, but is not correct code.
1742 68. The host_aton() function was not handling scoped IPv6 addresses (those
1743 with, for example, "%eth0" on the end) correctly.
1745 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1746 negated items (that is, ~something) in unsigned ints. Some compilers
1747 apparently mutter when there is no cast.
1749 70. If an address verification called from an ACL failed, and did not produce a
1750 user-specific message (i.e. there was only a "system" message), nothing was
1751 put in $acl_verify_message. In this situation, it now puts the system
1754 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1755 session; change 4.31/43 added the unwanted input to the log line - except
1756 that it did not do this in the start of session case. It now does.
1758 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1759 This is wrong and can cause the other end to generate a synchronization
1760 error if it is another Exim or anything else that does the synchronization
1761 check. A QUIT command is no longer sent after a timeout.
1763 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1764 during host lookups.
1766 74. The defer_ok option of callout verification was not working if it was used
1767 when verifying addresses in header lines, that is, for this case:
1769 verify = header_sender/callout=defer_ok
1771 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1772 those file descriptors could be used for SMTP connections. If anything
1773 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1774 could be sent to the SMTP client, causing chaos. The daemon now opens
1775 stdin, stdout, and stderr to /dev/null when it puts itself into the
1778 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1779 log by default. The user can override this with suitable Perl magic.
1781 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1782 the log message when discard triggers, was not working for the DATA ACL or
1783 for the non-SMTP ACL.
1785 78. Error message wording change in sieve.c.
1787 79. If smtp_accept_max_per_host was set, the number of connections could be
1788 restricted to fewer than expected, because the daemon was trying to set up
1789 a new connection before checking whether the processes handling previous
1790 connections had finished. The check for completed processes is now done
1791 earlier. On busy systems, this bug wouldn't be noticed because something
1792 else would have woken the daemon, and it would have reaped the completed
1795 80. If a message was submitted locally by a user whose login name contained one
1796 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1797 It caused a spool format error. I have fixed the spool reading code. A
1798 related problem was that the "from" clause in the Received: line became
1799 illegal because of the space(s). It is now covered by ${quote_local_part.
1801 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1804 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1806 83. Patch from Sieve maintainer to fix unterminated string problem in
1807 "vacation" handling.
1809 84. Some minor changes to the Linux configuration files to help with other
1810 OS variants using glibc.
1812 85. One more patch for Sieve to update vacation handling to latest spec.
1815 ----------------------------------------------------
1816 See the note above about the 4.44 and 4.50 releases.
1817 ----------------------------------------------------
1823 1. Change 4.43/35 introduced a bug that caused file counts to be
1824 incorrectly computed when quota_filecount was set in an appendfile
1827 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1828 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1831 3. Give more explanation in the error message when the command for a transport
1832 filter fails to execute.
1834 4. There are several places where Exim runs a non-Exim command in a
1835 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1836 was being done only for the command run by the queryprogram router. It is
1837 now done for all such subprocesses. The other cases are: ${run, transport
1838 filters, and the commands run by the lmtp and pipe transports.
1840 5. Some older OS have a limit of 256 on the maximum number of file
1841 descriptors. Exim was using setrlimit() to set 1000 as a large value
1842 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1843 systems. I've change it so that if it can't get 1000, it tries for 256.
1845 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1846 was an oversight, and furthermore, ever since the addition of extra
1847 controls (e.g. 4.43/32), the checks on when to allow different forms of
1848 "control" were broken. There should now be diagnostics for all cases when a
1849 control that does not make sense is encountered.
1851 7. $recipients is now available in the predata ACL (oversight).
1853 8. Tidy the search cache before the fork to do a delivery from a message
1854 received from the command line. Otherwise the child will trigger a lookup
1855 failure and thereby defer the delivery if it tries to use (for example) a
1856 cached ldap connection that the parent has called unbind on.
1858 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1859 of $address_data from the recipient verification was clobbered by the
1860 sender verification.
1862 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1863 was its contents. (It was OK if the option was not defined at all.)
1865 11. A "Completed" log line is now written for messages that are removed from
1866 the spool by the -Mrm option.
1868 12. $host_address is now set to the target address during the checking of
1869 ignore_target_hosts.
1871 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1872 being passed; this would have caused $sender_host_name to have been used if
1873 matching the list had actually called for a host name (not very likely,
1874 since this list is usually IP addresses). A host name is now passed as
1877 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1878 code that specifies a non-restarting handler (typically sigaction() in
1879 modern systems) in an attempt to fix a rare and obscure crash bug.
1881 15. Narrowed the window for a race in the daemon that could cause it to ignore
1882 SIGCHLD signals. This is not a major problem, because they are used only to
1883 wake it up if nothing else does.
1885 16. A malformed maildirsize file could cause Exim to calculate negative values
1886 for the mailbox size or file count. Odd effects could occur as a result.
1887 The maildirsize information is now recalculated if the size or filecount
1890 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1891 support for a long time. Removed HAVE_SYS_VFS_H.
1893 18. Updated exipick to current release from John Jetmore.
1895 19. Allow an empty sender to be matched against a lookup in an address list.
1896 Previously the only cases considered were a regular expression, or an
1899 20. Exim went into a mad DNS lookup loop when doing a callout where the
1900 host was specified on the transport, if the DNS lookup yielded more than
1903 21. The RFC2047 encoding function was originally intended for short strings
1904 such as real names; it was not keeping to the 75-character limit for
1905 encoded words that the RFC imposes. It now respects the limit, and
1906 generates multiple encoded words if necessary. To be on the safe side, I
1907 have increased the buffer size for the ${rfc2047: expansion operator from
1910 22. Failure to deliver a bounce message always caused it to be frozen, even if
1911 there was an errors_to setting on the router. The errors_to setting is now
1914 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1915 canonical form (fully expanded) before being placed in
1916 $sender_host_address.
1918 24. Updated eximstats to version 1.33
1920 25. Include certificate and key file names in error message when GnuTLS fails
1921 to set them up, because the GnuTLS error message doesn't include the name
1922 of the failing file when there is a problem reading it.
1924 26. Expand error message when OpenSSL has problems setting up cert/key files.
1927 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1928 (this can affect the format of dates).
1930 28. exim_tidydb, when checking for the continued existence of a message for
1931 which it has found a message-specific retry record, was not finding
1932 messages that were in split spool directories. Consequently, it was
1933 deleting retry records that should have stayed in existence.
1935 29. eximstats updated to version 1.35
1936 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1937 1.35 - bugfix such that pie charts by volume are generated correctly
1939 30. The SPA authentication driver was not abandoning authentication and moving
1940 on to the next authenticator when an expansion was forced to fail,
1941 contradicting the general specification for all authenticators. Instead it
1942 was generating a temporary error. It now behaves as specified.
1944 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1945 (the order specifies the preference for clients). The order is now AES256,
1946 AES128, 3DES, ARCFOUR128.
1948 31. Small patch to Sieve code - explicitly set From: when generating an
1951 32. Exim crashed if a remote delivery caused a very long error message to be
1952 recorded - for instance if somebody sent an entire SpamAssassin report back
1953 as a large number of 550 error lines. This bug was coincidentally fixed by
1954 increasing the size of one of Exim's internal buffers (big_buffer) that
1955 happened as part of the Exiscan merge. However, to be on the safe side, I
1956 have made the code more robust (and fixed the comments that describe what
1959 33. Some experimental protocols are using DNS PTR records for new purposes. The
1960 keys for these records are domain names, not reversed IP addresses. The
1961 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1962 leaves it alone. Component reversal etc. now happens only for IP addresses.
1965 34. The host_aton() function is supposed to be passed a string that is known
1966 to be a valid IP address. However, in the case of IPv6 addresses, it was
1967 not checking this. This is a hostage to fortune. Exim now panics and dies
1968 if the condition is not met. A case was found where this could be provoked
1969 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1970 components; fortuitously, this particular loophole had already been fixed
1971 by change 4.50/55 or 4.44/33 above.
1973 If there are any other similar loopholes, the new check in host_aton()
1974 itself should stop them being exploited. The report I received stated that
1975 data on the command line could provoke the exploit when Exim was running as
1976 exim, but did not say which command line option was involved. All I could
1977 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1978 running as the user.
1981 35. There was a buffer overflow vulnerability in the SPA authentication code
1982 (which came originally from the Samba project). I have added a test to the
1983 spa_base64_to_bits() function which I hope fixes it.
1986 36. The daemon start-up calls getloadavg() while still root for those OS that
1987 need the first call to be done as root, but it missed one case: when
1988 deliver_queue_load_max is set with deliver_drop_privilege. This is
1989 necessary for the benefit of the queue runner, because there is no re-exec
1990 when deliver_drop_privilege is set.
1992 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1993 was in use, was not putting the data itself into the right store pool;
1994 consequently, it could be overwritten for a subsequent message in the same
1995 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1998 38. Sometimes the final signoff response after QUIT could fail to get
1999 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2000 < 0 before doing a fflush(). This bug looks as though it goes back to the
2001 introduction of TLS in release 3.20, but "sometimes" must have been rare
2002 because the tests only now provoked it.
2008 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2009 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2010 Now it returns an integer. A number of calls were still expecting a T/F
2011 return. Fortuitously, in all cases, the tests worked in OK situations,
2012 which is the norm. However, things would have gone wrong on any write
2013 failures on the smtp file descriptor. This function is used when sending
2014 messages over SMTP and also when doing verify callouts.
2016 2. When Exim is called to do synchronous delivery of a locally submitted
2017 message (the -odf or -odi options), it no longer closes stderr before doing
2020 3. Implemented the mua_wrapper option.
2022 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2024 5. Implemented the functions header_remove(), header_testname(),
2025 header_add_at_position(), and receive_remove_recipient(), and exported them
2028 6. If an ACL "warn" statement specified the addition of headers, Exim already
2029 inserted X-ACL-Warn: at the start if there was no header name. However, it
2030 was not making this test for the second and subsequent header lines if
2031 there were newlines in the string. This meant that an invalid header could
2032 be inserted if Exim was badly configured.
2034 7. Allow an ACL "warn" statement to add header lines at the start or after all
2035 the Received: headers, as well as at the end.
2037 8. Added the rcpt_4xx retry error code.
2039 9. Added postmaster_mailfrom=xxx to callout verification option.
2041 10. Added mailfrom=xxxx to the callout verification option, for verify=
2044 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2045 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2047 12. Inserted some casts to stop certain compilers warning when using pointer
2048 differences as field lengths or precisions in printf-type calls (mostly
2049 affecting debugging statements).
2051 13. Added optional readline() support for -be (dynamically loaded).
2053 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2054 same clock tick as a message's arrival, so that its received time was the
2055 same as the "first fail" time on the retry record, and that message
2056 remained on the queue past the ultimate address timeout, every queue runner
2057 would try a delivery (because it was past the ultimate address timeout) but
2058 after another failure, the ultimate address timeout, which should have then
2059 bounced the address, did not kick in. This was a "< instead of <=" error;
2060 in most cases the first failure would have been in the next clock tick
2061 after the received time, and all would be well.
2063 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2064 being recognized when the domain list was tested by the match_domain
2065 condition in an expansion string.
2067 16. Added the ${str2b64: operator.
2069 17. Exim was always calling setrlimit() to set a large limit for the number of
2070 processes, without checking whether the existing limit was already
2071 adequate. (It did check for the limit on file descriptors.) Furthermore,
2072 errors from getrlimit() and setrlimit() were being ignored. Now they are
2073 logged to the main and panic logs, but Exim does carry on, to try to do its
2074 job under whatever limits there are.
2076 18. Imported PCRE 5.0.
2078 19. Trivial typo in log message " temporarily refused connection" (the leading
2081 20. If the log selector return_path_on_delivery was set and an address was
2082 redirected to /dev/null, the delivery process crashed because it assumed
2083 that a return path would always be set for a "successful" delivery. In this
2084 case, the whole delivery is bypassed as an optimization, and therefore no
2087 21. Internal re-arrangement: the function for sending a challenge and reading
2088 a response while authentication was assuming a zero-terminated challenge
2089 string. It's now changed to take a pointer and a length, to allow for
2090 binary data in such strings.
2092 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2094 23. Exim was not respecting finduser_retries when seeking the login of the
2095 uid under which it was called; it was always trying 10 times. (The default
2096 setting of finduser_retries is zero.) Also, it was sleeping after the final
2097 failure, which is pointless.
2099 24. Implemented tls_on_connect_ports.
2101 25. Implemented acl_smtp_predata.
2103 26. If the domain in control=submission is set empty, Exim assumes that the
2104 authenticated id is a complete email address when it generates From: or
2105 Sender: header lines.
2107 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2108 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2109 chown and chgrp in /bin and hostname in /usr/bin.
2111 28. Exim was keeping the "process log" file open after each use, just as it
2112 does for the main log. This opens the possibility of it remaining open for
2113 long periods when the USR1 signal hits a daemon. Occasional processlog
2114 errors were reported, that could have been caused by this. Anyway, it seems
2115 much more sensible not to leave this file open at all, so that is what now
2118 29. The long-running daemon process does not normally write to the log once it
2119 has entered its main loop, and it closes the log before doing so. This is
2120 so that log files can straightforwardly be renamed and moved. However,
2121 there are a couple of unusual error situations where the daemon does write
2122 log entries, and I had neglected to close the log afterwards.
2124 30. The text of an SMTP error response that was received during a remote
2125 delivery was being truncated at 512 bytes. This is too short for some of
2126 the long messages that one sometimes sees. I've increased the limit to
2129 31. It is now possible to make retry rules that apply only when a message has a
2130 specific sender, in particular, an empty sender.
2132 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2133 it possible to be selective about when SMTP synchronization is enforced.
2135 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2137 32. Implemented hosts_connection_nolog.
2139 33. Added an ACL for QUIT.
2141 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2144 35. Added mailbox_size and mailbox_filecount to appendfile.
2146 36. Added control = no_multiline_responses to ACLs.
2148 37. There was a bug in the logic of the code that waits for the clock to tick
2149 in the case where the clock went backwards by a substantial amount such
2150 that the microsecond fraction of "now" was more than the microsecond
2151 fraction of "then" (but the whole seconds number was less).
2153 38. Added support for the libradius Radius client library this is found on
2154 FreeBSD (previously only the radiusclient library was supported).
2160 1. When certain lookups returned multiple values in the form name=value, the
2161 quoting of the values was not always being done properly. Specifically:
2162 (a) If the value started with a double quote, but contained no whitespace,
2164 (b) If the value contained whitespace other than a space character (i.e.
2165 tabs or newlines or carriage returns) it was not quoted.
2166 This fix has been applied to the mysql and pgsql lookups by writing a
2167 separate quoting function and calling it from the lookup code. The fix
2168 should probably also be applied to nisplus, ibase and oracle lookups, but
2169 since I cannot test any of those, I have not disturbed their existing code.
2171 2. A hit in the callout cache for a specific address caused a log line with no
2172 reason for rejecting RCPT. Now it says "Previous (cached) callout
2173 verification failure".
2175 3. There was an off-by-one bug in the queryprogram router. An over-long
2176 return line was truncated at 256 instead of 255 characters, thereby
2177 overflowing its buffer with the terminating zero. As well as fixing this, I
2178 have increased the buffer size to 1024 (and made a note to document this).
2180 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2181 when Exim is waiting for an SMTP response from a remote server, Exim
2182 restarts its select() call on the socket, thereby resetting its timeout.
2183 This is not a problem when such interrupts are rare. Somebody set up a cron
2184 job to run exiwhat every 2 minutes, which is less than the normal select()
2185 timeout (5 or 10 minutes). This meant that the select() timeout never
2186 kicked in because it was always reset. I have fixed this by comparing the
2187 time when an interrupt arrives with the time at the start of the first call
2188 to select(). If more time than the timeout has elapsed, the interrupt is
2189 treated as a timeout.
2191 5. Some internal re-factoring in preparation for the addition of Sieve
2192 extensions (by MH). In particular, the "personal" test is moved to a
2193 separate function, and given an option for scanning Cc: and Bcc: (which is
2194 not set for Exim filters).
2196 6. When Exim created an email address using the login of the caller as the
2197 local part (e.g. when creating a From: or Sender: header line), it was not
2198 quoting the local part when it contained special characters such as @.
2200 7. Installed new OpenBSD configuration files.
2202 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2203 try to make them clearer.
2205 9. Callout options, other than the timeout value, were being ignored when
2206 verifying sender addresses in header lines. For example, when using
2208 verify = header_sender/callout=no_cache
2210 the cache was (incorrectly) being used.
2212 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2213 only the Cygwin environment.
2215 11. When return_path_on_delivery was set as a log selector, if different remote
2216 addresses in the same message used different return paths and parallel
2217 remote delivery occurred, the wrong values would sometimes be logged.
2218 (Whenever a remote delivery process finished, the return path value from
2219 the most recently started remote delivery process was logged.)
2221 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2222 lines when AUTH and/or TLS are in use. This is the "received protocol"
2223 field. Exim used to use "asmtp" for authenticated SMTP, without any
2224 indication (in the protocol name) for TLS use. Now it follows the RFC and
2225 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2226 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2227 names appear in log lines as well as in Received: header lines.
2229 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2230 extensions, and comparison tests, and to fix some bugs.
2232 14. Changes to the "personal" filter test:
2234 (1) The test was buggy in that it was just doing the equivalent of
2235 "contains" tests on header lines. For example, if a user's address was
2236 anne@some.where, the "personal" test would incorrectly be true for
2238 To: susanne@some.where
2240 This test is now done by extracting each address from the header in turn,
2241 and checking the entire address. Other tests that are part of "personal"
2242 are now done using regular expressions (for example, to check local parts
2243 of addresses in From: header lines).
2245 (2) The list of non-personal local parts in From: addresses has been
2246 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2247 taken from the Sieve specification recommendations.
2249 (3) If the message contains any header line starting with "List-" it is
2250 treated as non-personal.
2252 (4) The test for "circular" in the Subject: header line has been removed
2253 because it now seems ill-conceived.
2255 15. Minor typos in src/EDITME comments corrected.
2257 16. Installed latest exipick from John Jetmore.
2259 17. If headers_add on a router specified a text string that was too long for
2260 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2261 of string_sprintf() is now avoided.
2263 18. $message_body_size was not set (it was always zero) when running the DATA
2264 ACL and the local_scan() function.
2266 19. For the "mail" command in an Exim filter, no default was being set for
2267 the once_repeat time, causing a random time value to be used if "once" was
2268 specified. (If the value happened to be <= 0, no repeat happened.) The
2269 default is now 0s, meaning "never repeat". The "vacation" command was OK
2270 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2271 (I found it when inspecting the code).
2273 20. There is now an overall timeout for performing a callout verification. It
2274 defaults to 4 times the callout timeout, which applies to individual SMTP
2275 commands during the callout. The overall timeout applies when there is more
2276 than one host that can be tried. The timeout is checked before trying the
2277 next host. This prevents very long delays if there are a large number of
2278 hosts and all are timing out (e.g. when the network connections are timing
2279 out). The value of the overall timeout can be changed by specifying an
2280 additional sub-option for "callout", called "maxwait". For example:
2282 verify = sender/callout=5s,maxwait=20s
2284 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2285 to the end before writing, but this should make it even safer).
2287 22. Exim was forgetting that it had advertised PIPELINING for the second and
2288 subsequent messages on an SMTP connection. It was also not resetting its
2289 memory on STARTTLS and an internal HELO.
2291 23. When Exim logs an SMTP synchronization error within a session, it now
2292 records whether PIPELINING has been advertised or not.
2294 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2295 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2296 rather than long int.
2298 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2300 26. Added the never_mail option to autoreply.
2306 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2307 crash if the getsockname() call failed; this can happen if a connection is
2308 closed very soon after it is established. The problem was simply in the
2309 order in which certain operations were done, causing Exim to try to write
2310 to the SMTP stream before it had set up the file descriptor. The bug has
2311 been fixed by making things happen in the correct order.
2317 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2318 before the connection was closed, thus losing the rejection response.
2320 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2321 some early Solaris releases, but causes trouble in current releases where
2322 socklen_t is defined.
2324 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2327 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2330 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2331 that releasing the top of it at the end releases what was used for sub-
2332 expansions (unless the block got too big). However, discard this block if
2333 the first thing is a variable or header, so that we can use its block when
2334 it is dynamic (useful for very large $message_headers, for example).
2336 6. Lookups now cache *every* query, not just the most recent. A new, separate
2337 store pool is used for this. It can be recovered when all lookup caches are
2338 flushed. Lookups now release memory at the end of their result strings.
2339 This has involved some general refactoring of the lookup sources.
2341 7. Some code has been added to the store_xxx() functions to reduce the amount
2342 of flapping under certain conditions.
2344 8. log_incoming_interface used to affect only the <= reception log lines. Now
2345 it causes the local interface and port to be added to several more SMTP log
2346 lines, for example "SMTP connection from", and rejection lines.
2348 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2350 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2352 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2353 could be overwritten at the end of the current message (or the start of a
2354 new message if it was set in a HELO ACL). The value is now preserved for
2355 the duration of the SMTP connection.
2357 12. If a transport had a headers_rewrite setting, and a matching header line
2358 contained an unqualified address, that address was qualified, even if it
2359 did not match any rewriting rules. The underlying bug was that the values
2360 of the flags that permit the existence of unqualified sender and recipient
2361 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2362 non-local messages, and by -bnq for local messages) were not being
2363 preserved with the message after it was received.
2365 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2366 "next input=" as part of the text comprising the host identity instead of
2367 the correct text. The code was using the same buffer for two different
2368 strings. However, depending on which order the printing function evaluated
2369 its arguments, the bug did not always show up. Under Linux, for example, my
2370 test suite worked just fine.
2372 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2373 doesn't work with some older versions of Perl. It has been changed to "my",
2374 which in any case is probably the better facility to use.
2376 15. A really picky compiler found some instances of statements for creating
2377 error messages that either had too many or two few arguments for the format
2380 16. The size of the buffer for calls to the DNS resolver has been increased
2381 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2382 for addresses that have a lot of PTR records. This alleviates a problem; it
2383 does not fully solve it.
2385 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2386 buffer now truncates the list and logs the incident, which is the same
2387 action as happens when Exim is looking up a host name and its aliases.
2388 Previously in this situation something unpredictable would happen;
2389 sometimes it was "internal error: store_reset failed".
2391 18. If a server dropped the connection unexpectedly when an Exim client was
2392 using GnuTLS and trying to read a response, the client delivery process
2393 crashed while trying to generate an error log message.
2395 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2396 string, for example:
2398 warn message = H1: something\nH2: something
2400 the text was added as a single header line from Exim's point of view
2401 though it ended up OK in the delivered message. However, searching for the
2402 second and subsequent header lines using $h_h2: did not work. This has been
2403 fixed. Similarly, if a system filter added multiple headers in this way,
2404 the routers could not see them.
2406 20. Expanded the error message when iplsearch is called with an invalid key to
2407 suggest using net-iplsearch in a host list.
2409 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2410 ACLs are no longer actually imposed (and a message to that effect is
2413 22. If a "gecos" field in a passwd entry contained escaped characters, in
2414 particular, if it contained a \" sequence, Exim got it wrong when building
2415 a From: or a Sender: header from that name. A second bug also caused
2416 incorrect handling when an unquoted " was present following a character
2417 that needed quoting.
2419 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2420 was not being matched caselessly.
2422 24. Arranged for all hyphens in the exim.8 source to be escaped with
2425 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2426 properly. Recipient callout cache records were still being keyed to include
2427 the sender, even when use_sender was set false. This led to far more
2428 callouts that were necessary. The sender is no longer included in the key
2429 when use_sender is false.
2431 26. Added "control = submission" modifier to ACLs.
2433 27. Added the ${base62d: operator to decode base 62 numbers.
2435 28. dnsdb lookups can now access SRV records.
2437 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2438 the configuration file.
2440 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2441 (-v) mode. This makes the output for a verbose queue run more intelligible.
2443 31. Added a use_postmaster feature to recipient callouts.
2445 32. Added the $body_zerocount variable, containing the number of binary zero
2446 bytes in the message body.
2448 33. The time of last modification of the "new" subdirectory is now used as the
2449 "mailbox time last read" when there is a quota error for a maildir
2452 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2454 35. Added +ignore_unknown as a special item in host lists.
2456 36. Code for decoding IPv6 addresses in host lists is now included, even if
2457 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2458 address was recognized as an IP address, but was then not correctly decoded
2459 into binary, causing unexpected and incorrect effects when compared with
2466 1. Very minor rewording of debugging text in manualroute to say "list of
2467 hosts" instead of "hostlist".
2469 2. If verify=header_syntax was set, and a header line with an unqualified
2470 address (no domain) and a large number of spaces between the end of the
2471 name and the colon was received, the reception process suffered a buffer
2472 overflow, and (when I tested it) crashed. This was caused by some obsolete
2473 code that should have been removed. The fix is to remove it!
2475 3. When running in the test harness, delay a bit after writing a bounce
2476 message to get a bit more predictability in the log output.
2478 4. Added a call to search_tidyup() just before forking a reception process. In
2479 theory, someone could use a lookup in the expansion of smtp_accept_max_
2480 per_host which, without the tidyup, could leave open a database connection.
2482 5. Added the variables $recipient_data and $sender_data which get set from a
2483 lookup success in an ACL "recipients" or "senders" condition, or a router
2484 "senders" option, similar to $domain_data and $local_part_data.
2486 6. Moved the writing of debug_print from before to after the "senders" test
2489 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2490 problems for message scanning, either using a data ACL, or using
2491 local_scan() because the Received: header was not generated till after they
2492 were called (in order to set the time as the time of reception completion).
2493 I have revised the way this works. The header is now generated after the
2494 body is received, but before the ACL or local_scan() are called. After they
2495 are run, the timestamp in the header is updated.
2501 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2502 before starting a queue runner without re-exec. This happened only when
2503 deliver_drop_privilege was set or when the Exim user was set to root. The
2504 effect of the bug was that timeouts during subsequent deliveries caused
2505 crashes instead of being properly handled. The handler is now left at its
2506 default (and expected) setting.
2508 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2509 message, again when deliver_drop_privilege is set or Exim is run as root.
2510 The bug described in (1) was not present in this case, but the tidying up
2511 of the other signals was missing. I have made the two cases consistent.
2513 3. The ignore_target_hosts setting on a manualroute router was being ignored
2514 for hosts that were looked up using the /MX notation.
2516 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2519 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2520 operated on the sender address. After changing the $sender_address to <>
2521 for the sender address verify, Exim was re-instated it as the original
2522 (before rewriting) address, but remembering that it had rewritten it, so it
2523 wasn't rewriting it again. This bug also had the effect of breaking the
2524 sender address verification caching when the sender address was rewritten.
2526 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2527 This has been changed so that if the ip literal address matches
2528 ignore_target_hosts, the router declines.
2530 7. Added expansion conditions match_domain, match_address, and match_local_
2531 part (NOT match_host).
2533 8. The placeholder for the Received: header didn't have a length field set.
2535 9. Added code to Exim itself and to exim_lock to test for a specific race
2536 condition that could lead to file corruption when using MBX delivery. The
2537 issue is with the lockfile that is created in /tmp. If this file is removed
2538 after a process has opened it but before that process has acquired a lock,
2539 there is the potential for a second process to recreate the file and also
2540 acquire a lock. This could lead to two Exim processes writing to the file
2541 at the same time. The added code performs the same test as UW imapd; it
2542 checks after acquiring the lock that its file descriptor still refers to
2543 the same named file.
2545 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2546 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2549 11. Added the smtp_active_hostname option. If used, this will typically be made
2550 to depend on the incoming interface address. Because $interface_address is
2551 not set up until the daemon has forked a reception process, error responses
2552 that can happen earlier (such as "too many connections") no longer contain
2555 12. If an expansion in a condition on a "warn" statement fails because a lookup
2556 defers, the "warn" statement is abandoned, and the next ACL statement is
2557 processed. Previously this caused the whole ACL to be aborted.
2559 13. Added the iplsearch lookup type.
2561 14. Added ident_timeout as a log selector.
2563 15. Added tls_certificate_verified as a log selector.
2565 16. Added a global option tls_require_ciphers (compare the smtp transport
2566 option of the same name). This controls incoming TLS connections.
2568 17. I finally figured out how to make tls_require_ciphers do a similar thing
2569 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2570 before starting the TLS session.
2572 18. Tabs are now shown as \t in -bP output.
2574 19. If the log selector return_path_on_delivery was set, Exim crashed when
2575 bouncing a message because it had too many Received: header lines.
2577 20. If two routers both had headers_remove settings, and the first one included
2578 a superfluous trailing colon, the final name in the first list and the
2579 first name in the second list were incorrectly joined into one item (with a
2580 colon in the middle).
2586 1. Added -C and -D options to the exinext utility, mainly to make it easier
2587 to include in the automated testing, but these could be helpful when
2588 multiple configurations are in use.
2590 2. The exinext utility was not formatting the output nicely when there was
2591 an alternate port involved in the retry record key, nor when there was a
2592 message id as well (for retries that were specific to a specific message
2593 and a specific host). It was also confused by IPv6 addresses, because of
2594 the additional colons they contain. I have fixed the IPv4 problem, and
2595 patched it up to do a reasonable job for IPv6.
2597 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2598 delivery, the log line now contains "pipelined" if PIPELINING was used.
2600 4. An SMTP transport process used to panic and die if the bind() call to set
2601 an explicit outgoing interface failed. This has been changed; it is now
2602 treated in the same way as a connect() failure.
2604 5. A reference to $sender_host_name in the part of a conditional expansion
2605 that was being skipped was still causing a DNS lookup. This no longer
2608 6. The def: expansion condition was not recognizing references to header lines
2609 that used bh_ and bheader_.
2611 7. Added the _cache feature to named lists.
2613 8. The code for checking quota_filecount in the appendfile transport was
2614 allowing one more file than it should have been.
2616 9. For compatibility with Sendmail, the command line option
2624 and sets the incoming protocol and host name (for trusted callers). The
2625 host name and its colon can be omitted when only the protocol is to be set.
2626 Note the Exim already has two private options, -pd and -ps, that refer to
2627 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2628 "s", but I don't think that's a major issue.
2630 10. A number of refactoring changes to the code, none of which should affect
2633 (a) The number of logging options was getting close to filling up the
2634 32-bit word that was used as a bit map. I have split them into two classes:
2635 those that are passed in the argument to log_write(), and those that are
2636 only ever tested independently outside of that function. These are now in
2637 separate 32-bit words, so there is plenty of room for expansion again.
2638 There is no change in the user interface or the logging behaviour.
2640 (b) When building, for example, log lines, the code previously used a
2641 macro that called string_cat() twice, in order to add two strings. This is
2642 not really sufficiently general. Furthermore, there was one instance where
2643 it was actually wrong because one of the argument was used twice, and in
2644 one call a function was used. (As it happened, calling the function twice
2645 did not affect the overall behaviour.) The macro has been replaced by a
2646 function that can join an arbitrary number of extra strings onto a growing
2649 (c) The code for expansion conditions now uses a table and a binary chop
2650 instead of a serial search (which was left over from when there were very
2651 few conditions). Also, it now recognizes conditions like "pam" even when
2652 the relevant support is not compiled in: a suitably worded error message is
2653 given if an attempt is made to use such a condition.
2655 11. Added ${time_interval:xxxxx}.
2657 12. A bug was causing one of the ddress fields not to be passed back correctly
2658 from remote delivery subprocesses. The field in question was not being
2659 subsequently used, so this caused to problems in practice.
2661 13. Added new log selectors queue_time and deliver_time.
2663 14. Might have fixed a bug in maildirsizefile handling that threw up
2664 "unexpected character" debug warnings, and recalculated the data
2665 unnecessarily. In any case, I expanded the warning message to give more
2668 15. Added the message "Restricted characters in address" to the statements in
2669 the default ACL that block characters like @ and % in local parts.
2671 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2672 Three changes have been made:
2674 (a) There was a serious bug; a negative response to MAIL caused the whole
2675 recipient domain to be cached as invalid, thereby blocking all messages
2676 to all local parts at the same domain, from all senders. This bug has
2677 been fixed. The domain is no longer cached after a negative response to
2678 MAIL if the sender used is not empty.
2680 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2683 (c) A new callout option, "use_sender" has been added for people who want
2684 the modified behaviour.
2690 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2693 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2694 indeed breaks things for older releases.
2696 3. Added additional logging to the case where there is a problem reading data
2697 from a filter that is running in a subprocess using a pipe, in order to
2698 try to track down a specific problem.
2700 4. Testing facility fudge: when running in the test harness and attempting
2701 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2702 getting "No route to host". Convert this to a timeout.
2704 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2707 6. Some OS don't have socklen_t but use size_t instead. This affects the
2708 fifth argument of getsockopt() amongst other things. This is now
2709 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2710 can be set for individual OS. I have set it for SunOS5, OSF1, and
2711 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2712 some earlier ones do not.
2714 7. Change 4.30/15 was not doing the test caselessly.
2716 8. The standard form for an IPv6 address literal was being rejected by address
2717 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2718 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2719 this, as well as the form without the "IPv6" on the front (but only when
2720 address literals are enabled, of course).
2722 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2724 10. Exim crashed if a message with an empty sender address specified by -f
2725 encountered a router with an errors_to setting. This could be provoked only
2726 by a command such as
2730 where an empty string was supplied; "<>" did not hit this bug.
2732 11. Installed PCRE release 4.5.
2734 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2735 remained set. It is now erased.
2737 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2738 times from message ids (which are base 36 rather than the normal 62).
2740 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2741 were being counted as actual protocol errors, and logged if the log
2742 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2743 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2744 and DATA following a set of rejected RCPTs do not count as protocol errors.
2745 In other words, Exim assumes they were pipelined, though this may not
2746 actually be the case. Of course, in all cases the client gets an
2747 appropriate error code.
2749 15. If a lookup fails in an ACL condition, a message about the failure may
2750 be available; it is used if testing the ACL cannot continue, because most
2751 such messages specify what the cause of the deferral is. However, some
2752 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2753 that caused an old message to be retained and used if a later statement
2754 caused a defer, replacing the real cause of the deferral.
2756 16. If an IP address had so many PTR records that the DNS lookup buffer
2757 was not large enough to hold them, Exim could crash while trying to process
2758 the truncated data. It now detects and logs this case.
2760 17. Further to 4.21/58, another change has been made: if (and only if) the
2761 first line of a message (the first header line) ends with CRLF, a bare LF
2762 in a subsequent header line has a space inserted after it, so as not to
2763 terminate the header.
2765 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2766 unnecessarily, used atoi() instead of strtol(), and didn't check the
2767 termination when getting file sizes from file names by regex.
2769 19. Completely re-implemented the support for maildirsize files, in the light
2770 of a number of problems with the previous contributed implementation
2771 (4.30/29). In particular:
2773 . If the quota is zero, the maildirsize file is maintained, but no quota is
2776 . If the maildir directory does not exist, it is created before any attempt
2777 to write a maildirsize file.
2779 . The quota value in the file is just a cache; if the quota is changed in
2780 the transport, the new value overrides.
2782 . A regular expression is available for excluding directories from the
2785 20. The autoreply transport checks the characters in options that define the
2786 message's headers; it allows continued headers, but it was checking with
2787 isspace() after an embedded newline instead of explicitly looking for a
2790 21. If all the "regular" hosts to which an address was routed had passed their
2791 expiry times, and had not reached their retry times, the address was
2792 bounced, even if fallback hosts were defined. Now Exim should go on to try
2795 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2796 equivalent code in the SMTP transport. Some hosts send humungous responses
2797 to HELO/EHLO, more than 1024 it seems.
2799 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2800 gives compiler warnings in some environments. I've now done it "properly",
2803 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2804 (because of problems with the built-in one) was declared to return uschar *
2805 instead of char *, causing compiler failure.
2807 25. Fixed a file descriptor leak when processing alias/forward files.
2809 26. Fixed a minor format string issue in dbfn.c.
2811 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2813 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2814 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2817 29. When a sender address is verified, it is cached, to save repeating the test
2818 when there is more than one recipient in a message. However, when the
2819 verification involves a callout, it is possible for different callout
2820 options to be set for different recipients. It is too complicated to keep
2821 track of this in the cache, so now Exim always runs a verification when a
2822 callout is required, relying on the callout cache for the optimization.
2823 The overhead is duplication of the address routing, but this should not be
2826 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2827 to be verified with callout=postmaster, and the main callout worked but the
2828 postmaster check failed, the verification correctly failed. However, if a
2829 subsequent RCPT command asked for sender verification *without* the
2830 postmaster check, incorrect caching caused this verification also to fail,
2833 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2834 it was not caching the DNS options (qualify_single, search_parents) that
2835 were used when the lookup failed. A subsequent lookup with different
2836 options therefore always gave the same answer, though there were cases
2837 where it should not have. (Example: a "domains = !$mx_any" option on a
2838 dnslookup router: the "domains" option is always processed without any
2839 widening, but the router might have qualify_single set.) Now Exim uses the
2840 cached value only when the same options are set.
2842 32. Added John Jetmore's "exipick" utility to the distribution.
2844 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2845 than a timeout (e.g. a certificate is required, and is not provided), an
2846 Exim server now closes the connection immediately. Previously it waited for
2847 the client to close - but if the client is SSL, it seems that they each
2848 wait for each other, leading to a delay before one of them times out.
2850 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2851 maintained 0.8.x compatibility because I don't think many are using it, and
2852 it is clearly obsolete.
2854 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2857 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2858 client certificate was expired. A simple patch fixes this, though I don't
2859 understand the full logic of why the verify callback is called multiple
2862 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2863 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2864 which causes problems with some clients (such as the Certicom SSL Plus
2865 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2866 disables the coutermeasure allowing Eudora to connect."
2868 38. Exim was not checking that a write() to a log file succeeded. This could
2869 lead to Bad Things if a log got too big, in particular if it hit a file
2870 size limit. Exim now panics and dies if it cannot write to a log file, just
2871 as it does if it cannot open a log file.
2873 39. Modified OS/Makefile-Linux so that it now contains
2875 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2877 The two -D definitions ensure that Exim is compiled with large file
2878 support, which makes it possible to handle log files that are bigger than
2881 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2882 instance) a domain was checked against a named list that involved a lookup,
2883 causing $domain_data to be set, then another domain was checked against the
2884 same list, then the first domain was re-checked, the value of $domain_data
2885 after the final check could be wrong. In particular, if the second check
2886 failed, it could be set empty. This bug probably also applied to
2889 41. The strip_trailing_dot option was not being applied to the address given
2890 with the -f command-line option.
2892 42. The code for reading a message's header from the spool was incrementing
2893 $received_count, but never initializing it. This meant that the value was
2894 incorrect (doubled) while delivering a message in the same process in which
2895 it was received. In the most common configuration of Exim, this never
2896 happens - a fresh exec is done - but it can happen when
2897 deliver_drop_privilege is set.
2899 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2900 it now includes up to 150 characters of the unexpected data in the log
2903 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2904 and building data strings. The size of both of these buffers was 10 000
2905 bytes - far larger than anybody would *ever* want, thought I. Needless to
2906 say, somebody hit the limit. I have increased the maximum line length to
2907 20 000 and the maximum data length of concatenated lines to 100 000. I have
2908 also fixed two bugs, because there was no checking on these buffers. Tsk,
2909 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2910 buffer is too small.
2912 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2913 lsearch lookups. Now it does.
2915 46. When parsing a route_list item in a manualroute router, a fixed-length
2916 buffer was used for the list of hosts. I made this 1024 bytes long,
2917 thinking that nobody would ever have a list of hosts that long. Wrong.
2918 Somebody had a whole pile of complicated expansion conditions, and the
2919 string was silently truncated, leading to an expansion error. It turns out
2920 that it is easier to change to an unlimited length (owing to other changes
2921 that have happened since this code was originally written) than to build
2922 structure for giving a limitation error. The length of the item that
2923 expands into the list of hosts is now unlimited.
2925 47. The lsearch lookup could not handle data where the length of text line was
2926 more than 4095 characters. Such lines were truncated, leading to shortened
2927 data being returned. It should now handle lines of any length.
2929 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2930 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2933 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2934 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2935 no longer happens in comment lines. A list of replacements is now placed
2936 at the head of all of the source files, except those whose only change is
2937 to replace PERL_COMMAND in the very first #! line.
2939 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2940 messages on the queue, with a bottom-up merge sort, using code contributed
2941 by Michael Haardt. This should make operations like -bp somewhat faster on
2942 large queues. It won't affect queue runners, except when queue_run_in_order
2945 51. Installed eximstats 1.31 in the distribution.
2947 52. Added support for SRV lookups to the dnslookup router.
2949 53. If an ACL referred to $message_body or $message_body_end, the value was not
2950 reset for any messages that followed in the same SMTP session.
2952 54. The store-handling optimization for building very long strings was not
2953 differentiating between the different store pools. I don't think this
2954 actually made any difference in practice, but I've tidied it.
2956 55. While running the routers to verify a sender address, $sender_address
2957 was still set to the sender address. This is wrong, because when routing to
2958 send a bounce to the sender, it would be empty. Therefore, I have changed
2959 it so that, while verifying a sender address, $sender_address is set to <>.
2960 (There is no change to what happens when verifying a recipient address.)
2962 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2963 target A or AAAA records (if not already returned) without resetting the
2964 qualify_single or search_parents options of the DNS resolver. These are
2965 inappropriate in this case because the targets of MX and SRV records must
2966 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2967 target that, when qualified, matched something in the local domain. These
2968 two options are now turned off when doing these lookups.
2970 57. It seems that at least some releases of Reiserfs (which does not have the
2971 concept of a fixed number of inodes) returns zero and not -1 for the
2972 number of available inodes. This interacted badly with check_spool_inodes,
2973 which assumed that -1 was the "no such thing" setting. What I have done is
2974 to check that the total number of inodes is greater than zero before doing
2975 the test of how many are available.
2977 58. When a "warn" ACL statement has a log_message modifier, the message is
2978 remembered, and not repeated. This is to avoid a lot of repetition when a
2979 message has many recipients that cause the same warning to be written.
2980 Howewer, Exim was preserving the list of already written lines for an
2981 entire SMTP session, which doesn't seem right. The memory is now reset if a
2982 new message is started.
2984 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2985 debugging output unless log_rewrite was also set.
2987 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2988 of (int)(handle) when we know that handle contains (void *)(-1).
2990 61. The Exim daemon panic-logs an error return when it closes the incoming
2991 connection. However "connection reset by peer" seems to be common, and
2992 isn't really an error worthy of noting specially, so that particular error
2995 62. When Exim is trying to find all the local interfaces, it used to panic and
2996 die if the ioctl to get the interface flags failed. However, it seems that
2997 on at least one OS (Solaris 9) it is possible to have an interface that is
2998 included in the list of interfaces, but for which you get a failure error
2999 for this call. This happens when the interface is not "plumbed" into a
3000 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3001 failure of the "get flags" call assumes that the interface is down.
3003 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3004 makes life easier for people who are doing arithmetic on fields extracted
3005 from dates, where you often get leading zeros that should not be
3006 interpreted as octal.
3008 64. Added qualify_domain to the redirect router, to override the global
3011 65. If a pathologically long header line contained very many addresses (the
3012 report of this problem mentioned 10 000) and each of them was rewritten,
3013 Exim could use up a very large amount of memory. (It kept on making new
3014 copies of the header line as it rewrote, and never released the old ones.)
3015 At the expense of a bit more processing, the header rewriting function has
3016 been changed so that it no longer eats memory in this way.
3018 66. The generation of the Received: header has been moved from the time that a
3019 message starts to be received, to the time that it finishes. The timestamp
3020 in the Received: header should now be very close to that of the <= log
3021 line. There are two side-effects of this change:
3023 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3024 logged header lines no longer include the local Received: line, because
3025 it has not yet been created. The same applies to a copy of the message
3026 that is returned to a non-SMTP sender when a message is rejected.
3028 (b) When a filter file is tested using -bf, no additional Received: header
3029 is added to the test message. After some thought, I decided that this
3032 This change does not affect the value of $received_for. It is still set
3033 after address rewriting, but before local_scan() is called.
3035 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3037 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3038 gave an unhelpful panic error message, and a defer error. I have managed to
3039 change this behaviour so that it now rejects any supplied certificate,
3040 which seems right, as the list of acceptable certificates is empty.
3042 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3043 gave an unhelpful defer error. I have not managed to make this reject any
3044 supplied certificates, but the error message it gives is "no certificate
3045 supplied", which is not helpful.
3047 70. exigrep's output now also includes lines that are not associated with any
3048 message, but which match the given pattern. Implemented by a patch from
3049 Martin Sluka, which also tidied up the Perl a bit.
3051 71. Recipient callout verification, like sender verification, was using <> in
3052 the MAIL FROM command. This isn't really the right thing, since the actual
3053 sender may affect whether the remote host accepts the recipient or not. I
3054 have changed it to use the actual sender in the callout; this means that
3055 the cache record is now keyed on a recipient/sender pair, not just the
3056 recipient address. There doesn't seem to be a real danger of callout loops,
3057 since a callout by the remote host to check the sender would use <>.
3058 [SEE ABOVE: changed after hitting problems.]
3060 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3061 temporary errors. However, in the case of such a code being given after
3062 the end of a data transmission (i.e. after ".") Exim was failing to write
3063 a retry record for the message. (Yes, there was some broken host that was
3064 actually sending 8xx at this point.)
3066 73. An unknown lookup type in a host list could cause Exim to panic-die when
3067 the list was checked. (An example that provoked this was putting <; in the
3068 middle of a list instead of at the start.) If this happened during a DATA
3069 ACL check, a -D file could be left lying around. This kind of configuration
3070 error no longer causes Exim to die; instead it causes a defer errror. The
3071 incident is still logged to the main and panic logs.
3073 74. Buglet left over from Exim 3 conversion. The message "too many messages
3074 in one connection" was written to the rejectlog but not the mainlog, except
3075 when address rewriting (yes!) was being logged.
3077 75. Added write_rejectlog option.
3079 76. When a system filter was run not as root (that is, when system_filter_user
3080 was set), the values of the $n variables were not being returned to the
3081 main process; thus, they were not subsequently available in the $sn
3084 77. Added +return_path_on_delivery log selector.
3086 78. A connection timeout was being treated differently from recipients deferred
3087 when testing hosts_max_try with a message that was older than the host's
3088 retry timeout. (The host should not be counted, thus allowing all hosts to
3089 be tried at least once before bouncing.) This may have been the cause of an
3090 occasionally reported bug whereby a message would remain on the queue
3091 longer than the retry timeout, but would be bounced if a delivery was
3092 forced. I say "may" because I never totally pinned down the problem;
3093 setting up timeout/retry tests is difficult. See also the next item.
3095 79. The ultimate address timeout was not being applied to errors that involved
3096 a combination of host plus message (for example, a timeout on a MAIL
3097 command). When an address resolved to a number of possible hosts, and they
3098 were not all tried for each delivery (e.g. because of hosts_max_try), a
3099 message could remain on the queue longer than the retry timeout.
3101 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3104 81. Fixed an obscure SMTP outgoing bug which required at least the following
3105 conditions: (a) there was another message waiting for the same server;
3106 (b) the server returned 5xx to all RCPT commands in the first message so
3107 that the message was not completed; (c) the server dropped the connection
3108 or gave a negative response to the RSET that Exim sends to abort the
3109 transaction. The observed case was a dropped connection after DATA that had
3110 been sent in pipelining mode. That is, the server had advertised PIPELINING
3111 but was not implementing it correctly. The effect of the bug was incorrect
3112 behaviour, such as trying another host, and this could lead to a crash.
3118 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3119 and daemon.c were passed as pointers to ints; they should have been
3120 pointers to socklen_t variables (which are typically unsigned ints).
3122 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3125 3. Fixed a really odd bug that affected only the testing scheme; patching a
3126 certain fixed string in the binary changed the value of another string that
3127 happened to be identical to the end of the original first string.
3129 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3130 name", it returns that address as the IP address. On some operating
3131 systems (e.g. Solaris), it also passes back the IP address string as the
3132 "host name". However, on others (e.g. Linux), it passes back an empty
3133 string. Exim wasn't checking for this, and was changing the host name to an
3134 empty string, assuming it had been canonicized.
3136 5. Although rare, it is permitted to have more than one PTR record for a given
3137 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3138 all the names associated with an address, because they do in Solaris.
3139 However, it seems that they do not in Linux for data that comes from the
3140 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3141 I found this out when I moved to a new Linux workstation and tried to run
3142 the Exim test suite.
3144 To get round this problem I have changed the code so that it now does its
3145 own call to the DNS to look up PTR records when searching for a host name.
3146 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3147 addresses that are only in /etc/hosts are still found.
3149 This behaviour is, however, controlled by an option called host_lookup_
3150 order, which defaults to "bydns:byaddr". If people want to use the other
3151 order, or indeed, just use one or the other means of lookup, they can
3152 specify it in this variable.
3154 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3155 some operating systems, this comes back from gethostbyaddr() as an empty
3156 string, and this is what Exim used to test for. However, it seems that in
3157 other systems, "." is yielded. Exim now tests for this case too.
3159 7. The values of check_spool_space and check_log_space are now held internally
3160 as a number of kilobytes instead of an absolute number of bytes. If a
3161 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3162 kilobyte. This means that much larger values can be stored.
3164 8. Exim monitor: an attempt to get the action menu when not actually pointing
3165 at a message produces an empty menu entitled "No message selected". This
3166 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3167 no entries in it ("Shell widget menu has zero width and/or height"). So I
3168 have added a single, blank menu entry in this case.
3170 9. Added ${quote_local_part.
3172 10. MIME decoding is now applied to the contents of Subject: header lines when
3175 11. Now that a reference to $sender_host_address automatically causes a reverse
3176 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3177 host lookup before query-style lookups in lists that might use this
3178 variable. This has therefore been abolished, and the "net-" prefix is no
3179 longer necessary for query-style lookups.
3181 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3182 have been a typo for LFLAGS, so it has been changed.
3184 13. The install script calls Exim with "-C /dev/null" in order to find the
3185 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3186 to be output. Howeve, since Exim outputs its version number before the
3187 error, it didn't break the script. It just looked ugly. I fixed this by
3188 always allowing "-C /dev/null" if the caller is root.
3190 14. Ignore overlarge ACL variable number when reading spool file - insurance
3191 against a later release with more variables having written the file.
3193 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3194 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3195 this, as well as the form without the "IPv6" on the front.
3197 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3198 OS/Makefile-Darwin file.
3200 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3201 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3203 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3204 variable. However, some OS use ints for inodes. I've added cast to long int
3205 to get rid of the compiler warning.
3207 19. I had forgotten to lock out "/../" in configuration file names when
3208 ALT_CONFIG_PREFIX was set.
3210 20. Routers used for verification do not need to specify transports. However,
3211 if such a router generated a host list, and callout was configured, Exim
3212 crashed, because it could not find a port number from the (non-existent)
3213 transport. It now assumes port 25 in this circumstance.
3215 21. Added the -t option to exigrep.
3217 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3218 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3219 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3220 them is set, LOOKUP_LSEARCH is forced.
3222 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3224 24. Added sender and host information to the "rejected by local_scan()" log
3225 line; previously there was no indication of these.
3227 25. Added .include_if_exists.
3229 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3230 Linux. It turns out that not all file systems support this. Apparently some
3231 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3232 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3233 not supported on the file descriptor, is now ignored when Exim is trying to
3234 sync a directory. This applies only to Linux.
3236 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3238 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3239 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3242 29. Added support for maildirsize files from supplied patch (modified a bit).
3244 30. The use of :fail: followed by an empty string could lead Exim to respond to
3245 sender verification failures with (e.g.):
3247 550 Verification failed for <xxx>
3248 550 Sender verify failed
3250 where the first response line was missing the '-' that indicates it is not
3251 the final line of the response.
3253 31. The loop for finding the name of the user that called Exim had a hardwired
3254 limit of 10; it now uses the value of finduser_retries, which is used for
3255 all other user lookups.
3257 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3260 33. Exim was neglecting to zero errno before one call of strtol() when
3261 expanding a string and expecting an integer value. On some systems this
3262 resulted in spurious "integer overflow" errors. Also, it was casting the
3263 result into an int without checking.
3265 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3266 did not work. The code looks as if it has *never* worked, though it appears
3267 to have been documented since at least releast 1.62. I have made it work.
3269 35. The "timeout_DNS" error in retry rules, also documented since at least
3270 1.62, also never worked. As it isn't clear exactly what this means, and
3271 clearly it isn't a major issue, I have abolished the feature by treating it
3272 as "timeout", and writing a warning to the main and panic logs.
3274 36. The display of retry rules for -brt wasn't always showing the error code
3277 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3278 timeout_connect_A, timeout_connect_MX.
3280 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3281 to the empty sender.
3283 39. The daemon was not analysing the content of -oX till after it had closed
3284 stderr and disconnected from the controlling terminal. This meant that any
3285 syntax errors were only noted on the panic log, and the return code from
3286 the command was 0. By re-arranging the code a little, I've made the
3287 decoding happen first, so such errors now appear on stderr, and the return
3288 code is 1. However, the actual setting up of the sockets still happens in
3289 the disconnected process, so errors there are still only recorded on the
3292 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3293 connections (as happens on some IP stacks) was logged at start up time as
3294 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3295 it from "IPv6 and IPv4", which means that two separate sockets are being
3298 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3299 says whether AF_INET or AF_INET6 was passed as an argument.
3301 42. Exiwhat output was messed up when time zones were included in log
3304 43. Exiwhat now gives more information about the daemon's listening ports,
3305 and whether -tls-on-connect was used.
3307 44. The "port" option of the smtp transport is now expanded.
3309 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3310 silently ignored. Now an error message is written to the main and panic
3313 46. There's a new ACL modifier called "logwrite" which writes to a log file
3314 as soon as it is encountered.
3316 47. Added $local_user_uid and $local_user_gid at routing time.
3318 48. Exim crashed when trying to verify a sender address that was being
3321 49. Exim was recognizing only a space character after ".include". It now also
3322 recognizes a tab character.
3324 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3325 extracting the relevant information from the specification. The man page no
3326 longer contains scrambled data for the -d option, and I've added a section
3327 at the front about calling Exim under different names.
3329 51. Added "extra_headers" argument to the "mail" command in filter files.
3331 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3334 53. Installed eximstats 1.29.
3336 54. Added transport_filter_timeout as a generic transport option.
3338 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3339 Cc: header lines. This was required by RFC 822, but it not required by RFC
3342 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3343 incoming messages that did not have them. Now it does so only if the
3344 message originates locally, that is, if there is no associated remote host
3345 address. When Resent- header lines are present, this applies to the Resent-
3346 lines rather than the non-Resent- lines.
3348 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3349 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3351 58. Messages for configuration errors now include the name of the main
3352 configuration file - useful now that there may be more than one file in a
3353 list (.included file names were always shown).
3355 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3356 for those rare installations that do not start the daemon as root or run it
3357 setuid root. I've cut out the call to initgroups() if the daemon is not
3360 60. The Exim user and group can now be bound into the binary as text strings
3361 that are looked up at the start of Exim's processing.
3363 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3365 62. Added $mailstore_basename variable.
3367 63. Installed patch to sieve.c from Michael Haardt.
3369 64. When Exim failed to open the panic log after failing to open the main log,
3370 the original message it was trying to log was written to stderr and debug
3371 output, but if they were not available (the usual case in production), it
3372 was lost. Now it is written to syslog before the two lines that record the
3373 failures to open the logs.
3375 65. Users' Exim filters run in subprocesses under the user's uid. It is
3376 possible for a "deliver" command or an alias in a "personal" command to
3377 provoke an address rewrite. If logging of address rewriting is configured,
3378 this fails because the process is not running as root or exim. There may be
3379 a better way of dealing with this, but for the moment (because 4.30 needs
3380 to be released), I have disabled address rewrite logging when running a
3381 filter in a non-root, non-exim process.
3387 1. The buildconfig auxiliary program wasn't quoting the value set for
3388 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3389 not defined. This bug was masked in 4.22 by the effect that was fixed in
3392 2. Some messages that were rejected after a message id was allocated were
3393 shown as "incomplete" by exigrep. It no longer does this for messages that
3394 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3396 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3397 have allow_domain_literals set, the ID did not get logged in the <= line.
3398 Domain literals are now always recognized in Message-ID: header lines.
3400 4. The first argument for a ${extract expansion item is the key name or field
3401 number. Leading and trailing spaces in this item were not being ignored,
3402 causing some misleading effects.
3404 5. When deliver_drop_privilege was set, single queue runner processes started
3405 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3406 same command in the process it spins off) were not dropping privilege.
3408 6. When the daemon running as "exim" started a queue runner, it always
3409 re-executed Exim in the spun-off process. This is a waste of effort when
3410 deliver_drop_privilege is set. The new process now just calls the
3411 queue-runner function directly.
3417 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3420 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3421 ignored. Though the use of -oP was forcing the writing of a pid file, it
3422 was always written to the default place.
3424 3. If the message "no IP address found for host xxxx" is generated during
3425 incoming verification, it is now followed by identification of the incoming
3426 connection (so you can more easily find what provoked it).
3428 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3430 5. Added some features to "harden" Exim a bit more against certain attacks:
3432 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3433 be put in Local/Makefile. This is like the never_users runtime option,
3434 but it cannot be overridden. The default setting is "root".
3436 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3437 prefix string with which any file named in a -C command line option
3440 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3441 is retained for -C and -D only if the caller of Exim is root. Without
3442 it, the exim user may also use -C and -D and retain privilege.
3444 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3445 command line option is disabled.
3447 6. Macro names set by the -D option must start with an upper case letter, just
3448 like macro names defined in the configuration file.
3450 7. Added "dereference=" facility to LDAP.
3452 8. Two instances of the typo "uknown" in the source files are fixed.
3454 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3455 the Configure-Makefile script screwed up while processing it.
3457 10. Incorporated PCRE 4.4.
3459 11. The SMTP synchronization check was not operating right at the start of an
3460 SMTP session. For example, it could not catch a HELO sent before the client
3461 waited for the greeting. There is now a check for outstanding input at the
3462 point when the greeting is written. Because of the duplex, asynchronous
3463 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3464 way, but not yet received, when the check is performed.
3466 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3467 on TCP/IP sockets, because this apparently causes some broken clients to
3470 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3471 unchanged) from the Cygwin maintainer.
3473 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3474 was supported instead of testing for mbx. Effectively a typo.
3476 15. The spa authenticator server code was not checking that the input it
3477 received was valid base64.
3479 16. The debug output line for the "set" modifier in ACLs was not showing the
3480 name of the variable that was being set.
3482 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3484 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3485 lookup on its own. Something else was needed to trigger the lookup. For
3486 example, a match in host_lookup or the need for a host name in a host list.
3487 Now, if $sender_host_name is referenced and the host name has not yet been
3488 looked up, a lookup is performed. If the lookup fails, the variable remains
3489 empty, and $host_lookup_failed is set to "1".
3491 19. Added "eqi" as a case-independent comparison operator.
3493 20. The saslauthd authentication condition could segfault if neither service
3494 nor realm was specified.
3496 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3497 error message that was logged was misleading, and incoming SMTP
3498 connections were dropped. The message is now more accurate, and temporary
3499 errors are given to SMTP connections.
3501 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3502 (except RSET) with a 421 error, until QUIT is received. However, it was
3503 failing to send a response to QUIT.
3505 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3506 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3507 ACL is now run after the helo_try_verify_hosts code.
3509 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3510 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3511 case-independent, so other case variants are also recognized). Apparently
3512 some systems use these upper case variants.
3514 25. If more than two messages were waiting for the same host, and a transport
3515 filter was specified for the transport, Exim sent two messages over the
3516 same TCP/IP connection, and then failed with "socket operation on non-
3517 socket" when it tried to send the third.
3519 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3521 27. The extern definition of crypt16() in expand.c was not being excluded when
3522 the OS had its own crypt16() function.
3524 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3525 as a preferred synonym for return_size_limit, both as an option and as an
3528 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3530 30. Changed the default configuration ACL to relax the local part checking rule
3531 for addresses that are not in any local domains. For these addresses,
3532 slashes and pipe symbols are allowed within local parts, but the sequence
3533 /../ is explicitly forbidden.
3535 31. SPA server authentication was not clearing the challenge buffer before
3538 32. log_message in a "warn" ACL statement was writing to the reject log as
3539 well as to the main log, which contradicts the documentation and doesn't
3540 seem right (because no rejection is happening). So I have stopped it.
3542 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3543 However, I am unable to do any testing of this.
3545 34. Fixed an infelicity in the appendfile transport. When checking directories
3546 for a mailbox, to see if any needed to be created, it was accidentally
3547 using path names with one or more superfluous leading slashes; tracing
3548 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3550 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3551 contents are added to the log line that is written for every discarded
3552 recipient. (Previously a log_message setting was ignored.)
3554 36. The ${quote: operator now quotes the string if it is empty.
3556 37. The install script runs exim in order to find its version number. If for
3557 some reason other than non-existence or emptiness, which it checks, it
3558 could not run './exim', it was installing it with an empty version number,
3559 i.e. as "exim-". This error state is now caught, and the installation is
3562 38. An argument was missing from the function that creates an error message
3563 when Exim fails to connect to the socket for saslauthd authentication.
3564 This could cause Exim to crash, or give a corrupted message.
3566 39. Added isip, isip4, and isip6 to ${if conditions.
3568 40. The ACL variables $acl_xx are now saved with the message, and can be
3569 accessed later in routers, transports, and filters.
3571 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3572 strings in the file are not string-expanded.
3574 42. If a MAIL command specified a SIZE value that was too large to fit into an
3575 int variable, the check against message_size_limit failed. Such values are
3576 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3577 day this will have to be increased, but I don't think I want to be around
3578 when emails are that large.
3585 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3586 iconv() is not standard in FreeBSD.
3588 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3589 IPv6 enabled. The observed symptom was a segmentation fault on return from
3590 the function os_common_find_running_interfaces() in src/os.c.
3592 3. In the check_special_case() function in daemon.c I had used "errno" as an
3593 argument name, which causes warnings on some systems. This was basically a
3594 typo, since it was named "eno" in the comments!
3596 4. The code that waits for the clock to tick (at a resolution of some fraction
3597 of a second) so as to ensure message-id uniqueness was always waiting for
3598 at least one whole tick, when it could have waited for less. [This is
3599 almost certainly not relevant at current processor speeds, where it is
3600 unlikely to ever wait at all. But we try to future-proof.]
3602 5. The function that sleeps for a time interval that includes fractions of a
3603 second contained a race. It did not block SIGALRM between setting the
3604 timer, and suspending (a couple of lines later). If the interval was short
3605 and the sigsuspend() was delayed until after it had expired, the suspension
3606 never ended. On busy systems this could lead to processes getting stuck for
3609 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3610 process, before it forks any delivery processes. The open lookup caching
3611 mechanism meant that the open file or database connection was passed into
3612 the delivery process. The problem was that delivery processes always tidy
3613 up cached lookup data. This could cause a problem for the next delivery
3614 process started by the queue runner, because the external queue runner
3615 process does not know about the closure. So the next delivery process
3616 still has data in the lookup cache. In the case of a file lookup, there was
3617 no problem because closing a file descriptor in a subprocess doesn't affect
3618 the parent. However, if the lookup was caching a connection to a database,
3619 the connection was closed, and the second delivery process was likely to
3620 see errors such as "PGSQL: query failed: server closed the connection
3621 unexpectedly". The problem has been fixed by closing all cached lookups
3622 in a queue runner before running a delivery process.
3624 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3625 seem to have the "const" qualifier which it has on other OS. I've
3628 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3629 *of the same type* (client or server) with the same public name that an
3630 error should be diagnosed.
3632 9. When Exim looked up a host name for an IP address, but failed to find the
3633 original IP address when looking up the host name (a safety check), it
3634 output the message "<ip address> does not match any IP for NULL", which was
3635 confusing, to say the least. The bug was that the host name should have
3636 appeared instead of "NULL".
3638 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3639 user that is built into the binary, and the -C or -D options is used, root
3640 privilege is dropped before the configuration file is read. In addition,
3641 logging is switched to stderr instead of the normal log files. If the
3642 configuration then re-defines the Exim user, the unprivileged environment
3643 is probably not what is expected, so Exim logs a panic warning message (but
3646 However, if deliver_drop_privilege is set, the unprivileged state may well
3647 be exactly what is intended, so the warning has been cut out in that case,
3648 and Exim is allowed to try to write to its normal log files.
3654 1. smtp_return_error_details was not giving details for temporary sender
3655 or receiver verification errors.
3657 2. Diagnose a configuration error if two authenticators have the same public
3660 3. Exim used not to create the message log file for a message until the first
3661 delivery attempt. This could be confusing when incoming messages were held
3662 for policy or load reasons. The message log file is now created at the time
3663 the message is received, and an initial "Received" line is written to it.
3665 4. The automatically generated man page for command line options had a minor
3666 bug that caused no ill effects; however, a more serious problem was that
3667 the procedure for building the man page automatically didn't always
3668 operate. Consequently, release 4.20 contains an out-of-date version. This
3669 shouldn't happen again.
3671 5. When building Exim with embedded Perl support, the script that builds the
3672 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3673 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3675 6. The freeze_tell option was not being used for messages that were frozen on
3676 arrival, either by an ACL or by local_scan().
3678 7. Added the smtp_incomplete_transaction log selector.
3680 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3681 was accepting AUTH without a new EHLO.
3683 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3684 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3687 10. Logging of TCP/IP connections (when configured) now happens in the main
3688 daemon process instead of the child process, so that the TCP/IP connection
3689 count is more accurate (but it can never be perfect).
3691 11. The use of "drop" in a nested ACL was not being handled correctly in the
3692 outer ACL. Now, if condition failure induced by the nested "drop" causes
3693 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3694 or "require"), the connection is dropped.
3696 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3697 that yield "discard" can now be used with an "accept" or a "discard" verb,
3698 but an error is generated for any others (because I can't see a useful way
3699 to define what should happen).
3701 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3702 are now processed in the same way as lines in the Exim configuration file.
3703 In particular, continuation lines are supported.
3705 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3707 15. Added -ti meaning -t -i.
3709 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3710 domains, and warn by logging if others are found.
3712 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3713 returned from GIFCONF when Exim is trying to find the list of interfaces on
3714 a host. The code in os.c has been modified to copy each ifreq to an aligned
3715 structure in all cases.
3717 Also, in some cases, the returned ifreq's were being copied to a 'struct
3718 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3719 means the last couple of bytes of an IPv6 address could be chopped if the
3720 ifreq contained only a normal sockaddr (14 bytes storage).
3722 18. Named domain lists were not supported in the hosts_treat_as_local option.
3723 An entry such as +xxxx was not recognized, and was treated as a literal
3726 19. Ensure that header lines added by a DATA ACL are included in the reject log
3727 if the ACL subsequently rejects the message.
3729 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3730 MD5 (which is deprecated).
3732 21. When testing a filter file using -bf, Exim was writing a message when it
3733 took the sender from a "From " line in the message, but it was not doing so
3734 when it took $return_path from a Return-Path: header line. It now does.
3736 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3737 with a valid header line field name (a series of printing characters
3738 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3740 23. Changed "disc" in the source to "disk" to conform to the documentation and
3741 the book and for uniformity.
3743 24. Ignore Sendmail's -Ooption=value command line item.
3745 25. When execve() failed while trying to run a command in a pipe transport,
3746 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3747 could be confused with a return value of 69 from the command itself. This
3748 has been changed to 127, the value the shell returns if it is asked to run
3749 a non-existent command. The wording for the related log line suggests a
3750 non-existent command as the problem.
3752 26. If received_header_text expands to an empty string, do not add a Received:
3753 header line to the message. (Well, it adds a token one on the spool, but
3754 marks it "old" so that it doesn't get used or transmitted.)
3756 27. Installed eximstats 1.28 (addition of -nt option).
3758 28. There was no check for failure on the call to getsockname() in the daemon
3759 code. This can fail if there is a shortage of resources on the system, with
3760 ENOMEM, for example. A temporary error is now given on failure.
3762 29. Contrary to the C standard, it seems that in some environments, the
3763 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3764 program. Exim now does this explicitly; it affects the formatting of
3765 timestamps using strftime().
3767 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3768 complaints. I've now initialized all the variables, to avoid this.
3770 32. Header lines added by a system filter were not being "seen" during
3771 transport-time rewrites.
3773 33. The info_callback() function passed to OpenSSL is set up with type void
3774 (*)(SSL *, int, int), as described somewhere. However, when calling the
3775 function (actually a macro) that sets it up, the type void(*)() is
3776 expected. I've put in a cast to prevent warnings from picky compilers.
3778 34. If a DNS black list lookup found a CNAME record, but there were no A
3779 records associated with the domain it pointed at, Exim crashed.
3781 35. If a DNS black list lookup returned more than one A record, Exim ignored
3782 all but the first. It now scans all returned addresses if a particular IP
3783 value is being sought. In this situation, the contents of the
3784 $dnslist_value variable are a list of all the addresses, separated by a
3787 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3788 to accept a host name and all its aliases if the forward lookup for any of
3789 them yielded the IP address of the incoming connection. Now it accepts only
3790 those names whose forward lookup yields the correct IP address. Any other
3791 names are discarded. This closes a loophole whereby a rogue DNS
3792 administrator could create reverse DNS records to break through a
3793 wildcarded host restriction in an ACL.
3795 37. If a user filter or a system filter that ran in a subprocess used any of
3796 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3797 the wrong values were passed to the pipe command ($thisaddress had the
3798 value of $0, $0 had the value of $1, etc). This bug was introduced by
3799 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3801 38. Improved the line breaking for long SMTP error messages from ACLs.
3802 Previously, if there was no break point between 40 and 75 characters, Exim
3803 left the rest of the message alone. Two changes have been made: (a) I've
3804 reduced the minimum length to 35 characters; (b) if it can't find a break
3805 point between 35 and 75 characters, it looks ahead and uses the first one
3806 that it finds. This may give the occasional overlong line, but at least the
3807 remaining text gets split now.
3809 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3810 file descriptors might be low, and that setting 1000 would always raise it.
3811 It turns out that in some environments, the limit is already over 1000 and
3812 that lowering it causes trouble. So now Exim takes care not to decrease it.
3814 40. When delivering a message, the value of $return_path is set to $sender_
3815 address at the start of routing (routers may change the value). By an
3816 oversight, this default was not being set up when an address was tested by
3817 -bt or -bv, which affected the outcome if any router or filter referred to
3820 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3821 log only when "message" or "log_message" are set. However, if one of the
3822 conditions was an address verification, or a call to a nested ACL, the
3823 messages generated by the underlying test were being passed through. This
3824 no longer happens. The underlying message is available in $acl_verify_
3825 message for both "message" and "log_message" expansions, so it can be
3826 passed through if needed.
3828 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3829 new expansion $bh_ to give the encoded byte string without charset
3830 translation. Translation happens only if iconv() is available; HAVE_ICONV
3831 indicates this at build time. HEADERS_CHARSET gives the charset to
3832 translate to; headers_charset can change it in the configuration, and
3833 "headers charset" can change it in an individual filter file.
3835 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3836 that creates RFC 2047 encoded "words" labels them as that charset instead
3837 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3838 expansion operator; (ii) when Exim creates a From: line for a local
3839 message; (iii) when a header line is rewritten to include a "phrase" part.
3841 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3842 buggy, causing it to skip the first lines of messages whose message ID
3843 ended in 'D'. This would not have bitten before Exim release 4.14, because
3844 message IDs were unlikely to end in 'D' before then. The effect was to have
3845 incorrect size information for certain domains.
3847 45. #include "config.h" was missing at the start of the crypt16.c module. This
3848 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3851 46. If there was a timeout during a "random" callout check, Exim treated it as
3852 a failure of the random address, and carried on sending RSET and the real
3853 address. If the delay was just some slowness somewhere, the response to the
3854 original RCPT would be taken as a response to RSET and so on, causing
3855 mayhem of various kinds.
3857 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3858 when I implemented it. It didn't allow for the fact that some option values
3859 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3860 the right test for positive values.
3862 48. Domain names in DNS records are case-independent. Exim always looks them up
3863 in lower case. Some resolvers return domain names in exactly the case they
3864 appear in the zone file, that is, they may contain uppercase letters. Not
3865 all resolvers do this - some return always lower case. Exim was treating a
3866 change of case by a resolver as a change of domain, similar to a widening
3867 of a domain abbreviation. This triggered its re-routing code and so it was
3868 trying to route what was effectively the same domain again. This normally
3869 caused routing to fail (because the router wouldn't handle the domain
3870 twice). Now Exim checks for this case specially, and just changes the
3871 casing of the domain that it ultimately uses when it transmits the message
3874 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3877 50. If a filter generated a file delivery with a non-absolute name (possible if
3878 no home directory exists for the router), the forbid_file option was not
3881 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3882 the existing equality matching.
3884 52. Exim was using ints instead of ino_t variables in some places where it was
3885 dealing with inode numbers.
3887 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3888 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3889 called TMPDIR, and if it finds it is different, it changes its value.
3891 54. The smtp_printf() function is now made available to local_scan() so
3892 additional output lines can be written before returning. There is also an
3893 smtp_fflush() function to enable the detection of a dropped connection.
3894 The variables smtp_input and smtp_batched_input are exported to
3897 55. Changed the default runtime configuration: the message "Unknown user"
3898 has been removed from the ACL, and instead placed on the localuser router,
3899 using the cannot_route_message feature. This means that any verification
3900 failures that generate their own messages won't get overridden. Similarly,
3901 the "Unrouteable address" message that was in the ACL for unverifiable
3902 relay addresses has also been removed.
3904 56. Added hosts_avoid_esmtp to the smtp transport.
3906 57. The exicyclog script was not checking for the esoteric option
3907 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3908 will work only if exicyclog is run under the appropriate euid.
3910 58. Following a discussion on the list, the rules by which Exim recognises line
3911 endings on incoming messages have been changed. The -dropcr and drop_cr
3912 options are now no-ops, retained only for backwards compatibility. The
3913 following line terminators are recognized: LF CRLF CR. However, special
3914 processing applies to CR:
3916 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3917 nor a local message in the state where . is a terminator.
3919 (ii) If a bare CR is encountered in a header line, an extra space is added
3920 after the line terminator so as not to end the header. The reasoning
3921 behind this is that bare CRs in header lines are most likely either
3922 to be mistakes, or people trying to play silly games.
3924 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3925 was being incorrectly given as 18 bytes larger than it should have been.
3926 This is a VOB (very old bug).
3928 60. This may never have affected anything current, but just in case it has:
3929 When the local host is found other than at the start of a list of hosts,
3930 the local host, those with the same MX, and any that follow, are discarded.
3931 When the list in question was part of a longer list of hosts, the following
3932 hosts (not currently being processed) were also being discarded. This no
3933 longer happens. I'm not sure if this situation could ever has previously
3936 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3939 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3940 header. This is something that is recommended in a new Internet Draft, and
3941 is something that is documented as being done by Sendmail. There are two
3942 possible values. For messages generated by the autoreply transport, Exim
3945 Auto-Submitted: auto-replied
3947 whereas for all other generated messages (e.g. bounces) it adds
3949 Auto-Submitted: auto-generated
3951 63. The "personal" condition in filters now includes a test for the
3952 Auto-Submitted: header. If it contains the string "auto-" the message it
3953 not considered personal.
3955 64. Added rcpt_include_affixes as a generic transport option.
3957 65. Added queue_only_override (default true).
3959 66. Added the syslog_duplication option.
3961 67. If what should have been the first header line of a message consisted of
3962 a space followed by a colon, Exim was mis-interpreting it as a header line.
3963 It isn't of course - it is syntactically invalid and should therefore be
3964 treated as the start of the message body. The misbehaviour could have
3965 caused a number of strange effects, including loss of data in subsequent
3966 header lines, and spool format errors.
3968 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3969 client host had authenticated. This control can now be exercised by an ACL
3970 for more flexibility.
3972 69. By default, callouts do not happen when testing with -bh. There is now a
3973 variant, -bhc, which does actually run the callout code, including
3974 consulting and updating the callout cache.
3976 70. Added support for saslauthd authentication, courtesy of Alexander
3979 71. If statvfs() failed on the spool or log directories while checking their
3980 size for availability, Exim confusingly gave the error "space shortage".
3981 Furthermore, in debugging mode it crashed with a floating point exception.
3982 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3983 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3984 serious problem, Exim now writes to the main and panic logs when this
3985 happens, with details of the failure. It then refuses to accept the
3986 incoming message, giving the message "spool directory problem" or "log
3987 directory problem" with a 421 code for SMTP messages.
3989 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3990 0, 1, and 2 exist, because some OS complain for execs without them (see
3991 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3992 descriptors. However, the code omitted to check that the open succeeded,
3993 causing mysterious errors if for some reason the permissions on /dev/null
3994 got screwed. Now Exim writes a message to the main and panic logs, and
3995 bombs out if it can't open /dev/null.
3997 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3998 interact so that it is all more flexible. It is supposed to remain
3999 backwards compatible. Also added extra_local_interfaces.
4001 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4002 to bomb out with an assertion failure - to the client this appears as a
4003 connection drop. This problem occurs in the part of the code that was taken
4004 from the Samba project. Fortunately, the assertion is in a very simple
4005 function, so I have fixed this by reproducing the function inline in the
4006 one place where it is called, and arranging for authentication to fail
4007 instead of killing the process with assert().
4009 75. The SPA client code was not working when the server requested OEM rather
4010 than Unicode encoding.
4012 76. Added code to make require_files with a specific uid setting more usable in
4013 the case where statting the file as root fails - usually a non-root-mounted
4014 NFS file system. When this happens and the failure is EACCES, Exim now
4015 forks a subprocess and does the per-uid checking as the relevant uid.
4017 77. Added process_log_path.
4019 78. If log_file_path was not explicitly set, a setting of check_log_space or
4020 check_log_inodes was ignored.
4022 79. If a space check for the spool or log partitions fails, the incident is now
4023 logged. Of course, in the latter case the data may get lost...
4025 80. Added the %p formatting code to string_format() so that it can be used to
4026 print addresses in debug_print(). Adjusted all the address printing in the
4027 debugging in store.c to use %p rather than %d.
4029 81. There was a concern that a line of code in smtp_in.c could overflow a
4030 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4031 initially expressed, the concern was not well-founded, because trailing
4032 spaces are removed early. However, if the trailing spaces were followed by
4033 a NULL, they did not get removed, so the overflow was possible. Two fixes
4036 (a) I re-wrote the offending code in a cleaner fashion.
4037 (b) If an incoming SMTP command contains a NULL character, it is rejected
4040 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4041 runs initgroups(), so that if the Exim user is in any additional groups,
4042 they will be used during message reception.
4048 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob