1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.374 2006/07/13 14:46:05 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 SC/01 Use a glob alias rather than an array ref in eximstats generated
10 parser. This improves both readability and performance.
12 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
13 Don't display local sender or destination tables in eximstats unless
14 there is data to show.
15 Added average volumes into the eximstats top table text output.
17 SC/03 Collect data on the number of addresses (recipients) as well
18 as the number of messages in eximstats.
20 TF/01 Correct an error in the documentation for the redirect router. Exim
21 does (usually) call initgroups() when daemonizing.
23 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
24 with consistent privilege compared to when running as a daemon.
26 TF/03 Note in the spec that $authenticated_id is not set for local
27 submissions from trusted users.
29 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
30 Thanks to Dean Brooks <dean@iglou.com> for the patch.
32 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
33 by adding some example configuration directives to the default
34 configuration file. A little bit of work is required to uncomment the
35 directives and define how usernames and passwords are checked, but
36 there is now a framework to start from.
38 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
39 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
40 without this. I don't know how relevant this is to other LDAP libraries.
42 PH/02 Add the verb name to the "unknown ACL verb" error.
44 PH/03 Magnus Holmgren's patch for filter_prepend_home.
46 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
48 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
49 directory not expanded when it should be if an expanded home directory
50 was set for the address (which is overridden by the transport).
52 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
55 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
56 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
57 because it is too late at that time, and has no effect.
59 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
60 security issue with \' (bugzilla #107). I could not use the
61 PQescapeStringConn() function, because it needs a PGconn value as one of
64 PH/08 When testing addresses using -bt, indicate those final addresses that
65 are duplicates that would not cause an additional delivery. At least one
66 person was confused, thinking that -bt output corresponded to deliveries.
67 (Suppressing duplicates isn't a good idea as you lose the information
68 about possibly different redirections that led to the duplicates.)
70 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
71 systems where poll() doesn't work, in particular OS X.
73 PH/10 Added more information to debugging output for retry time not reached.
75 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
76 operations in malware.c.
78 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
81 PH/13 If write_rejectlog was set false when logging was sent to syslog with
82 syslog_duplication set false, log lines that would normally be written
83 both the the main log and to the reject log were not written to syslog at
86 PH/14 In the default configuration, change the use of "message" in ACL warn
87 statements to "add_header".
89 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
90 not followed by a command (e.g. "seen endif").
92 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
93 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
96 PH/17 Added an extra condition to the default value of delay_warning_condition
100 { match{$h_precedence:}{(?i)bulk|list|junk} }
101 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} }
104 The Auto-Submitted: header is standardised, whereas I don't think
105 Precedence: ever was.
111 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
112 other effects) broke the use of negated acl sub-conditions.
114 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
117 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
118 "Deny" causes Exim to reject the incoming connection with a 554 error.
119 Unfortunately, if there is a major crisis, such as a disk failure,
120 tcp-wrappers gives "deny", whereas what one would like would be some
121 kind of temporary error. A kludge has been added to help with this.
122 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
123 554 error is used if errno is still zero or contains ENOENT (which occurs
124 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
127 PH/03 Add -lutil to the default FreeBSD LIBS setting.
129 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
130 errors. Otherwise a message that provokes a temporary error (when other
131 messages do not) can cause a whole host to time out.
133 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
134 addresses were routed directly to files or pipes from a redirect router.
135 File deliveries just didn't batch; pipe deliveries might have suffered
138 PH/06 A failure to get a lock for a hints database would erroneously always say
139 "Failed to get write lock", even when it was really a read lock.
141 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
142 of 0600. This has been changed to use the value of the lockfile_mode
143 option (which defaults to 0600).
145 PH/08 Applied small patch from the Sieve maintainer.
147 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
148 folder from quota calculations, a direct delivery into this folder messed
149 up the contents of the maildirsize file. This was because the regex was
150 used only to exclude .Trash (or whatever) when the size of the mailbox
151 was calculated. There was no check that a delivery was happening into an
152 excluded directory. This bug has been fixed by ignoring all quota
153 processing for deliveries into excluded directories.
155 PH/10 Added the maildirfolder_create_regex option to appendfile.
161 PH/01 The code for finding all the local interface addresses on a FreeBSD
162 system running IPv6 was broken. This may well have applied to all BSD
163 systems, as well as to others that have similar system calls. The broken
164 code found IPv4 interfaces correctly, but gave incorrect values for the
165 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
166 that it would not match correctly against @[] and not recognize the IPv6
169 PH/02 The ipliteral router was not recognizing addresses of the form user@
170 [ipv6:....] because it didn't know about the "ipv6:" prefix.
172 PH/03 Added disable_ipv6.
174 PH/04 Changed $reply_address to use the raw form of the headers instead of the
175 decoded form, because it is most often used to construct To: headers
176 lines in autoreplies, and the decoded form may well be syntactically
177 invalid. However, $reply_address has leading white space removed, and all
178 newlines turned into spaces so that the autoreply transport does not
181 PH/05 If group was specified without a user on a router, and no group or user
182 was specified on a transport, the group from the router was ignored.
184 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
185 for visible compile-time settings that can be used to change these
186 numbers, for those that want even more. Backwards compatibility with old
187 spool files has been maintained. However, going back to a previous Exim
188 release will lost any variables that are in spool files.
190 PH/07 Two small changes when running in the test harness: increase delay when
191 passing a TCP/IP connection to a new process, in case the original
192 process has to generate a bounce, and remove special handling of
193 127.0.0.2 (sic), which is no longer necessary.
195 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
196 be the same on different OS.
198 PH/09 Moved a debug statement in filter processing to avoid a race problem when
201 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
202 whether --show-vars was specified or not
204 JJ/02 exipick: Added support for new ACL variable spool format introduced
207 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
208 syntactically invalid From: or Reply-to: line, and a filter used this to
209 generate an autoreply, and therefore failed to obtain an address for the
210 autoreply, Exim could try to deliver to a non-existent relative file
211 name, causing unrelated and misleading errors. What now happens is that
212 it logs this as a hard delivery error, but does not attempt to create a
215 PH/11 The exinext utility has a -C option for testing purposes, but although
216 the given file was scanned by exinext itself; it wasn't being passed on
217 when Exim was called.
219 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
220 an end-of-file indication when reading a command response.
222 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
223 compiled. In many other places in Exim, IPv6 addresses are always
224 recognized, so I have changed this. It also means that IPv4 domain
225 literals of the form [IPV4:n.n.n.n] are now always recognized.
227 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
228 used if the router is not running as root, for example, when verifying at
229 ACL time, or when using -bh. The debugging output from this situation was
230 non-existent - all you got was a failure to exec. I have made two
233 (a) Failures to set uid/gid, the current directory, or a process leader
234 in a subprocess such as that created by queryprogram now generate
235 suitable debugging ouput when -d is set.
237 (b) The queryprogram router detects when it is not running as root,
238 outputs suitable debugging information if -d is set, and then runs
239 the subprocess without attempting to change uid/gid.
241 PH/15 Minor change to Makefile for building test_host (undocumented testing
244 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
245 additional section of a DNS packet that returns MX or SRV records.
246 Instead, it always explicitly searches for A/AAAA records. This avoids
247 major problems that occur when a DNS server includes only records of one
248 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
249 fixed another bug: if SRV records were looked up and the corresponding
250 address records were *not* found in the additional section, the port
251 values from the SRV records were lost.
253 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
254 using the correct key (the original address) when searching the retry
255 rules in order to find which one to use for generating the retry hint.
257 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
258 adding the default one. Similarly, if it contains a Reply-To: header, the
259 errors_reply_to option, if set, is not used.
261 PH/19 When calculating a retry time, Exim used to measure the "time since
262 failure" by looking at the "first failed" field in the retry record. Now
263 it does not use this if it is later than than the arrival time of the
264 message. Instead it uses the arrival time. This makes for better
265 behaviour in cases where some deliveries succeed, thus re-setting the
266 "first failed" field. An example is a quota failure for a huge message
267 when small messages continue to be delivered. Without this change, the
268 "time since failure" will always be short, possible causing more frequent
269 delivery attempts for the huge message than are intended.
270 [Note: This change was subsequently modified - see PH/04 for 4.62.]
272 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
273 $1, $2, $3) because the numerical variables can be reset during some
274 expansion items (e.g. "match"), thereby losing the authentication data.
276 PH/21 Make -bV show the size of off_t variables so that the test suite can
277 decide whether to run tests for quotas > 2G.
279 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
280 mailbox_size, and mailbox_filecount in the appendfile transport. If a
281 filecount value is greater than 2G or if a quota value is greater than 2G
282 on a system where the size of off_t is not greater than 4, a panic error
285 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
286 never match. The debug and -bh output now contains an explicit error
287 message indicating a malformed IPv4 address or mask.
289 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
290 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
293 PH/25 Do not write to syslog when running in the test harness. The only
294 occasion when this arises is a failure to open the main or panic logs
295 (for which there is an explicit test).
297 PH/26 Added the /no_tell option to "control=freeze".
299 PH/27 If a host name lookup failed very early in a connection, for example, if
300 the IP address matched host_lookup and the reverse lookup yielded a name
301 that did not have a forward lookup, an error message of the form "no IP
302 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
303 could be logged. Now it outputs the IP address instead of "NULL".
305 PH/28 An enabling patch from MH: add new function child_open_exim2() which
306 allows the sender and the authenticated sender to be set when
307 submitting a message from within Exim. Since child_open_exim() is
308 documented for local_scan(), the new function should be too.
310 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
311 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
312 results in an empty string is now treated as unset.
314 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
316 PH/31 Added sender_verify_fail logging option.
318 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
319 needed by RFC 822 but not by RFC 2822 was commented out. I have now
320 tidied the source and removed it altogether.
322 PH/33 When a queue run was abandoned because the load average was too high, a
323 log line was always written; now it is written only if the queue_run log
324 selector is set. In addition, the log line for abandonment now contains
325 information about the queue run such as the pid. This is always present
326 in "start" and "stop" lines but was omitted from the "abandon" line.
328 PH/34 Omit spaces between a header name and the colon in the error message that
329 is given when verify = headers_syntax fails (if there are lots of them,
330 the message gets confusing).
332 PH/35 Change the default for dns_check_names_pattern to allow slashes within
333 names, as there are now some PTR records that contain slashes. This check
334 is only to protect against broken name servers that fall over on strange
335 characters, so the fact that it applies to all lookups doesn't matter.
337 PH/36 Now that the new test suite is complete, we can remove some of the
338 special code in Exim that was needed for the old test suite. For example,
339 sorting DNS records because real resolvers return them in an arbitrary
340 order. The new test suite's fake resolver always returns records in the
343 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
344 bounces) except when queue_only is set, to avoid logging races between
345 the different processes.
347 PH/38 Panic-die if .include specifies a non-absolute path.
349 PH/39 A tweak to the "H" retry rule from its user.
351 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
352 a label. They prevented compilation on older perls.
354 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
355 a warning to be raised on newish perls.
357 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
358 on queue. Changes to match documented behaviour of showing count of
359 messages matching specified criteria.
361 PH/40 Changed the default ident timeout from 30s to 5s.
363 PH/41 Added support for the use of login_cap features, on those BSD systems
364 that have them, for controlling the resources used by pipe deliveries.
366 PH/42 The content-scanning code uses fopen() to create files in which to put
367 message data. Previously it was not paying any attention to the mode of
368 the files. Exim runs with umask(0) because the rest of the code creates
369 files with open(), and sets the required mode explicitly. Thus, these
370 files were ending up world-writeable. This was not a big issue, because,
371 being within the spool directory, they were not world-accessible. I have
372 created a function called modefopen, which takes an additional mode
373 argument. It sets umask(777), creates the file, chmods it to the required
374 mode, then resets the umask. All the relevant calls to fopen() in the
375 content scanning code have been changed to use this function.
377 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
378 to 24 hours. This avoids potential overflow problems when processing G
379 and H retry rules. I suspect nobody ever tinkers with this value.
381 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
383 PH/45 When the plaintext authenticator is running as a client, the server's
384 challenges are checked to ensure they are valid base64 strings. By
385 default, the authentication attempt is cancelled if an invalid string is
386 received. Setting client_ignore_invalid_base64 true ignores these errors.
387 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
388 they are received. Thus, the responses can be made to depend on the
389 challenges. If an invalid string is ignored, an empty string is placed in
392 PH/46 Messages that are created by the autoreply transport now contains a
393 References: header, in accordance with RFCs 2822 and 3834.
395 PH/47 Added authenticated_sender_force to the smtp transport.
397 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
399 PH/49 Installed latest patch from the Sieve maintainer.
401 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
402 was also set, the appendfile transport was unnecessarily scanning a
403 directory of message files (e.g. for maildir delivery) to find the count
404 of files (along with the size), even though it did not need this
405 information. It now does the scan only if it needs to find either the
406 size of the count of files.
408 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
410 PH/52 Two bugs concerned with error handling when the smtp transport is
413 (i) Exim was not creating retry information for temporary errors given
414 for individual recipients after the DATA command when the smtp transport
415 was used in LMTP mode. This meant that they could be retried too
416 frequently, and not timed out correctly.
418 (ii) Exim was setting the flag that allows error details to be returned
419 for LMTP errors on RCPT commands, but not for LMTP errors for individual
420 recipients that were returned after the DATA command.
422 PH/53 This is related to PH/52, but is more general: for any failing address,
423 when detailed error information was permitted to be returned to the
424 sender, but the error was temporary, then after the final timeout, only
425 "retry timeout exceeded" was returned. Now it returns the full error as
426 well as "retry timeout exceeded".
428 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
429 do this, and (what is worse) MTAs that accept it.
431 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
432 will now be deprecated.
434 PH/56 New os.c-cygwin from the Cygwin maintainer.
436 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
437 formats (previously only available in exim formats via -bpr, -bpru,
438 and -bpra. Now also available in native and exiqgrep formats)
440 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
441 with very large, slow to parse queues
443 JJ/08 exipick: added ! as generic prefix to negate any criteria format
445 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
447 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
448 responses to authentication challenges, though it was showing the
449 challenges; (ii) I've removed the CR characters from the debug output for
452 PH/58 Allow for the insertion of a newline as well as a space when a string
453 is turned into more than one encoded-word during RFC 2047 encoding. The
454 Sieve code now uses this.
456 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
457 data_4xx, lost_connection, tls_required.
459 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
460 message was being sent as an SMTP response.
462 PH/61 Add -l and -k options to exicyclog.
464 PH/62 When verifying, if an address was redirected to one new address, so that
465 verification continued, and the new address failed or deferred after
466 having set something in $address_data, the value of $address_data was not
467 passed back to the ACL. This was different to the case when no
468 redirection occurred. The value is now passed back in both cases.
470 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
471 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
472 use login_cap.h, so on its own it isn't the distinguishing feature. The
473 new name refers directly to the setclassresources() function.
475 PH/65 Added configuration files for NetBSD3.
477 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
479 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
480 is preferred over IPv4.
482 PH/68 The bounce_return_message and bounce_return_body options were not being
483 honoured for bounces generated during the reception of non-SMTP messages.
484 In particular, this applied to messages rejected by the ACL. This bug has
485 been fixed. However, if bounce_return_message is true and bounce_return_
486 body is false, the headers that are returned for a non-SMTP message
487 include only those that have been read before the error was detected.
488 (In the case of an ACL rejection, they have all been read.)
490 PH/69 The HTML version of the specification is now built in a directory called
491 spec_html instead of spec.html, because the latter looks like a path with
492 a MIME-type, and this confuses some software.
494 PH/70 Catch two compiler warnings in sieve.c.
496 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
497 function verify_get_ident() calls ip_connect() to connect a socket, but
498 if the "connect()" function timed out, ip_connect() used to close the
499 socket. However, verify_get_ident() also closes the socket later, and in
500 between Exim writes to the log, which may get opened at this point. When
501 the socket was closed in ip_connect(), the log could get the same file
502 descriptor number as the socket. This naturally causes chaos. The fix is
503 not to close the socket in ip_connect(); the socket should be closed by
504 the function that creates it. There was only one place in the code where
505 this was missing, in the iplookup router, which I don't think anybody now
506 uses, but I've fixed it anyway.
508 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
509 well as to direct DNS lookups. Otherwise the handling of names in host
510 lists is inconsistent and therefore confusing.
516 PH/01 Two changes to the default runtime configuration:
518 (1) Move the checks for relay_from_hosts and authenticated clients from
519 after to before the (commented out) DNS black list checks.
521 (2) Add control=submission to the relay_from_hosts and authenticated
522 clients checks, on the grounds that messages accepted by these
523 statements are most likely to be submissions.
525 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
527 (1) Generate an error if the third argument for the ${prvs expansion is
530 (2) Treat a missing third argument of ${prvscheck as if it were an empty
533 (3) Reset the variables that are obtained from the first argument of
534 ${prvscheck and used in the second argument before leaving the code,
535 because their memory is reclaimed, so using them afterwards may do
538 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
539 one (it's much easier than Tom thought :-).
541 (5) Because of (4), we can now allow for the use of $prvscheck_result
542 inside the third argument.
544 PH/03 For some reason, the default setting of PATH when running a command from
545 a pipe transport was just "/usr/bin". I have changed it to
548 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
549 anything to be listed in the output from -bV.
551 PH/05 When a filter generated an autoreply, the entire To: header line was
552 quoted in the delivery log line, like this:
554 => >A.N.Other <ano@some.domain> <original@ddress> ...
556 This has been changed so that it extracts the operative address. There
557 may be more than one such address. If so, they are comma-separated, like
560 => >ano@some.domain,ona@other.domain <original@ddress> ...
562 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
563 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
564 not being looked up in the rDNS to get a host name, Exim was showing the
565 IP address twice in Received: lines, even though the IP addresses were
566 identical. For example:
568 Received: from [1.2.3.4] (helo=[1.2.3.4])
570 However, if the real host name was known, it was omitting the HELO data
571 if it matched the actual IP address. This has been tidied up so that it
572 doesn't show the same IP address twice.
574 PH/07 When both +timestamp and +memory debugging was on, the value given by
575 $tod_xxx expansions could be wrong, because the tod_stamp() function was
576 called by the debug printing, thereby overwriting the timestamp buffer.
577 Debugging no longer uses the tod_stamp() function when +timestamp is set.
579 PH/08 When the original message was included in an autoreply transport, it
580 always said "this is a copy of the message, including all the headers",
581 even if body_only or headers_only was set. It now gives an appropriate
584 PH/09 Applied a patch from the Sieve maintainer which:
586 o fixes some comments
587 o adds the (disabled) notify extension core
588 o adds some debug output for the result of if/elsif tests
589 o points to the current vacation draft in the documentation
590 and documents the missing references header update
594 o fixes a bug in processing the envelope test (when testing
595 multiple envelope elements, the last element determinted the
598 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
599 Electronic Mail") by including:
601 Auto-submitted: auto-generated
603 in the messages that it generates (bounce messages and others, such as
604 warnings). In the case of bounce messages for non-SMTP mesages, there was
605 also a typo: it was using "Auto_submitted" (underscore instead of
606 hyphen). Since every message generated by Exim is necessarily in response
607 to another message, thes have all been changed to:
609 Auto-Submitted: auto-replied
611 in accordance with these statements in the RFC:
613 The auto-replied keyword:
615 - SHOULD be used on messages sent in direct response to another
616 message by an automatic process,
618 - MUST NOT be used on manually-generated messages,
620 - MAY be used on Delivery Status Notifications (DSNs) and Message
621 Disposition Notifications (MDNs),
623 - MUST NOT be used on messages generated by automatic or periodic
624 processes, except for messages which are automatic responses to
627 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
628 to the default Received: header definition.
630 PH/12 Added log selector acl_warn_skipped (default on).
632 PH/13 After a successful wildlsearch lookup, discard the values of numeric
633 variables because (a) they are in the wrong storage pool and (b) even if
634 they were copied, it wouldn't work properly because of the caching.
636 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
637 checking when decoding. Apparently there are clients that generate
638 overlong encoded strings. Why am I not surprised?
640 PH/15 If the first argument of "${if match_address" was not empty, but did not
641 contain an "@" character, Exim crashed. Now it writes a panic log message
642 and treats the condition as false.
644 PH/16 In autoreply, treat an empty string for "once" the same as unset.
646 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
647 extension "envelope-auth". The code is finished and in agreement with
648 other implementations, but there is no documentation so far and in fact,
649 nobody wrote the draft yet. This extension is currently #undef'ed, thus
650 not changing the active code.
652 Print executed "if" and "elsif" statements when debugging is used. This
653 helps a great deal to understand what a filter does.
655 Document more things not specified clearly in RFC3028. I had all this
656 sorted out, when out of a sudden new issues came to my mind. Oops."
658 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
661 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
662 canonical form (as documented). However, after a host name lookup from
663 the IP address, check_host() was doing a simple string comparison with
664 addresses acquired from the DNS when checking that the found name did
665 have the original IP as one of its addresses. Since any found IPv6
666 addresses are likely to be in abbreviated form, the comparison could
667 fail. Luckily, there already exists a function for doing the comparison
668 by converting both addresses to binary, so now that is used instead of
671 PH/20 There was another similar case to PH/19, when a complete host name was
672 given in a host list; looking up its IP address could give an abbreviated
673 form, whereas the current host's name might or might not be abbreviated.
674 The same fix has been applied.
680 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
681 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
684 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
685 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
687 PH/03 Typo: missing ".o" in src/pcre/Makefile.
689 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
690 header line, restrict the check to what is listed in RFCs 2369 and 2929.
691 Also, for "Auto-Submitted", treat anything other than "no" as
692 non-personal, in accordance with RFC 3834. (Previously it treated
693 anything starting "auto-" as non-personal.)
695 TF/01 The control=submission/name=... option had a problem with syntax
696 errors if the name included a slash character. The /name= option
697 now slurps the rest of the string, so it can include any characters
698 but it must come last in the list of options (after /sender_retain
701 PH/05 Some modifications to the interface to the fake nameserver for the new
709 TK/01 Added the "success_on_redirect" address verification option. See
710 NewStuff for rationale and an example.
712 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
714 PH/02 Patch to exigrep to allow it to work on syslog lines.
716 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
717 fread() to skip over the body file's header line, because in Cygwin the
718 header line is locked and is inaccessible.
720 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
721 co-exist for some time) to make it clear that it is the Exim ID that is
722 referenced, not the Message-ID: header line.
724 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
725 string_format() function, because snprintf() does not exist on all
728 PH/06 The use of forbid_filter_existstest now also locks out the use of the
729 ${stat: expansion item.
731 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
732 protocol synchronization error", to keep the pedants happy.
734 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
735 well as for IRIX systems, when gcc is being used. See the host.c source
738 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
740 PH/10 Named domain lists were not working if used in a queue_smtp_domains
743 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
744 transport and to the smtp transport in LMTP mode.
746 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
748 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
749 run a filter in a subprocess. This could lead to confusion in subsequent
750 lookups in the parent process. There should also be a search_tidyup() at
751 the end of the subprocess.
753 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
754 only if the host matched helo_try_verify_hosts, which caused the
755 verification to occur when the EHLO/HELO command was issued. The ACL just
756 tested the remembered result. Now, if a previous verification attempt has
757 not happened, "verify = helo" does it there and then.
759 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
761 TK/03 Fix log output including CR from clamd.
763 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
764 exist provoked a memory error which could cause a segfault.
766 PH/15 Installed PCRE 6.2
768 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
770 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
771 of the problem. Specifically, suggested +O2 rather than +O1 for the
774 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
776 PH/20 If a delivery was routed to a non-standard port by means of an SRV
777 record, the port was not correctly logged when the outgoing_port log
778 selector was set (it logged the transort's default port).
780 PH/21 Added support for host-specific ports to manualroute, queryprogram,
781 fallback_hosts, and "hosts" in the smtp transport.
783 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
784 host errors such as "Connection refused".
786 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
787 authentication with radiusclient 0.4.9:
789 - Error returned from rc_read_config was caught wrongly
790 - Username/password not passed on to radius server due to wrong length.
792 The presumption is that some radiusclient API changes for 4.51/PH/17
793 were not taken care of correctly. The code is still untested by me (my
794 Linux distribution still has 0.3.2 of radiusclient), but it was
795 contributed by a Radius user.
797 PH/24 When doing a callout, the value of $domain wasn't set correctly when
798 expanding the "port" option of the smtp transport.
800 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
801 while reading a MIME header. Thanks to Tom Hughes for a patch.
803 PH/24 Include config.h inside local_scan.h so that configuration settings are
806 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
807 that in an ACL for RCPT (for example), you can examine exactly what was
810 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
811 commands, but it was not correctly comparing the address with the actual
812 client host address. Thus, it would show the EHLO address in Received:
813 header lines when this was not necessary.
815 PH/27 Added the % operator to ${eval:}.
817 PH/28 Exim tries to create and chdir to its spool directory when it starts;
818 it should be ignoring failures (because with -C, for example, it has lost
819 privilege). It wasn't ignoring creation failures other than "already
822 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
825 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
826 because an input line was too long, either on its own, or by virtue of
827 too many continuations, the temporary file was not being removed, and the
828 return code was incorrect.
830 PH/31 Missing "BOOL" in function definition in filtertest.c.
832 PH/32 Applied Sieve patches from the maintainer.
834 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
836 PH/33 Added "verify = not_blind".
838 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
839 Local/Makefile (with some defaults set). These are used in built scripts
840 such as exicyclog, but they have never been used in the exim_install
841 script (though there are many overriding facilities there). I have
842 arranged that the exim_install script now takes note of these two
845 PH/35 Installed configuration files for Dragonfly.
847 PH/36 When a locally submitted message by a trusted user did not contain a
848 From: header, and the sender address was obtained from -f or from an SMTP
849 MAIL command, and the trusted user did not use -F to supply a sender
850 name, $originator_name was incorrectly used when constructing a From:
851 header. Furthermore, $originator_name was used for submission mode
852 messages from external hosts without From: headers in a similar way,
853 which is clearly wrong.
855 PH/37 Added control=suppress_local_fixups.
857 PH/38 When log_selector = +received_sender was set, and the addition of the
858 sender made the log line's construction buffer exactly full, or one byte
859 less than full, an overflow happened when the terminating "\n" was
862 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
863 when the result of a list match is failure because a DNS lookup failed.
865 PH/40 RM_COMMAND is now used in the building process.
867 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
868 the "build-* directories that it finds.
870 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
871 address in a domain literal was a prefix of an interface address.
873 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
874 when verifying a sender address, unless rewrite_headers is false.
876 PH/44 Wrote a long comment about why errors_to addresses are verified as
877 recipients, not senders.
879 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
880 the ratelimit ACL was added.
882 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
884 PH/46 Added extra information about PostgreSQL errors to the error string.
886 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
887 avoiding the need to install special zones in a real server. This is
888 backwards compatible; if it can't find the fake resolver, it drops back.
889 Thus, both old and new test suites can be run.
891 TF/02 Added util/ratelimit.pl
893 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
894 clock is set back in time.
896 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
897 Candler <B.Candler@pobox.com>.
899 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
900 OK for addresses that are the result of redirections.
902 PH/48 A number of further additions for the benefit of the new test suite,
903 including a fake gethostbyname() that interfaces to the fake DNS resolver
906 TF/06 The fix for widen_domains has also been applied to qualify_single and
907 search_parents which are the other dnslookup options that can cause
910 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
913 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
915 TF/07 Exim produced the error message "an SRV record indicated no SMTP
916 service" if it encountered an MX record with an empty target hostname.
917 The message is now "an MX or SRV record indicated no SMTP service".
919 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
920 if the DNS of the sending site is misconfigured. This is quite a
921 common situation. This change restores the behaviour of treating a
922 helo verification defer as a failure.
924 PH/51 If self=fail was set on a router, the bounce message did not include the
925 actual error message.
931 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
933 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
934 command itself ended in error, the underlying message about the transport
935 filter timeout was being overwritten with the pipe command error. Now the
936 underlying error message should be appended to the second error message.
938 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
940 PH/02 Reduce the amount of output that "make" produces by default. Full output
941 can still be requested.
943 PH/03 The warning log line about a condition test deferring for a "warn" verb
944 was being output only once per connection, rather than after each
945 occurrence (because it was using the same function as for successful
946 "warn" verbs). This seems wrong, so I have changed it.
948 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
949 it should not have, which might have caused a crash in the right
950 circumstances, but probably never did.
952 PH/04 Installed a modified version of Tony Finch's patch to make submission
953 mode fix the return path as well as the Sender: header line, and to
954 add a /name= option so that you can make the user's friendly name appear
957 TF/03 Added the control = fakedefer ACL modifier.
959 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
960 Mark Lowes for thorough testing.
962 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
964 TK/03 Merged latest SRS patch from Miles Wilton.
966 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
967 with the definition in sysexits.h (which is #included earlier).
968 Fortunately, Exim does not actually use EX_OK. The code used to try to
969 preserve the sysexits.h value, by assumimg that macro definitions were
970 scanned for macro replacements. I have been disabused of this notion,
971 so now the code just undefines EX_OK before #including unistd.h.
973 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
974 in the smtp transport. When a block could not be written in a single
975 write() function, the timeout was being re-applied to each part-write.
976 This seems wrong - if the receiver was accepting one byte at a time it
977 would take for ever. The timeout is now adjusted when this happens. It
978 doesn't have to be particularly precise.
980 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
981 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
983 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
984 without a domain if the check to <postmaster@domain> fails.
986 SC/01 Eximstats: added -xls and the ability to specify output files
987 (patch written by Frank Heydlauf).
989 SC/02 Eximstats: use FileHandles for outputing results.
991 SC/03 Eximstats: allow any combination of xls, txt, and html output.
993 SC/04 Eximstats: fixed display of large numbers with -nvr option
995 SC/05 Eximstats: fixed merging of reports with empty tables.
997 SC/06 Eximstats: added the -include_original_destination flag
999 SC/07 Eximstats: removed tabs and trailing whitespace.
1001 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1003 TK/06 MBOX spool code: Add real "From " MBOX separator line
1004 so the .eml file is really in mbox format (even though
1005 most programs do not really care). Patch from Alex Miller.
1007 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1008 The latter is generated from $received_to and is only set if the
1009 message has one envelope recipient. SA can use these headers,
1010 obviously out-of-the-box. Patch from Alex Miller.
1012 PH/08 The ${def test on a variable was returning false if the variable's
1013 value was "0", contrary to what the specification has always said!
1014 The result should be true unless the variable is empty.
1016 PH/09 The syntax error of a character other than { following "${if
1017 def:variable_name" (after optional whitespace) was not being diagnosed.
1018 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1019 accidental colon was present, for example, could give incorrect results.
1021 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1022 result is used (not including appendfile, where other changes are about
1025 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1026 This involved changing a lot of size variables from int to off_t. It
1027 should work with maildirs and everything.
1029 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1030 spamd dying while we are connected to it.
1032 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1035 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1036 function for BDB 4.3.
1038 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1040 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1041 the sources for PCRE have been re-arranged and more files are now
1044 PH/15 The code I had for printing potentially long long variables in PH/11
1045 above was not the best (it lost precision). The length of off_t variables
1046 is now inspected at build time, and an appropriate printing format (%ld
1047 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1048 to be "long long int" or "long int". This is needed for the internal
1049 formatting function string_vformat().
1051 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1052 the configuration file to be ":syslog", then the script "guesses" where
1053 the logs files are, rather than using the compiled in default. In our
1054 case the guess is not the same as the compiled default, so the script
1055 suddenly stopped working when I started to use syslog. The patch checks
1056 to see if log_file_path is "". If so, it attempts to read it from exim
1057 with no configuration file to get the compiled in version, before it
1058 falls back to the previous guessing code."
1060 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1061 implementing BATV in an Exim configuration. See NewStuff for the gory
1064 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1065 Makefile that are specific to HP-UX.
1067 PH/18 If the "use_postmaster" option was set for a recipient callout together
1068 with the "random" option, the postmaster address was used as the MAIL
1069 FROM address for the random test, but not for the subsequent recipient
1070 test. It is now used for both.
1072 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1073 patch removes a few documentation additions to RFC 3028, because the
1074 latest draft now contains them. It adds the new en;ascii-case comparator
1075 and a new error check for 8bit text in MIME parts. Comparator and
1076 require names are now matched exactly. I enabled the subaddress
1077 extension, but it is not well tested yet (read: it works for me)."
1079 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1080 rework some of the code of TK/09 above to avoid the hardwired use of
1081 "%lld" and "long long". Replaced the call to snprintf() with a call to
1084 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1085 records point to non-existent hosts", "retry timeout exceeded", and
1086 "retry time not reached for any host after a long failure period".
1088 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1089 experimental DomainKeys support:
1091 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1092 (2) On an error such as an illegally used "control", the wrong name for
1093 the control was given.
1095 These problems did NOT occur unless DomainKeys support was compiled.
1097 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1099 PH/24 Added ${if match_ip condition.
1101 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1102 databases so that it will be absolutely obvious if a crash occurs in the
1103 DB library. This is a regular occurrence (often caused by mis-matched
1106 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1107 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1108 picked up on a user's system that detects such things. There doesn't seem
1109 to be a gcc warning option for this - only an attribute that has to be
1110 put on the function's prototype. It seems that in Fedora Core 4 they have
1111 set this on a number of new functions. No doubt there will be more in due
1114 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1115 specify a transport. However, if an address that was verified by such a
1116 router was the subject of a callout, Exim crashed because it tried to
1117 read the rcpt_include_affixes from the non-existent transport. Now it
1118 just assumes that the setting of that option is false. This bug was
1119 introduced by 4.51/PH/31.
1121 PH/28 Changed -d+all to exclude +memory, because that information is very
1122 rarely of interest, but it makes the output a lot bigger. People tend to
1123 do -d+all out of habit.
1125 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1126 code in os-type was giving problems when libc.so lives in lib64, like on
1129 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1130 aren't the modern standard, and it seems that some systems' include files
1131 don't always have them. Exim was already checking for some of the newer
1132 ones like T_AAAA, and defining it itself. I've added checks for all the
1133 record types that Exim uses.
1135 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1136 not automatically generating a new one, as it is supposed to. This
1137 prevented TLS from working. If the file did exist, but contained invalid
1138 data, a new version was generated, as expected. It was only the case of a
1139 non-existent file that was broken.
1141 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1142 with a change in libdomainkeys > 0.64.
1144 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1145 from DNS. If the selector record carries the flag, it now has
1146 precedence over the domain-wide flag.
1148 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1150 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1151 the use of a port name that isn't defined in /etc/services) occurred, the
1152 message was deferred as in a normal delivery, and thus remained on the
1153 spool, instead of being failed because of the mua_wrapper setting. This
1154 is now fixed, and I tidied up some of the mua_wrapper messages at the
1157 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1158 the messages in a hash of arrays rather than using individual hashes.
1159 This is a bit cleaner and results in dramatic memory savings, albeit
1160 at a slight CPU cost.
1162 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1163 as requested by Marc Sherman.
1165 SC/10 Eximstats: added histograms for user specified patterns as requested
1168 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1170 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1171 fopen() in the content-scanning modules that did not already have it.
1177 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1178 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1180 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1182 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1184 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1185 to test Sieve filters that use "vacation".
1187 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1188 that changes the way the GnuTLS parameters are stored in the cache file.
1189 The new format can be generated externally. For backward compatibility,
1190 if the data in the cache doesn't make sense, Exim assumes it has read an
1191 old-format file, and it generates new data and writes a new file. This
1192 means that you can't go back to an older release without removing the
1195 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1196 work if there are any delivery delays because "one_time" forces the
1197 parent to be marked "delivered", so its unseen clone is never tried
1198 again. For this reason, Exim now forbids the simultaneous setting of
1201 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1202 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1203 entry if you want to know the details. The fix, however, neglected to
1204 consider the case when local delivery batching is involved. The test for
1205 "previously delivered" was not happening when checking to see if an
1206 address could be batched with a previous (undelivered) one; under
1207 certain circumstances this could lead to multiple deliveries to the same
1210 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1211 in its include files, and this causes problems building Exim.
1213 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1214 header_syntax) but Exim was just ignoring anything given after a slash.
1215 In particular, this caused confusion with an attempt to use "verify =
1216 reverse_host_lookup/defer_ok". An error is now given when options are
1217 supplied for verify items that do not have them. (Maybe reverse_host_
1218 lookup should have a defer_ok option, but that's a different point.)
1220 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1221 defined by RFC 821) to 2048, because there were problems with some AUTH
1222 commands, and RFC 1869 says the size should be increased for extended
1223 SMTP commands that take arguments.
1225 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1228 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1229 "unknown" error; now it says that the functionality isn't in the binary.
1231 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1232 an address' error message when a string expansion fails (syntax or
1233 whatever). Otherwise the password may appear in the log. Following change
1234 PH/42 below, there is no longer a chance of it appearing in a bounce
1237 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1239 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1240 of its addresses was ever tried. (Bugzilla bug #2.)
1242 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1243 the result incorrectly in the debug output. (It correctly added a newline
1244 to what was transported.)
1246 TF/01 Added $received_time.
1248 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1249 commented out examples of how to interface to a virus scanner and to
1250 SpamAssassin. Also added commented examples of av_scanner and
1251 spamd_address settings.
1253 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1254 and controls are allowed in which ACLs. There were a couple of minor
1255 errors. Some of the entries in the conditions table (which is a table of
1256 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1257 negation of where the condition IS allowed.
1259 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1261 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1262 header file does not have a version number, so I've had to invent a new
1263 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1264 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1265 radiusclient), but it was contributed by a Radius user.
1267 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1268 files or directories, for OpenSSL.
1270 PH/19 When an Exim process that is running as root has to create an Exim log
1271 file, it does so in a subprocess that runs as exim:exim so as to get the
1272 ownership right at creation (otherwise, other Exim processes might see
1273 the file with the wrong ownership). There was no test for failure of this
1274 fork() call, which would lead to the process getting stuck as it waited
1275 for a non-existent subprocess. Forks do occasionally fail when resources
1276 run out. I reviewed all the other calls to fork(); they all seem to check
1279 PH/20 When checking for unexpected SMTP input at connect time (before writing
1280 the banner), Exim was not dealing correctly with a non-positive return
1281 from the read() function. If the client had disconnected by this time,
1282 the result was a log entry for a synchronization error with an empty
1283 string after "input=" when read() returned zero. If read() returned -1
1284 (an event I could not check), uninitialized data bytes were printed.
1285 There were reports of junk text (parts of files, etc) appearing after
1288 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1290 PH/22 Added support for macro redefinition, and (re)definition in between
1291 driver and ACL definitions.
1293 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1294 forgetting to use the resulting value; it was using the unexpanded value.
1296 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1297 hadn't been configured. The fix is from Juergen Kreileder, who
1298 understands it better than I do:
1300 "Here's what I see happening with three configured cyrus_sasl
1301 authenticators configured (plain, login, cram-md5):
1303 On startup auth_cyrus_sasl_init() gets called for each of these.
1304 This means three calls to sasl_listmech() without a specified mech_list.
1305 => SASL tests which mechs of all available mechs actually work
1306 => three warnings about OTP not working
1307 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1309 With the patch, sasl_listmech() also gets called three times. But now
1310 SASL's mech_list option is set to the server_mech specified in the the
1311 authenticator. Or in other words, the answer from sasl_listmech()
1312 gets limited to just the mech you're testing for (which is different
1314 => the return list contains just 'plain' or 'login', 'cram-md5' or
1315 nothing depending on the value of ob->server_mech.
1317 I've just tested the patch: Authentication still works fine,
1318 unavailable mechs specified in the exim configuration are still
1319 caught, and the auth.log warnings about OTP are gone."
1321 PH/25 When debugging is enabled, the contents of the command line are added
1322 to the debugging output, even when log_selector=+arguments is not
1325 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1326 answer is "GNU", and only if the return is "GNU/something" is the answer
1329 PH/27 $acl_verify_message is now set immediately after the failure of a
1330 verification in an ACL, and so is available in subsequent modifiers. In
1331 particular, the message can be preserved by coding like this:
1333 warn !verify = sender
1334 set acl_m0 = $acl_verify_message
1336 Previously, $acl_verify_message was set only while expanding "message"
1337 and "log_message" when a very denied access.
1339 PH/28 Modified OS/os.c-Linux with
1341 -#ifndef OS_LOAD_AVERAGE
1342 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1344 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1345 nomenclature these days.)
1347 PH/29 Installed patch from the Sieve maintainer that adds the options
1348 sieve_useraddress and sieve_subaddress to the redirect router.
1350 PH/30 In these circumstances:
1351 . Two addresses routed to the same list of hosts;
1352 . First host does not offer TLS;
1353 . First host accepts first address;
1354 . First host gives temporary error to second address;
1355 . Second host offers TLS and a TLS session is established;
1356 . Second host accepts second address.
1357 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1358 and peerdn, if requested) that were in fact used only for the second
1361 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1362 attention to any local part prefix or suffix that was matched by the
1363 router that accepted the address. It now behaves in the same way as it
1364 does for delivery: the affixes are removed from the local part unless
1365 rcpt_include_affixes is set on the transport.
1367 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1368 timeout during the DATA phase of an incoming message.
1370 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1371 applied a patch sanctioned by the Sieve maintainer.
1373 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1374 the uid or gid is negative. A case of a negative gid caused this to be
1375 noticed. The fix allows for either to be negative.
1377 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1378 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1381 PH/36 The change PH/12 above was broken. Fixed it.
1383 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1384 the grounds that routing the same address twice would always produce the
1385 same answer. This might have been true once, but it is certainly no
1386 longer true now. Routing a child address may depend on the previous
1387 routing that produced that child. Some complicated redirection strategies
1388 went wrong when messages had multiple recipients, and made Exim's
1389 behaviour dependent on the order in which the addresses were given.
1391 I have moved the duplicate checking until after the routing is complete.
1392 Exim scans the addresses that are assigned to local and remote
1393 transports, and removes any duplicates. This means that more work will be
1394 done, as duplicates will always all be routed, but duplicates are
1395 presumably rare, so I don't expect this is of any significance.
1397 For deliveries to pipes, files, and autoreplies, the duplicate checking
1398 still happens during the routing process, since they are not going to be
1401 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1402 It corrects a timeout issue with spamd. This is Ian's comment: "The
1403 background is that sometimes spamd either never reads data from a
1404 connection it has accepted, or it never writes response data. The exiscan
1405 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1406 blindly assumes that writes won't block so it may never time out."
1408 PH/39 Allow G after quota size as well as K and M.
1410 PH/40 The value set for $authenticated_id in an authenticator may not contain
1411 binary zeroes or newlines because the value is written to log lines and
1412 to spool files. There was no check on this. Now the value is run through
1413 the string_printing() function so that such characters are converted to
1414 printable escape sequences.
1416 PH/41 $message_linecount is a new variable that contains the total number of
1417 lines in the message. Compare $body_linecount, which is the count for the
1420 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1421 bounce and delay warning messages, except in certain special cases, which
1424 (a) An SMTP error message from a remote host;
1425 (b) A message specified in a :fail: redirection;
1426 (c) A message specified in a "fail" command in a system filter;
1427 (d) A message specified in a FAIL return from the queryprogram router;
1428 (e) A message specified by the cannot_route_message router option.
1430 In these cases only, Exim does include the error details in bounce and
1431 warning messages. There are also a few cases where bland messages such
1432 as "unrouteable address" or "local delivery error" are given.
1434 PH/43 $value is now also set for the "else" part of a ${run expansion.
1436 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1437 being worked on, but at least Exim now implements the latest version to
1440 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1441 process to complete was treated as a delivery failure, a timeout while
1442 writing the message to the pipe was logged, but erroneously treated as a
1443 successful delivery. Such timeouts include transport filter timeouts. For
1444 consistency with the overall process timeout, these timeouts are now
1445 treated as errors, giving rise to delivery failures by default. However,
1446 there is now a new Boolean option for the pipe transport called
1447 timeout_defer, which, if set TRUE, converts the failures into defers for
1448 both kinds of timeout. A transport filter timeout is now identified in
1451 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1452 systems where "make" and "gmake" are different, calling "gmake" at top
1453 level broke things. I've arranged for the value of $(MAKE) to be passed
1454 from the Makefile to this script so that it can call the same version of
1458 A note about Exim versions 4.44 and 4.50
1459 ----------------------------------------
1461 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1462 changes of various kinds. As a consequence, a big documentation update was
1463 needed. This delayed the release for rather longer than seemed good, especially
1464 in the light of a couple of (minor) security issues. Therefore, the changes
1465 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1466 release. So 4.44 and 4.50 are in effect two different branches that both start
1469 I have left the 4.50 change log unchanged; it contains all the changes since
1470 4.43. The change log for 4.44 is below; many of its items are identical to
1471 those for 4.50. This seems to be the most sensible way to preserve the
1472 historical information.
1478 1. Minor wording change to the doc/README.SIEVE file.
1480 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1481 computation of the current number of files was incorrect.
1483 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1484 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1487 4. Give more explanation in the error message when the command for a transport
1488 filter fails to execute.
1490 5. There are several places where Exim runs a non-Exim command in a
1491 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1492 was being done only for the command run by the queryprogram router. It is
1493 now done for all such subprocesses. The other cases are: ${run, transport
1494 filters, and the commands run by the lmtp and pipe transports.
1496 6. Added CONFIGURE_GROUP build-time option.
1498 7. Some older OS have a limit of 256 on the maximum number of file
1499 descriptors. Exim was using setrlimit() to set 1000 as a large value
1500 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1501 systems. I've change it so that if it can't get 1000, it tries for 256.
1503 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1504 was an oversight, and furthermore, ever since the addition of extra
1505 controls (e.g. 4.43/32), the checks on when to allow different forms of
1506 "control" were broken. There should now be diagnostics for all cases when a
1507 control that does not make sense is encountered.
1509 9. Added the /retain_sender option to "control=submission".
1511 10. $recipients is now available in the predata ACL (oversight).
1513 11. Tidy the search cache before the fork to do a delivery from a message
1514 received from the command line. Otherwise the child will trigger a lookup
1515 failure and thereby defer the delivery if it tries to use (for example) a
1516 cached ldap connection that the parent has called unbind on.
1518 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1519 of $address_data from the recipient verification was clobbered by the
1520 sender verification.
1522 13. The value of address_data from a sender verification is now available in
1523 $sender_address_data in subsequent conditions in the ACL statement.
1525 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1527 15. Added a new option "connect=<time>" to callout options, to set a different
1530 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1531 was its contents. (It was OK if the option was not defined at all.)
1533 17. A "Completed" log line is now written for messages that are removed from
1534 the spool by the -Mrm option.
1536 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1537 information about exactly what failed.
1539 19. Added -dd to debug only the daemon process.
1541 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1542 handles timeouts, both on the server side and network timeouts. Renamed the
1543 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1545 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1546 It is now set to "smtps".
1548 22. $host_address is now set to the target address during the checking of
1549 ignore_target_hosts.
1551 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1552 being passed; this would have caused $sender_host_name to have been used if
1553 matching the list had actually called for a host name (not very likely,
1554 since this list is usually IP addresses). A host name is now passed as
1557 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1558 code that specifies a non-restarting handler (typically sigaction() in
1559 modern systems) in an attempt to fix a rare and obscure crash bug.
1561 25. Narrowed the window for a race in the daemon that could cause it to ignore
1562 SIGCHLD signals. This is not a major problem, because they are used only to
1563 wake it up if nothing else does.
1565 26. A malformed maildirsize file could cause Exim to calculate negative values
1566 for the mailbox size or file count. Odd effects could occur as a result.
1567 The maildirsize information is now recalculated if the size or filecount
1570 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1571 support for a long time. Removed HAVE_SYS_VFS_H.
1573 28. Installed the latest version of exipick from John Jetmore.
1575 29. In an address list, if the pattern was not a regular expression, an empty
1576 subject address (from a bounce message) matched only if the pattern was an
1577 empty string. Non-empty patterns were not even tested. This was the wrong
1578 because it is perfectly reasonable to use an empty address as part of a
1579 database query. An empty address is now tested by patterns that are
1580 lookups. However, all the other forms of pattern expect the subject to
1581 contain a local part and a domain, and therefore, for them, an empty
1582 address still always fails if the pattern is not itself empty.
1584 30. Exim went into a mad DNS loop when attempting to do a callout where the
1585 host was specified on an smtp transport, and looking it up yielded more
1586 than one IP address.
1588 31. Re-factored the code for checking spool and log partition space into a
1589 function that finds that data and another that does the check. The former
1590 is then used to implement four new variables: $spool_space, $log_space,
1591 $spool_inodes, and $log_inodes.
1593 32. The RFC2047 encoding function was originally intended for short strings
1594 such as real names; it was not keeping to the 75-character limit for
1595 encoded words that the RFC imposes. It now respects the limit, and
1596 generates multiple encoded words if necessary. To be on the safe side, I
1597 have increased the buffer size for the ${rfc2047: expansion operator from
1600 33. It is now permitted to omit both strings after an "if" condition; if the
1601 condition is true, the result is "true". As before, when the second string
1602 is omitted, a false condition yields an empty string. This makes it less
1603 cumbersome to write custom ACL and router conditions.
1605 34. Failure to deliver a bounce message always caused it to be frozen, even if
1606 there was an errors_to setting on the router. The errors_to setting is now
1609 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1610 canonical form (fully expanded) before being placed in
1611 $sender_host_address.
1613 36. The table in the code that translates DNS record types into text (T_A to
1614 "A" for instance) was missing entries for NS and CNAME. It is just possible
1615 that this could have caused confusion if both these types were looked up
1616 for the same domain, because the text type is used as part of Exim's
1617 per-process caching. But the chance of anyone hitting this buglet seems
1620 37. The dnsdb lookup has been extended in a number of ways.
1622 (1) There is a new type, "zns", which walks up the domain tree until it
1623 finds some nameserver records. It should be used with care.
1625 (2) There is a new type, "mxh", which is like "mx" except that it returns
1626 just the host names, not the priorities.
1628 (3) It is now possible to give a list of domains (or IP addresses) to be
1629 looked up. The behaviour when one of the lookups defers can be
1630 controlled by a keyword.
1632 (4) It is now possible to specify the separator character for use when
1633 multiple records are returned.
1635 38. The dnslists ACL condition has been extended: it is now possible to supply
1636 a list of IP addresses and/or domains to be looked up in a particular DNS
1639 39. Added log_selector=+queue_time_overall.
1641 40. When running the queue in the test harness, wait just a tad after forking a
1642 delivery process, to get repeatability of debugging output.
1644 41. Include certificate and key file names in error message when GnuTLS fails
1645 to set them up, because the GnuTLS error message doesn't include the name
1646 of the failing file when there is a problem reading it.
1648 42. Allow both -bf and -bF in the same test run.
1650 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1652 44. The "Exiscan patch" is now merged into the mainline Exim source.
1654 45. Sometimes the final signoff response after QUIT could fail to get
1655 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1656 < 0 before doing a fflush(). This bug looks as though it goes back to the
1657 introduction of TLS in release 3.20, but "sometimes" must have been rare
1658 because the tests only now provoked it.
1660 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1661 (this can affect the format of dates).
1663 47. exim_tidydb, when checking for the continued existence of a message for
1664 which it has found a message-specific retry record, was not finding
1665 messages that were in split spool directories. Consequently, it was
1666 deleting retry records that should have stayed in existence.
1668 48. Steve fixed some bugs in eximstats.
1670 49. The SPA authentication driver was not abandoning authentication and moving
1671 on to the next authenticator when an expansion was forced to fail,
1672 contradicting the general specification for all authenticators. Instead it
1673 was generating a temporary error. It now behaves as specified.
1675 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1676 (the order specifies the preference for clients). The order is now AES256,
1677 AES128, 3DES, ARCFOUR128.
1679 51. Small patch to Sieve code - explicitly set From: when generating an
1682 52. Exim crashed if a remote delivery caused a very long error message to be
1683 recorded - for instance if somebody sent an entire SpamAssassin report back
1684 as a large number of 550 error lines. This bug was coincidentally fixed by
1685 increasing the size of one of Exim's internal buffers (big_buffer) that
1686 happened as part of the Exiscan merge. However, to be on the safe side, I
1687 have made the code more robust (and fixed the comments that describe what
1690 53. Now that there can be additional text after "Completed" in log lines (if
1691 the queue_time_overall log selector is set), a one-byte patch to exigrep
1692 was needed to allow it to recognize "Completed" as not the last thing in
1695 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1696 patch that reportedly fixes this has been added. I am not expert enough to
1697 create a test for it. This is what the patch creator wrote:
1699 "I found a little strange behaviour of ldap code when working with
1700 Windows 2003 AD Domain, where users was placed in more than one
1701 Organization Units. When I tried to give exim partial DN, the exit code
1702 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1703 But simultaneously result of request was absolutely normal ldap result,
1704 so I produce this patch..."
1706 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1707 REFERENCE, so I have modified the code to exclude the patch when that macro
1710 55. Some experimental protocols are using DNS PTR records for new purposes. The
1711 keys for these records are domain names, not reversed IP addresses. The
1712 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1713 leaves it alone. Component reversal etc. now happens only for IP addresses.
1716 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1718 57. Double the size of the debug message buffer (to 2048) so that more of very
1719 long debug lines gets shown.
1721 58. The exicyclog utility now does better if the number of log files to keep
1722 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1724 59. Two changes related to the smtp_active_hostname option:
1726 (1) $smtp_active_hostname is now available as a variable.
1727 (2) The default for smtp_banner uses $smtp_active_hostname instead
1728 of $primary_hostname.
1730 60. The host_aton() function is supposed to be passed a string that is known
1731 to be a valid IP address. However, in the case of IPv6 addresses, it was
1732 not checking this. This is a hostage to fortune. Exim now panics and dies
1733 if the condition is not met. A case was found where this could be provoked
1734 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1735 components; fortuitously, this particular loophole had already been fixed
1736 by change 4.50/55 above.
1738 If there are any other similar loopholes, the new check in host_aton()
1739 itself should stop them being exploited. The report I received stated that
1740 data on the command line could provoke the exploit when Exim was running as
1741 exim, but did not say which command line option was involved. All I could
1742 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1743 running as the user.
1746 61. There was a buffer overflow vulnerability in the SPA authentication code
1747 (which came originally from the Samba project). I have added a test to the
1748 spa_base64_to_bits() function which I hope fixes it.
1751 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1752 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1754 63. The daemon start-up calls getloadavg() while still root for those OS that
1755 need the first call to be done as root, but it missed one case: when
1756 deliver_queue_load_max is set with deliver_drop_privilege. This is
1757 necessary for the benefit of the queue runner, because there is no re-exec
1758 when deliver_drop_privilege is set.
1760 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1761 This has been fixed.
1763 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1764 was in use, was not putting the data itself into the right store pool;
1765 consequently, it could be overwritten for a subsequent message in the same
1766 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1769 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1771 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1772 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1773 the function were treating the return as a boolean value, which happened to
1774 work because 0=false and not-0=true, but is not correct code.
1776 68. The host_aton() function was not handling scoped IPv6 addresses (those
1777 with, for example, "%eth0" on the end) correctly.
1779 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1780 negated items (that is, ~something) in unsigned ints. Some compilers
1781 apparently mutter when there is no cast.
1783 70. If an address verification called from an ACL failed, and did not produce a
1784 user-specific message (i.e. there was only a "system" message), nothing was
1785 put in $acl_verify_message. In this situation, it now puts the system
1788 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1789 session; change 4.31/43 added the unwanted input to the log line - except
1790 that it did not do this in the start of session case. It now does.
1792 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1793 This is wrong and can cause the other end to generate a synchronization
1794 error if it is another Exim or anything else that does the synchronization
1795 check. A QUIT command is no longer sent after a timeout.
1797 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1798 during host lookups.
1800 74. The defer_ok option of callout verification was not working if it was used
1801 when verifying addresses in header lines, that is, for this case:
1803 verify = header_sender/callout=defer_ok
1805 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1806 those file descriptors could be used for SMTP connections. If anything
1807 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1808 could be sent to the SMTP client, causing chaos. The daemon now opens
1809 stdin, stdout, and stderr to /dev/null when it puts itself into the
1812 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1813 log by default. The user can override this with suitable Perl magic.
1815 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1816 the log message when discard triggers, was not working for the DATA ACL or
1817 for the non-SMTP ACL.
1819 78. Error message wording change in sieve.c.
1821 79. If smtp_accept_max_per_host was set, the number of connections could be
1822 restricted to fewer than expected, because the daemon was trying to set up
1823 a new connection before checking whether the processes handling previous
1824 connections had finished. The check for completed processes is now done
1825 earlier. On busy systems, this bug wouldn't be noticed because something
1826 else would have woken the daemon, and it would have reaped the completed
1829 80. If a message was submitted locally by a user whose login name contained one
1830 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1831 It caused a spool format error. I have fixed the spool reading code. A
1832 related problem was that the "from" clause in the Received: line became
1833 illegal because of the space(s). It is now covered by ${quote_local_part.
1835 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1838 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1840 83. Patch from Sieve maintainer to fix unterminated string problem in
1841 "vacation" handling.
1843 84. Some minor changes to the Linux configuration files to help with other
1844 OS variants using glibc.
1846 85. One more patch for Sieve to update vacation handling to latest spec.
1849 ----------------------------------------------------
1850 See the note above about the 4.44 and 4.50 releases.
1851 ----------------------------------------------------
1857 1. Change 4.43/35 introduced a bug that caused file counts to be
1858 incorrectly computed when quota_filecount was set in an appendfile
1861 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1862 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1865 3. Give more explanation in the error message when the command for a transport
1866 filter fails to execute.
1868 4. There are several places where Exim runs a non-Exim command in a
1869 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1870 was being done only for the command run by the queryprogram router. It is
1871 now done for all such subprocesses. The other cases are: ${run, transport
1872 filters, and the commands run by the lmtp and pipe transports.
1874 5. Some older OS have a limit of 256 on the maximum number of file
1875 descriptors. Exim was using setrlimit() to set 1000 as a large value
1876 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1877 systems. I've change it so that if it can't get 1000, it tries for 256.
1879 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1880 was an oversight, and furthermore, ever since the addition of extra
1881 controls (e.g. 4.43/32), the checks on when to allow different forms of
1882 "control" were broken. There should now be diagnostics for all cases when a
1883 control that does not make sense is encountered.
1885 7. $recipients is now available in the predata ACL (oversight).
1887 8. Tidy the search cache before the fork to do a delivery from a message
1888 received from the command line. Otherwise the child will trigger a lookup
1889 failure and thereby defer the delivery if it tries to use (for example) a
1890 cached ldap connection that the parent has called unbind on.
1892 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1893 of $address_data from the recipient verification was clobbered by the
1894 sender verification.
1896 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1897 was its contents. (It was OK if the option was not defined at all.)
1899 11. A "Completed" log line is now written for messages that are removed from
1900 the spool by the -Mrm option.
1902 12. $host_address is now set to the target address during the checking of
1903 ignore_target_hosts.
1905 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1906 being passed; this would have caused $sender_host_name to have been used if
1907 matching the list had actually called for a host name (not very likely,
1908 since this list is usually IP addresses). A host name is now passed as
1911 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1912 code that specifies a non-restarting handler (typically sigaction() in
1913 modern systems) in an attempt to fix a rare and obscure crash bug.
1915 15. Narrowed the window for a race in the daemon that could cause it to ignore
1916 SIGCHLD signals. This is not a major problem, because they are used only to
1917 wake it up if nothing else does.
1919 16. A malformed maildirsize file could cause Exim to calculate negative values
1920 for the mailbox size or file count. Odd effects could occur as a result.
1921 The maildirsize information is now recalculated if the size or filecount
1924 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1925 support for a long time. Removed HAVE_SYS_VFS_H.
1927 18. Updated exipick to current release from John Jetmore.
1929 19. Allow an empty sender to be matched against a lookup in an address list.
1930 Previously the only cases considered were a regular expression, or an
1933 20. Exim went into a mad DNS lookup loop when doing a callout where the
1934 host was specified on the transport, if the DNS lookup yielded more than
1937 21. The RFC2047 encoding function was originally intended for short strings
1938 such as real names; it was not keeping to the 75-character limit for
1939 encoded words that the RFC imposes. It now respects the limit, and
1940 generates multiple encoded words if necessary. To be on the safe side, I
1941 have increased the buffer size for the ${rfc2047: expansion operator from
1944 22. Failure to deliver a bounce message always caused it to be frozen, even if
1945 there was an errors_to setting on the router. The errors_to setting is now
1948 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1949 canonical form (fully expanded) before being placed in
1950 $sender_host_address.
1952 24. Updated eximstats to version 1.33
1954 25. Include certificate and key file names in error message when GnuTLS fails
1955 to set them up, because the GnuTLS error message doesn't include the name
1956 of the failing file when there is a problem reading it.
1958 26. Expand error message when OpenSSL has problems setting up cert/key files.
1961 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1962 (this can affect the format of dates).
1964 28. exim_tidydb, when checking for the continued existence of a message for
1965 which it has found a message-specific retry record, was not finding
1966 messages that were in split spool directories. Consequently, it was
1967 deleting retry records that should have stayed in existence.
1969 29. eximstats updated to version 1.35
1970 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1971 1.35 - bugfix such that pie charts by volume are generated correctly
1973 30. The SPA authentication driver was not abandoning authentication and moving
1974 on to the next authenticator when an expansion was forced to fail,
1975 contradicting the general specification for all authenticators. Instead it
1976 was generating a temporary error. It now behaves as specified.
1978 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1979 (the order specifies the preference for clients). The order is now AES256,
1980 AES128, 3DES, ARCFOUR128.
1982 31. Small patch to Sieve code - explicitly set From: when generating an
1985 32. Exim crashed if a remote delivery caused a very long error message to be
1986 recorded - for instance if somebody sent an entire SpamAssassin report back
1987 as a large number of 550 error lines. This bug was coincidentally fixed by
1988 increasing the size of one of Exim's internal buffers (big_buffer) that
1989 happened as part of the Exiscan merge. However, to be on the safe side, I
1990 have made the code more robust (and fixed the comments that describe what
1993 33. Some experimental protocols are using DNS PTR records for new purposes. The
1994 keys for these records are domain names, not reversed IP addresses. The
1995 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1996 leaves it alone. Component reversal etc. now happens only for IP addresses.
1999 34. The host_aton() function is supposed to be passed a string that is known
2000 to be a valid IP address. However, in the case of IPv6 addresses, it was
2001 not checking this. This is a hostage to fortune. Exim now panics and dies
2002 if the condition is not met. A case was found where this could be provoked
2003 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2004 components; fortuitously, this particular loophole had already been fixed
2005 by change 4.50/55 or 4.44/33 above.
2007 If there are any other similar loopholes, the new check in host_aton()
2008 itself should stop them being exploited. The report I received stated that
2009 data on the command line could provoke the exploit when Exim was running as
2010 exim, but did not say which command line option was involved. All I could
2011 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2012 running as the user.
2015 35. There was a buffer overflow vulnerability in the SPA authentication code
2016 (which came originally from the Samba project). I have added a test to the
2017 spa_base64_to_bits() function which I hope fixes it.
2020 36. The daemon start-up calls getloadavg() while still root for those OS that
2021 need the first call to be done as root, but it missed one case: when
2022 deliver_queue_load_max is set with deliver_drop_privilege. This is
2023 necessary for the benefit of the queue runner, because there is no re-exec
2024 when deliver_drop_privilege is set.
2026 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2027 was in use, was not putting the data itself into the right store pool;
2028 consequently, it could be overwritten for a subsequent message in the same
2029 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2032 38. Sometimes the final signoff response after QUIT could fail to get
2033 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2034 < 0 before doing a fflush(). This bug looks as though it goes back to the
2035 introduction of TLS in release 3.20, but "sometimes" must have been rare
2036 because the tests only now provoked it.
2042 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2043 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2044 Now it returns an integer. A number of calls were still expecting a T/F
2045 return. Fortuitously, in all cases, the tests worked in OK situations,
2046 which is the norm. However, things would have gone wrong on any write
2047 failures on the smtp file descriptor. This function is used when sending
2048 messages over SMTP and also when doing verify callouts.
2050 2. When Exim is called to do synchronous delivery of a locally submitted
2051 message (the -odf or -odi options), it no longer closes stderr before doing
2054 3. Implemented the mua_wrapper option.
2056 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2058 5. Implemented the functions header_remove(), header_testname(),
2059 header_add_at_position(), and receive_remove_recipient(), and exported them
2062 6. If an ACL "warn" statement specified the addition of headers, Exim already
2063 inserted X-ACL-Warn: at the start if there was no header name. However, it
2064 was not making this test for the second and subsequent header lines if
2065 there were newlines in the string. This meant that an invalid header could
2066 be inserted if Exim was badly configured.
2068 7. Allow an ACL "warn" statement to add header lines at the start or after all
2069 the Received: headers, as well as at the end.
2071 8. Added the rcpt_4xx retry error code.
2073 9. Added postmaster_mailfrom=xxx to callout verification option.
2075 10. Added mailfrom=xxxx to the callout verification option, for verify=
2078 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2079 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2081 12. Inserted some casts to stop certain compilers warning when using pointer
2082 differences as field lengths or precisions in printf-type calls (mostly
2083 affecting debugging statements).
2085 13. Added optional readline() support for -be (dynamically loaded).
2087 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2088 same clock tick as a message's arrival, so that its received time was the
2089 same as the "first fail" time on the retry record, and that message
2090 remained on the queue past the ultimate address timeout, every queue runner
2091 would try a delivery (because it was past the ultimate address timeout) but
2092 after another failure, the ultimate address timeout, which should have then
2093 bounced the address, did not kick in. This was a "< instead of <=" error;
2094 in most cases the first failure would have been in the next clock tick
2095 after the received time, and all would be well.
2097 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2098 being recognized when the domain list was tested by the match_domain
2099 condition in an expansion string.
2101 16. Added the ${str2b64: operator.
2103 17. Exim was always calling setrlimit() to set a large limit for the number of
2104 processes, without checking whether the existing limit was already
2105 adequate. (It did check for the limit on file descriptors.) Furthermore,
2106 errors from getrlimit() and setrlimit() were being ignored. Now they are
2107 logged to the main and panic logs, but Exim does carry on, to try to do its
2108 job under whatever limits there are.
2110 18. Imported PCRE 5.0.
2112 19. Trivial typo in log message " temporarily refused connection" (the leading
2115 20. If the log selector return_path_on_delivery was set and an address was
2116 redirected to /dev/null, the delivery process crashed because it assumed
2117 that a return path would always be set for a "successful" delivery. In this
2118 case, the whole delivery is bypassed as an optimization, and therefore no
2121 21. Internal re-arrangement: the function for sending a challenge and reading
2122 a response while authentication was assuming a zero-terminated challenge
2123 string. It's now changed to take a pointer and a length, to allow for
2124 binary data in such strings.
2126 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2128 23. Exim was not respecting finduser_retries when seeking the login of the
2129 uid under which it was called; it was always trying 10 times. (The default
2130 setting of finduser_retries is zero.) Also, it was sleeping after the final
2131 failure, which is pointless.
2133 24. Implemented tls_on_connect_ports.
2135 25. Implemented acl_smtp_predata.
2137 26. If the domain in control=submission is set empty, Exim assumes that the
2138 authenticated id is a complete email address when it generates From: or
2139 Sender: header lines.
2141 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2142 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2143 chown and chgrp in /bin and hostname in /usr/bin.
2145 28. Exim was keeping the "process log" file open after each use, just as it
2146 does for the main log. This opens the possibility of it remaining open for
2147 long periods when the USR1 signal hits a daemon. Occasional processlog
2148 errors were reported, that could have been caused by this. Anyway, it seems
2149 much more sensible not to leave this file open at all, so that is what now
2152 29. The long-running daemon process does not normally write to the log once it
2153 has entered its main loop, and it closes the log before doing so. This is
2154 so that log files can straightforwardly be renamed and moved. However,
2155 there are a couple of unusual error situations where the daemon does write
2156 log entries, and I had neglected to close the log afterwards.
2158 30. The text of an SMTP error response that was received during a remote
2159 delivery was being truncated at 512 bytes. This is too short for some of
2160 the long messages that one sometimes sees. I've increased the limit to
2163 31. It is now possible to make retry rules that apply only when a message has a
2164 specific sender, in particular, an empty sender.
2166 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2167 it possible to be selective about when SMTP synchronization is enforced.
2169 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2171 32. Implemented hosts_connection_nolog.
2173 33. Added an ACL for QUIT.
2175 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2178 35. Added mailbox_size and mailbox_filecount to appendfile.
2180 36. Added control = no_multiline_responses to ACLs.
2182 37. There was a bug in the logic of the code that waits for the clock to tick
2183 in the case where the clock went backwards by a substantial amount such
2184 that the microsecond fraction of "now" was more than the microsecond
2185 fraction of "then" (but the whole seconds number was less).
2187 38. Added support for the libradius Radius client library this is found on
2188 FreeBSD (previously only the radiusclient library was supported).
2194 1. When certain lookups returned multiple values in the form name=value, the
2195 quoting of the values was not always being done properly. Specifically:
2196 (a) If the value started with a double quote, but contained no whitespace,
2198 (b) If the value contained whitespace other than a space character (i.e.
2199 tabs or newlines or carriage returns) it was not quoted.
2200 This fix has been applied to the mysql and pgsql lookups by writing a
2201 separate quoting function and calling it from the lookup code. The fix
2202 should probably also be applied to nisplus, ibase and oracle lookups, but
2203 since I cannot test any of those, I have not disturbed their existing code.
2205 2. A hit in the callout cache for a specific address caused a log line with no
2206 reason for rejecting RCPT. Now it says "Previous (cached) callout
2207 verification failure".
2209 3. There was an off-by-one bug in the queryprogram router. An over-long
2210 return line was truncated at 256 instead of 255 characters, thereby
2211 overflowing its buffer with the terminating zero. As well as fixing this, I
2212 have increased the buffer size to 1024 (and made a note to document this).
2214 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2215 when Exim is waiting for an SMTP response from a remote server, Exim
2216 restarts its select() call on the socket, thereby resetting its timeout.
2217 This is not a problem when such interrupts are rare. Somebody set up a cron
2218 job to run exiwhat every 2 minutes, which is less than the normal select()
2219 timeout (5 or 10 minutes). This meant that the select() timeout never
2220 kicked in because it was always reset. I have fixed this by comparing the
2221 time when an interrupt arrives with the time at the start of the first call
2222 to select(). If more time than the timeout has elapsed, the interrupt is
2223 treated as a timeout.
2225 5. Some internal re-factoring in preparation for the addition of Sieve
2226 extensions (by MH). In particular, the "personal" test is moved to a
2227 separate function, and given an option for scanning Cc: and Bcc: (which is
2228 not set for Exim filters).
2230 6. When Exim created an email address using the login of the caller as the
2231 local part (e.g. when creating a From: or Sender: header line), it was not
2232 quoting the local part when it contained special characters such as @.
2234 7. Installed new OpenBSD configuration files.
2236 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2237 try to make them clearer.
2239 9. Callout options, other than the timeout value, were being ignored when
2240 verifying sender addresses in header lines. For example, when using
2242 verify = header_sender/callout=no_cache
2244 the cache was (incorrectly) being used.
2246 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2247 only the Cygwin environment.
2249 11. When return_path_on_delivery was set as a log selector, if different remote
2250 addresses in the same message used different return paths and parallel
2251 remote delivery occurred, the wrong values would sometimes be logged.
2252 (Whenever a remote delivery process finished, the return path value from
2253 the most recently started remote delivery process was logged.)
2255 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2256 lines when AUTH and/or TLS are in use. This is the "received protocol"
2257 field. Exim used to use "asmtp" for authenticated SMTP, without any
2258 indication (in the protocol name) for TLS use. Now it follows the RFC and
2259 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2260 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2261 names appear in log lines as well as in Received: header lines.
2263 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2264 extensions, and comparison tests, and to fix some bugs.
2266 14. Changes to the "personal" filter test:
2268 (1) The test was buggy in that it was just doing the equivalent of
2269 "contains" tests on header lines. For example, if a user's address was
2270 anne@some.where, the "personal" test would incorrectly be true for
2272 To: susanne@some.where
2274 This test is now done by extracting each address from the header in turn,
2275 and checking the entire address. Other tests that are part of "personal"
2276 are now done using regular expressions (for example, to check local parts
2277 of addresses in From: header lines).
2279 (2) The list of non-personal local parts in From: addresses has been
2280 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2281 taken from the Sieve specification recommendations.
2283 (3) If the message contains any header line starting with "List-" it is
2284 treated as non-personal.
2286 (4) The test for "circular" in the Subject: header line has been removed
2287 because it now seems ill-conceived.
2289 15. Minor typos in src/EDITME comments corrected.
2291 16. Installed latest exipick from John Jetmore.
2293 17. If headers_add on a router specified a text string that was too long for
2294 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2295 of string_sprintf() is now avoided.
2297 18. $message_body_size was not set (it was always zero) when running the DATA
2298 ACL and the local_scan() function.
2300 19. For the "mail" command in an Exim filter, no default was being set for
2301 the once_repeat time, causing a random time value to be used if "once" was
2302 specified. (If the value happened to be <= 0, no repeat happened.) The
2303 default is now 0s, meaning "never repeat". The "vacation" command was OK
2304 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2305 (I found it when inspecting the code).
2307 20. There is now an overall timeout for performing a callout verification. It
2308 defaults to 4 times the callout timeout, which applies to individual SMTP
2309 commands during the callout. The overall timeout applies when there is more
2310 than one host that can be tried. The timeout is checked before trying the
2311 next host. This prevents very long delays if there are a large number of
2312 hosts and all are timing out (e.g. when the network connections are timing
2313 out). The value of the overall timeout can be changed by specifying an
2314 additional sub-option for "callout", called "maxwait". For example:
2316 verify = sender/callout=5s,maxwait=20s
2318 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2319 to the end before writing, but this should make it even safer).
2321 22. Exim was forgetting that it had advertised PIPELINING for the second and
2322 subsequent messages on an SMTP connection. It was also not resetting its
2323 memory on STARTTLS and an internal HELO.
2325 23. When Exim logs an SMTP synchronization error within a session, it now
2326 records whether PIPELINING has been advertised or not.
2328 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2329 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2330 rather than long int.
2332 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2334 26. Added the never_mail option to autoreply.
2340 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2341 crash if the getsockname() call failed; this can happen if a connection is
2342 closed very soon after it is established. The problem was simply in the
2343 order in which certain operations were done, causing Exim to try to write
2344 to the SMTP stream before it had set up the file descriptor. The bug has
2345 been fixed by making things happen in the correct order.
2351 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2352 before the connection was closed, thus losing the rejection response.
2354 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2355 some early Solaris releases, but causes trouble in current releases where
2356 socklen_t is defined.
2358 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2361 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2364 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2365 that releasing the top of it at the end releases what was used for sub-
2366 expansions (unless the block got too big). However, discard this block if
2367 the first thing is a variable or header, so that we can use its block when
2368 it is dynamic (useful for very large $message_headers, for example).
2370 6. Lookups now cache *every* query, not just the most recent. A new, separate
2371 store pool is used for this. It can be recovered when all lookup caches are
2372 flushed. Lookups now release memory at the end of their result strings.
2373 This has involved some general refactoring of the lookup sources.
2375 7. Some code has been added to the store_xxx() functions to reduce the amount
2376 of flapping under certain conditions.
2378 8. log_incoming_interface used to affect only the <= reception log lines. Now
2379 it causes the local interface and port to be added to several more SMTP log
2380 lines, for example "SMTP connection from", and rejection lines.
2382 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2384 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2386 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2387 could be overwritten at the end of the current message (or the start of a
2388 new message if it was set in a HELO ACL). The value is now preserved for
2389 the duration of the SMTP connection.
2391 12. If a transport had a headers_rewrite setting, and a matching header line
2392 contained an unqualified address, that address was qualified, even if it
2393 did not match any rewriting rules. The underlying bug was that the values
2394 of the flags that permit the existence of unqualified sender and recipient
2395 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2396 non-local messages, and by -bnq for local messages) were not being
2397 preserved with the message after it was received.
2399 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2400 "next input=" as part of the text comprising the host identity instead of
2401 the correct text. The code was using the same buffer for two different
2402 strings. However, depending on which order the printing function evaluated
2403 its arguments, the bug did not always show up. Under Linux, for example, my
2404 test suite worked just fine.
2406 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2407 doesn't work with some older versions of Perl. It has been changed to "my",
2408 which in any case is probably the better facility to use.
2410 15. A really picky compiler found some instances of statements for creating
2411 error messages that either had too many or two few arguments for the format
2414 16. The size of the buffer for calls to the DNS resolver has been increased
2415 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2416 for addresses that have a lot of PTR records. This alleviates a problem; it
2417 does not fully solve it.
2419 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2420 buffer now truncates the list and logs the incident, which is the same
2421 action as happens when Exim is looking up a host name and its aliases.
2422 Previously in this situation something unpredictable would happen;
2423 sometimes it was "internal error: store_reset failed".
2425 18. If a server dropped the connection unexpectedly when an Exim client was
2426 using GnuTLS and trying to read a response, the client delivery process
2427 crashed while trying to generate an error log message.
2429 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2430 string, for example:
2432 warn message = H1: something\nH2: something
2434 the text was added as a single header line from Exim's point of view
2435 though it ended up OK in the delivered message. However, searching for the
2436 second and subsequent header lines using $h_h2: did not work. This has been
2437 fixed. Similarly, if a system filter added multiple headers in this way,
2438 the routers could not see them.
2440 20. Expanded the error message when iplsearch is called with an invalid key to
2441 suggest using net-iplsearch in a host list.
2443 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2444 ACLs are no longer actually imposed (and a message to that effect is
2447 22. If a "gecos" field in a passwd entry contained escaped characters, in
2448 particular, if it contained a \" sequence, Exim got it wrong when building
2449 a From: or a Sender: header from that name. A second bug also caused
2450 incorrect handling when an unquoted " was present following a character
2451 that needed quoting.
2453 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2454 was not being matched caselessly.
2456 24. Arranged for all hyphens in the exim.8 source to be escaped with
2459 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2460 properly. Recipient callout cache records were still being keyed to include
2461 the sender, even when use_sender was set false. This led to far more
2462 callouts that were necessary. The sender is no longer included in the key
2463 when use_sender is false.
2465 26. Added "control = submission" modifier to ACLs.
2467 27. Added the ${base62d: operator to decode base 62 numbers.
2469 28. dnsdb lookups can now access SRV records.
2471 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2472 the configuration file.
2474 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2475 (-v) mode. This makes the output for a verbose queue run more intelligible.
2477 31. Added a use_postmaster feature to recipient callouts.
2479 32. Added the $body_zerocount variable, containing the number of binary zero
2480 bytes in the message body.
2482 33. The time of last modification of the "new" subdirectory is now used as the
2483 "mailbox time last read" when there is a quota error for a maildir
2486 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2488 35. Added +ignore_unknown as a special item in host lists.
2490 36. Code for decoding IPv6 addresses in host lists is now included, even if
2491 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2492 address was recognized as an IP address, but was then not correctly decoded
2493 into binary, causing unexpected and incorrect effects when compared with
2500 1. Very minor rewording of debugging text in manualroute to say "list of
2501 hosts" instead of "hostlist".
2503 2. If verify=header_syntax was set, and a header line with an unqualified
2504 address (no domain) and a large number of spaces between the end of the
2505 name and the colon was received, the reception process suffered a buffer
2506 overflow, and (when I tested it) crashed. This was caused by some obsolete
2507 code that should have been removed. The fix is to remove it!
2509 3. When running in the test harness, delay a bit after writing a bounce
2510 message to get a bit more predictability in the log output.
2512 4. Added a call to search_tidyup() just before forking a reception process. In
2513 theory, someone could use a lookup in the expansion of smtp_accept_max_
2514 per_host which, without the tidyup, could leave open a database connection.
2516 5. Added the variables $recipient_data and $sender_data which get set from a
2517 lookup success in an ACL "recipients" or "senders" condition, or a router
2518 "senders" option, similar to $domain_data and $local_part_data.
2520 6. Moved the writing of debug_print from before to after the "senders" test
2523 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2524 problems for message scanning, either using a data ACL, or using
2525 local_scan() because the Received: header was not generated till after they
2526 were called (in order to set the time as the time of reception completion).
2527 I have revised the way this works. The header is now generated after the
2528 body is received, but before the ACL or local_scan() are called. After they
2529 are run, the timestamp in the header is updated.
2535 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2536 before starting a queue runner without re-exec. This happened only when
2537 deliver_drop_privilege was set or when the Exim user was set to root. The
2538 effect of the bug was that timeouts during subsequent deliveries caused
2539 crashes instead of being properly handled. The handler is now left at its
2540 default (and expected) setting.
2542 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2543 message, again when deliver_drop_privilege is set or Exim is run as root.
2544 The bug described in (1) was not present in this case, but the tidying up
2545 of the other signals was missing. I have made the two cases consistent.
2547 3. The ignore_target_hosts setting on a manualroute router was being ignored
2548 for hosts that were looked up using the /MX notation.
2550 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2553 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2554 operated on the sender address. After changing the $sender_address to <>
2555 for the sender address verify, Exim was re-instated it as the original
2556 (before rewriting) address, but remembering that it had rewritten it, so it
2557 wasn't rewriting it again. This bug also had the effect of breaking the
2558 sender address verification caching when the sender address was rewritten.
2560 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2561 This has been changed so that if the ip literal address matches
2562 ignore_target_hosts, the router declines.
2564 7. Added expansion conditions match_domain, match_address, and match_local_
2565 part (NOT match_host).
2567 8. The placeholder for the Received: header didn't have a length field set.
2569 9. Added code to Exim itself and to exim_lock to test for a specific race
2570 condition that could lead to file corruption when using MBX delivery. The
2571 issue is with the lockfile that is created in /tmp. If this file is removed
2572 after a process has opened it but before that process has acquired a lock,
2573 there is the potential for a second process to recreate the file and also
2574 acquire a lock. This could lead to two Exim processes writing to the file
2575 at the same time. The added code performs the same test as UW imapd; it
2576 checks after acquiring the lock that its file descriptor still refers to
2577 the same named file.
2579 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2580 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2583 11. Added the smtp_active_hostname option. If used, this will typically be made
2584 to depend on the incoming interface address. Because $interface_address is
2585 not set up until the daemon has forked a reception process, error responses
2586 that can happen earlier (such as "too many connections") no longer contain
2589 12. If an expansion in a condition on a "warn" statement fails because a lookup
2590 defers, the "warn" statement is abandoned, and the next ACL statement is
2591 processed. Previously this caused the whole ACL to be aborted.
2593 13. Added the iplsearch lookup type.
2595 14. Added ident_timeout as a log selector.
2597 15. Added tls_certificate_verified as a log selector.
2599 16. Added a global option tls_require_ciphers (compare the smtp transport
2600 option of the same name). This controls incoming TLS connections.
2602 17. I finally figured out how to make tls_require_ciphers do a similar thing
2603 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2604 before starting the TLS session.
2606 18. Tabs are now shown as \t in -bP output.
2608 19. If the log selector return_path_on_delivery was set, Exim crashed when
2609 bouncing a message because it had too many Received: header lines.
2611 20. If two routers both had headers_remove settings, and the first one included
2612 a superfluous trailing colon, the final name in the first list and the
2613 first name in the second list were incorrectly joined into one item (with a
2614 colon in the middle).
2620 1. Added -C and -D options to the exinext utility, mainly to make it easier
2621 to include in the automated testing, but these could be helpful when
2622 multiple configurations are in use.
2624 2. The exinext utility was not formatting the output nicely when there was
2625 an alternate port involved in the retry record key, nor when there was a
2626 message id as well (for retries that were specific to a specific message
2627 and a specific host). It was also confused by IPv6 addresses, because of
2628 the additional colons they contain. I have fixed the IPv4 problem, and
2629 patched it up to do a reasonable job for IPv6.
2631 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2632 delivery, the log line now contains "pipelined" if PIPELINING was used.
2634 4. An SMTP transport process used to panic and die if the bind() call to set
2635 an explicit outgoing interface failed. This has been changed; it is now
2636 treated in the same way as a connect() failure.
2638 5. A reference to $sender_host_name in the part of a conditional expansion
2639 that was being skipped was still causing a DNS lookup. This no longer
2642 6. The def: expansion condition was not recognizing references to header lines
2643 that used bh_ and bheader_.
2645 7. Added the _cache feature to named lists.
2647 8. The code for checking quota_filecount in the appendfile transport was
2648 allowing one more file than it should have been.
2650 9. For compatibility with Sendmail, the command line option
2658 and sets the incoming protocol and host name (for trusted callers). The
2659 host name and its colon can be omitted when only the protocol is to be set.
2660 Note the Exim already has two private options, -pd and -ps, that refer to
2661 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2662 "s", but I don't think that's a major issue.
2664 10. A number of refactoring changes to the code, none of which should affect
2667 (a) The number of logging options was getting close to filling up the
2668 32-bit word that was used as a bit map. I have split them into two classes:
2669 those that are passed in the argument to log_write(), and those that are
2670 only ever tested independently outside of that function. These are now in
2671 separate 32-bit words, so there is plenty of room for expansion again.
2672 There is no change in the user interface or the logging behaviour.
2674 (b) When building, for example, log lines, the code previously used a
2675 macro that called string_cat() twice, in order to add two strings. This is
2676 not really sufficiently general. Furthermore, there was one instance where
2677 it was actually wrong because one of the argument was used twice, and in
2678 one call a function was used. (As it happened, calling the function twice
2679 did not affect the overall behaviour.) The macro has been replaced by a
2680 function that can join an arbitrary number of extra strings onto a growing
2683 (c) The code for expansion conditions now uses a table and a binary chop
2684 instead of a serial search (which was left over from when there were very
2685 few conditions). Also, it now recognizes conditions like "pam" even when
2686 the relevant support is not compiled in: a suitably worded error message is
2687 given if an attempt is made to use such a condition.
2689 11. Added ${time_interval:xxxxx}.
2691 12. A bug was causing one of the ddress fields not to be passed back correctly
2692 from remote delivery subprocesses. The field in question was not being
2693 subsequently used, so this caused to problems in practice.
2695 13. Added new log selectors queue_time and deliver_time.
2697 14. Might have fixed a bug in maildirsizefile handling that threw up
2698 "unexpected character" debug warnings, and recalculated the data
2699 unnecessarily. In any case, I expanded the warning message to give more
2702 15. Added the message "Restricted characters in address" to the statements in
2703 the default ACL that block characters like @ and % in local parts.
2705 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2706 Three changes have been made:
2708 (a) There was a serious bug; a negative response to MAIL caused the whole
2709 recipient domain to be cached as invalid, thereby blocking all messages
2710 to all local parts at the same domain, from all senders. This bug has
2711 been fixed. The domain is no longer cached after a negative response to
2712 MAIL if the sender used is not empty.
2714 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2717 (c) A new callout option, "use_sender" has been added for people who want
2718 the modified behaviour.
2724 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2727 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2728 indeed breaks things for older releases.
2730 3. Added additional logging to the case where there is a problem reading data
2731 from a filter that is running in a subprocess using a pipe, in order to
2732 try to track down a specific problem.
2734 4. Testing facility fudge: when running in the test harness and attempting
2735 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2736 getting "No route to host". Convert this to a timeout.
2738 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2741 6. Some OS don't have socklen_t but use size_t instead. This affects the
2742 fifth argument of getsockopt() amongst other things. This is now
2743 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2744 can be set for individual OS. I have set it for SunOS5, OSF1, and
2745 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2746 some earlier ones do not.
2748 7. Change 4.30/15 was not doing the test caselessly.
2750 8. The standard form for an IPv6 address literal was being rejected by address
2751 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2752 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2753 this, as well as the form without the "IPv6" on the front (but only when
2754 address literals are enabled, of course).
2756 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2758 10. Exim crashed if a message with an empty sender address specified by -f
2759 encountered a router with an errors_to setting. This could be provoked only
2760 by a command such as
2764 where an empty string was supplied; "<>" did not hit this bug.
2766 11. Installed PCRE release 4.5.
2768 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2769 remained set. It is now erased.
2771 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2772 times from message ids (which are base 36 rather than the normal 62).
2774 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2775 were being counted as actual protocol errors, and logged if the log
2776 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2777 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2778 and DATA following a set of rejected RCPTs do not count as protocol errors.
2779 In other words, Exim assumes they were pipelined, though this may not
2780 actually be the case. Of course, in all cases the client gets an
2781 appropriate error code.
2783 15. If a lookup fails in an ACL condition, a message about the failure may
2784 be available; it is used if testing the ACL cannot continue, because most
2785 such messages specify what the cause of the deferral is. However, some
2786 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2787 that caused an old message to be retained and used if a later statement
2788 caused a defer, replacing the real cause of the deferral.
2790 16. If an IP address had so many PTR records that the DNS lookup buffer
2791 was not large enough to hold them, Exim could crash while trying to process
2792 the truncated data. It now detects and logs this case.
2794 17. Further to 4.21/58, another change has been made: if (and only if) the
2795 first line of a message (the first header line) ends with CRLF, a bare LF
2796 in a subsequent header line has a space inserted after it, so as not to
2797 terminate the header.
2799 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2800 unnecessarily, used atoi() instead of strtol(), and didn't check the
2801 termination when getting file sizes from file names by regex.
2803 19. Completely re-implemented the support for maildirsize files, in the light
2804 of a number of problems with the previous contributed implementation
2805 (4.30/29). In particular:
2807 . If the quota is zero, the maildirsize file is maintained, but no quota is
2810 . If the maildir directory does not exist, it is created before any attempt
2811 to write a maildirsize file.
2813 . The quota value in the file is just a cache; if the quota is changed in
2814 the transport, the new value overrides.
2816 . A regular expression is available for excluding directories from the
2819 20. The autoreply transport checks the characters in options that define the
2820 message's headers; it allows continued headers, but it was checking with
2821 isspace() after an embedded newline instead of explicitly looking for a
2824 21. If all the "regular" hosts to which an address was routed had passed their
2825 expiry times, and had not reached their retry times, the address was
2826 bounced, even if fallback hosts were defined. Now Exim should go on to try
2829 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2830 equivalent code in the SMTP transport. Some hosts send humungous responses
2831 to HELO/EHLO, more than 1024 it seems.
2833 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2834 gives compiler warnings in some environments. I've now done it "properly",
2837 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2838 (because of problems with the built-in one) was declared to return uschar *
2839 instead of char *, causing compiler failure.
2841 25. Fixed a file descriptor leak when processing alias/forward files.
2843 26. Fixed a minor format string issue in dbfn.c.
2845 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2847 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2848 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2851 29. When a sender address is verified, it is cached, to save repeating the test
2852 when there is more than one recipient in a message. However, when the
2853 verification involves a callout, it is possible for different callout
2854 options to be set for different recipients. It is too complicated to keep
2855 track of this in the cache, so now Exim always runs a verification when a
2856 callout is required, relying on the callout cache for the optimization.
2857 The overhead is duplication of the address routing, but this should not be
2860 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2861 to be verified with callout=postmaster, and the main callout worked but the
2862 postmaster check failed, the verification correctly failed. However, if a
2863 subsequent RCPT command asked for sender verification *without* the
2864 postmaster check, incorrect caching caused this verification also to fail,
2867 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2868 it was not caching the DNS options (qualify_single, search_parents) that
2869 were used when the lookup failed. A subsequent lookup with different
2870 options therefore always gave the same answer, though there were cases
2871 where it should not have. (Example: a "domains = !$mx_any" option on a
2872 dnslookup router: the "domains" option is always processed without any
2873 widening, but the router might have qualify_single set.) Now Exim uses the
2874 cached value only when the same options are set.
2876 32. Added John Jetmore's "exipick" utility to the distribution.
2878 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2879 than a timeout (e.g. a certificate is required, and is not provided), an
2880 Exim server now closes the connection immediately. Previously it waited for
2881 the client to close - but if the client is SSL, it seems that they each
2882 wait for each other, leading to a delay before one of them times out.
2884 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2885 maintained 0.8.x compatibility because I don't think many are using it, and
2886 it is clearly obsolete.
2888 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2891 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2892 client certificate was expired. A simple patch fixes this, though I don't
2893 understand the full logic of why the verify callback is called multiple
2896 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2897 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2898 which causes problems with some clients (such as the Certicom SSL Plus
2899 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2900 disables the coutermeasure allowing Eudora to connect."
2902 38. Exim was not checking that a write() to a log file succeeded. This could
2903 lead to Bad Things if a log got too big, in particular if it hit a file
2904 size limit. Exim now panics and dies if it cannot write to a log file, just
2905 as it does if it cannot open a log file.
2907 39. Modified OS/Makefile-Linux so that it now contains
2909 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2911 The two -D definitions ensure that Exim is compiled with large file
2912 support, which makes it possible to handle log files that are bigger than
2915 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2916 instance) a domain was checked against a named list that involved a lookup,
2917 causing $domain_data to be set, then another domain was checked against the
2918 same list, then the first domain was re-checked, the value of $domain_data
2919 after the final check could be wrong. In particular, if the second check
2920 failed, it could be set empty. This bug probably also applied to
2923 41. The strip_trailing_dot option was not being applied to the address given
2924 with the -f command-line option.
2926 42. The code for reading a message's header from the spool was incrementing
2927 $received_count, but never initializing it. This meant that the value was
2928 incorrect (doubled) while delivering a message in the same process in which
2929 it was received. In the most common configuration of Exim, this never
2930 happens - a fresh exec is done - but it can happen when
2931 deliver_drop_privilege is set.
2933 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2934 it now includes up to 150 characters of the unexpected data in the log
2937 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2938 and building data strings. The size of both of these buffers was 10 000
2939 bytes - far larger than anybody would *ever* want, thought I. Needless to
2940 say, somebody hit the limit. I have increased the maximum line length to
2941 20 000 and the maximum data length of concatenated lines to 100 000. I have
2942 also fixed two bugs, because there was no checking on these buffers. Tsk,
2943 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2944 buffer is too small.
2946 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2947 lsearch lookups. Now it does.
2949 46. When parsing a route_list item in a manualroute router, a fixed-length
2950 buffer was used for the list of hosts. I made this 1024 bytes long,
2951 thinking that nobody would ever have a list of hosts that long. Wrong.
2952 Somebody had a whole pile of complicated expansion conditions, and the
2953 string was silently truncated, leading to an expansion error. It turns out
2954 that it is easier to change to an unlimited length (owing to other changes
2955 that have happened since this code was originally written) than to build
2956 structure for giving a limitation error. The length of the item that
2957 expands into the list of hosts is now unlimited.
2959 47. The lsearch lookup could not handle data where the length of text line was
2960 more than 4095 characters. Such lines were truncated, leading to shortened
2961 data being returned. It should now handle lines of any length.
2963 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2964 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2967 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2968 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2969 no longer happens in comment lines. A list of replacements is now placed
2970 at the head of all of the source files, except those whose only change is
2971 to replace PERL_COMMAND in the very first #! line.
2973 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2974 messages on the queue, with a bottom-up merge sort, using code contributed
2975 by Michael Haardt. This should make operations like -bp somewhat faster on
2976 large queues. It won't affect queue runners, except when queue_run_in_order
2979 51. Installed eximstats 1.31 in the distribution.
2981 52. Added support for SRV lookups to the dnslookup router.
2983 53. If an ACL referred to $message_body or $message_body_end, the value was not
2984 reset for any messages that followed in the same SMTP session.
2986 54. The store-handling optimization for building very long strings was not
2987 differentiating between the different store pools. I don't think this
2988 actually made any difference in practice, but I've tidied it.
2990 55. While running the routers to verify a sender address, $sender_address
2991 was still set to the sender address. This is wrong, because when routing to
2992 send a bounce to the sender, it would be empty. Therefore, I have changed
2993 it so that, while verifying a sender address, $sender_address is set to <>.
2994 (There is no change to what happens when verifying a recipient address.)
2996 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2997 target A or AAAA records (if not already returned) without resetting the
2998 qualify_single or search_parents options of the DNS resolver. These are
2999 inappropriate in this case because the targets of MX and SRV records must
3000 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3001 target that, when qualified, matched something in the local domain. These
3002 two options are now turned off when doing these lookups.
3004 57. It seems that at least some releases of Reiserfs (which does not have the
3005 concept of a fixed number of inodes) returns zero and not -1 for the
3006 number of available inodes. This interacted badly with check_spool_inodes,
3007 which assumed that -1 was the "no such thing" setting. What I have done is
3008 to check that the total number of inodes is greater than zero before doing
3009 the test of how many are available.
3011 58. When a "warn" ACL statement has a log_message modifier, the message is
3012 remembered, and not repeated. This is to avoid a lot of repetition when a
3013 message has many recipients that cause the same warning to be written.
3014 Howewer, Exim was preserving the list of already written lines for an
3015 entire SMTP session, which doesn't seem right. The memory is now reset if a
3016 new message is started.
3018 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3019 debugging output unless log_rewrite was also set.
3021 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3022 of (int)(handle) when we know that handle contains (void *)(-1).
3024 61. The Exim daemon panic-logs an error return when it closes the incoming
3025 connection. However "connection reset by peer" seems to be common, and
3026 isn't really an error worthy of noting specially, so that particular error
3029 62. When Exim is trying to find all the local interfaces, it used to panic and
3030 die if the ioctl to get the interface flags failed. However, it seems that
3031 on at least one OS (Solaris 9) it is possible to have an interface that is
3032 included in the list of interfaces, but for which you get a failure error
3033 for this call. This happens when the interface is not "plumbed" into a
3034 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3035 failure of the "get flags" call assumes that the interface is down.
3037 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3038 makes life easier for people who are doing arithmetic on fields extracted
3039 from dates, where you often get leading zeros that should not be
3040 interpreted as octal.
3042 64. Added qualify_domain to the redirect router, to override the global
3045 65. If a pathologically long header line contained very many addresses (the
3046 report of this problem mentioned 10 000) and each of them was rewritten,
3047 Exim could use up a very large amount of memory. (It kept on making new
3048 copies of the header line as it rewrote, and never released the old ones.)
3049 At the expense of a bit more processing, the header rewriting function has
3050 been changed so that it no longer eats memory in this way.
3052 66. The generation of the Received: header has been moved from the time that a
3053 message starts to be received, to the time that it finishes. The timestamp
3054 in the Received: header should now be very close to that of the <= log
3055 line. There are two side-effects of this change:
3057 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3058 logged header lines no longer include the local Received: line, because
3059 it has not yet been created. The same applies to a copy of the message
3060 that is returned to a non-SMTP sender when a message is rejected.
3062 (b) When a filter file is tested using -bf, no additional Received: header
3063 is added to the test message. After some thought, I decided that this
3066 This change does not affect the value of $received_for. It is still set
3067 after address rewriting, but before local_scan() is called.
3069 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3071 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3072 gave an unhelpful panic error message, and a defer error. I have managed to
3073 change this behaviour so that it now rejects any supplied certificate,
3074 which seems right, as the list of acceptable certificates is empty.
3076 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3077 gave an unhelpful defer error. I have not managed to make this reject any
3078 supplied certificates, but the error message it gives is "no certificate
3079 supplied", which is not helpful.
3081 70. exigrep's output now also includes lines that are not associated with any
3082 message, but which match the given pattern. Implemented by a patch from
3083 Martin Sluka, which also tidied up the Perl a bit.
3085 71. Recipient callout verification, like sender verification, was using <> in
3086 the MAIL FROM command. This isn't really the right thing, since the actual
3087 sender may affect whether the remote host accepts the recipient or not. I
3088 have changed it to use the actual sender in the callout; this means that
3089 the cache record is now keyed on a recipient/sender pair, not just the
3090 recipient address. There doesn't seem to be a real danger of callout loops,
3091 since a callout by the remote host to check the sender would use <>.
3092 [SEE ABOVE: changed after hitting problems.]
3094 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3095 temporary errors. However, in the case of such a code being given after
3096 the end of a data transmission (i.e. after ".") Exim was failing to write
3097 a retry record for the message. (Yes, there was some broken host that was
3098 actually sending 8xx at this point.)
3100 73. An unknown lookup type in a host list could cause Exim to panic-die when
3101 the list was checked. (An example that provoked this was putting <; in the
3102 middle of a list instead of at the start.) If this happened during a DATA
3103 ACL check, a -D file could be left lying around. This kind of configuration
3104 error no longer causes Exim to die; instead it causes a defer errror. The
3105 incident is still logged to the main and panic logs.
3107 74. Buglet left over from Exim 3 conversion. The message "too many messages
3108 in one connection" was written to the rejectlog but not the mainlog, except
3109 when address rewriting (yes!) was being logged.
3111 75. Added write_rejectlog option.
3113 76. When a system filter was run not as root (that is, when system_filter_user
3114 was set), the values of the $n variables were not being returned to the
3115 main process; thus, they were not subsequently available in the $sn
3118 77. Added +return_path_on_delivery log selector.
3120 78. A connection timeout was being treated differently from recipients deferred
3121 when testing hosts_max_try with a message that was older than the host's
3122 retry timeout. (The host should not be counted, thus allowing all hosts to
3123 be tried at least once before bouncing.) This may have been the cause of an
3124 occasionally reported bug whereby a message would remain on the queue
3125 longer than the retry timeout, but would be bounced if a delivery was
3126 forced. I say "may" because I never totally pinned down the problem;
3127 setting up timeout/retry tests is difficult. See also the next item.
3129 79. The ultimate address timeout was not being applied to errors that involved
3130 a combination of host plus message (for example, a timeout on a MAIL
3131 command). When an address resolved to a number of possible hosts, and they
3132 were not all tried for each delivery (e.g. because of hosts_max_try), a
3133 message could remain on the queue longer than the retry timeout.
3135 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3138 81. Fixed an obscure SMTP outgoing bug which required at least the following
3139 conditions: (a) there was another message waiting for the same server;
3140 (b) the server returned 5xx to all RCPT commands in the first message so
3141 that the message was not completed; (c) the server dropped the connection
3142 or gave a negative response to the RSET that Exim sends to abort the
3143 transaction. The observed case was a dropped connection after DATA that had
3144 been sent in pipelining mode. That is, the server had advertised PIPELINING
3145 but was not implementing it correctly. The effect of the bug was incorrect
3146 behaviour, such as trying another host, and this could lead to a crash.
3152 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3153 and daemon.c were passed as pointers to ints; they should have been
3154 pointers to socklen_t variables (which are typically unsigned ints).
3156 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3159 3. Fixed a really odd bug that affected only the testing scheme; patching a
3160 certain fixed string in the binary changed the value of another string that
3161 happened to be identical to the end of the original first string.
3163 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3164 name", it returns that address as the IP address. On some operating
3165 systems (e.g. Solaris), it also passes back the IP address string as the
3166 "host name". However, on others (e.g. Linux), it passes back an empty
3167 string. Exim wasn't checking for this, and was changing the host name to an
3168 empty string, assuming it had been canonicized.
3170 5. Although rare, it is permitted to have more than one PTR record for a given
3171 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3172 all the names associated with an address, because they do in Solaris.
3173 However, it seems that they do not in Linux for data that comes from the
3174 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3175 I found this out when I moved to a new Linux workstation and tried to run
3176 the Exim test suite.
3178 To get round this problem I have changed the code so that it now does its
3179 own call to the DNS to look up PTR records when searching for a host name.
3180 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3181 addresses that are only in /etc/hosts are still found.
3183 This behaviour is, however, controlled by an option called host_lookup_
3184 order, which defaults to "bydns:byaddr". If people want to use the other
3185 order, or indeed, just use one or the other means of lookup, they can
3186 specify it in this variable.
3188 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3189 some operating systems, this comes back from gethostbyaddr() as an empty
3190 string, and this is what Exim used to test for. However, it seems that in
3191 other systems, "." is yielded. Exim now tests for this case too.
3193 7. The values of check_spool_space and check_log_space are now held internally
3194 as a number of kilobytes instead of an absolute number of bytes. If a
3195 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3196 kilobyte. This means that much larger values can be stored.
3198 8. Exim monitor: an attempt to get the action menu when not actually pointing
3199 at a message produces an empty menu entitled "No message selected". This
3200 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3201 no entries in it ("Shell widget menu has zero width and/or height"). So I
3202 have added a single, blank menu entry in this case.
3204 9. Added ${quote_local_part.
3206 10. MIME decoding is now applied to the contents of Subject: header lines when
3209 11. Now that a reference to $sender_host_address automatically causes a reverse
3210 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3211 host lookup before query-style lookups in lists that might use this
3212 variable. This has therefore been abolished, and the "net-" prefix is no
3213 longer necessary for query-style lookups.
3215 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3216 have been a typo for LFLAGS, so it has been changed.
3218 13. The install script calls Exim with "-C /dev/null" in order to find the
3219 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3220 to be output. Howeve, since Exim outputs its version number before the
3221 error, it didn't break the script. It just looked ugly. I fixed this by
3222 always allowing "-C /dev/null" if the caller is root.
3224 14. Ignore overlarge ACL variable number when reading spool file - insurance
3225 against a later release with more variables having written the file.
3227 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3228 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3229 this, as well as the form without the "IPv6" on the front.
3231 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3232 OS/Makefile-Darwin file.
3234 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3235 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3237 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3238 variable. However, some OS use ints for inodes. I've added cast to long int
3239 to get rid of the compiler warning.
3241 19. I had forgotten to lock out "/../" in configuration file names when
3242 ALT_CONFIG_PREFIX was set.
3244 20. Routers used for verification do not need to specify transports. However,
3245 if such a router generated a host list, and callout was configured, Exim
3246 crashed, because it could not find a port number from the (non-existent)
3247 transport. It now assumes port 25 in this circumstance.
3249 21. Added the -t option to exigrep.
3251 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3252 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3253 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3254 them is set, LOOKUP_LSEARCH is forced.
3256 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3258 24. Added sender and host information to the "rejected by local_scan()" log
3259 line; previously there was no indication of these.
3261 25. Added .include_if_exists.
3263 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3264 Linux. It turns out that not all file systems support this. Apparently some
3265 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3266 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3267 not supported on the file descriptor, is now ignored when Exim is trying to
3268 sync a directory. This applies only to Linux.
3270 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3272 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3273 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3276 29. Added support for maildirsize files from supplied patch (modified a bit).
3278 30. The use of :fail: followed by an empty string could lead Exim to respond to
3279 sender verification failures with (e.g.):
3281 550 Verification failed for <xxx>
3282 550 Sender verify failed
3284 where the first response line was missing the '-' that indicates it is not
3285 the final line of the response.
3287 31. The loop for finding the name of the user that called Exim had a hardwired
3288 limit of 10; it now uses the value of finduser_retries, which is used for
3289 all other user lookups.
3291 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3294 33. Exim was neglecting to zero errno before one call of strtol() when
3295 expanding a string and expecting an integer value. On some systems this
3296 resulted in spurious "integer overflow" errors. Also, it was casting the
3297 result into an int without checking.
3299 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3300 did not work. The code looks as if it has *never* worked, though it appears
3301 to have been documented since at least releast 1.62. I have made it work.
3303 35. The "timeout_DNS" error in retry rules, also documented since at least
3304 1.62, also never worked. As it isn't clear exactly what this means, and
3305 clearly it isn't a major issue, I have abolished the feature by treating it
3306 as "timeout", and writing a warning to the main and panic logs.
3308 36. The display of retry rules for -brt wasn't always showing the error code
3311 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3312 timeout_connect_A, timeout_connect_MX.
3314 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3315 to the empty sender.
3317 39. The daemon was not analysing the content of -oX till after it had closed
3318 stderr and disconnected from the controlling terminal. This meant that any
3319 syntax errors were only noted on the panic log, and the return code from
3320 the command was 0. By re-arranging the code a little, I've made the
3321 decoding happen first, so such errors now appear on stderr, and the return
3322 code is 1. However, the actual setting up of the sockets still happens in
3323 the disconnected process, so errors there are still only recorded on the
3326 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3327 connections (as happens on some IP stacks) was logged at start up time as
3328 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3329 it from "IPv6 and IPv4", which means that two separate sockets are being
3332 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3333 says whether AF_INET or AF_INET6 was passed as an argument.
3335 42. Exiwhat output was messed up when time zones were included in log
3338 43. Exiwhat now gives more information about the daemon's listening ports,
3339 and whether -tls-on-connect was used.
3341 44. The "port" option of the smtp transport is now expanded.
3343 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3344 silently ignored. Now an error message is written to the main and panic
3347 46. There's a new ACL modifier called "logwrite" which writes to a log file
3348 as soon as it is encountered.
3350 47. Added $local_user_uid and $local_user_gid at routing time.
3352 48. Exim crashed when trying to verify a sender address that was being
3355 49. Exim was recognizing only a space character after ".include". It now also
3356 recognizes a tab character.
3358 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3359 extracting the relevant information from the specification. The man page no
3360 longer contains scrambled data for the -d option, and I've added a section
3361 at the front about calling Exim under different names.
3363 51. Added "extra_headers" argument to the "mail" command in filter files.
3365 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3368 53. Installed eximstats 1.29.
3370 54. Added transport_filter_timeout as a generic transport option.
3372 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3373 Cc: header lines. This was required by RFC 822, but it not required by RFC
3376 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3377 incoming messages that did not have them. Now it does so only if the
3378 message originates locally, that is, if there is no associated remote host
3379 address. When Resent- header lines are present, this applies to the Resent-
3380 lines rather than the non-Resent- lines.
3382 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3383 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3385 58. Messages for configuration errors now include the name of the main
3386 configuration file - useful now that there may be more than one file in a
3387 list (.included file names were always shown).
3389 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3390 for those rare installations that do not start the daemon as root or run it
3391 setuid root. I've cut out the call to initgroups() if the daemon is not
3394 60. The Exim user and group can now be bound into the binary as text strings
3395 that are looked up at the start of Exim's processing.
3397 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3399 62. Added $mailstore_basename variable.
3401 63. Installed patch to sieve.c from Michael Haardt.
3403 64. When Exim failed to open the panic log after failing to open the main log,
3404 the original message it was trying to log was written to stderr and debug
3405 output, but if they were not available (the usual case in production), it
3406 was lost. Now it is written to syslog before the two lines that record the
3407 failures to open the logs.
3409 65. Users' Exim filters run in subprocesses under the user's uid. It is
3410 possible for a "deliver" command or an alias in a "personal" command to
3411 provoke an address rewrite. If logging of address rewriting is configured,
3412 this fails because the process is not running as root or exim. There may be
3413 a better way of dealing with this, but for the moment (because 4.30 needs
3414 to be released), I have disabled address rewrite logging when running a
3415 filter in a non-root, non-exim process.
3421 1. The buildconfig auxiliary program wasn't quoting the value set for
3422 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3423 not defined. This bug was masked in 4.22 by the effect that was fixed in
3426 2. Some messages that were rejected after a message id was allocated were
3427 shown as "incomplete" by exigrep. It no longer does this for messages that
3428 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3430 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3431 have allow_domain_literals set, the ID did not get logged in the <= line.
3432 Domain literals are now always recognized in Message-ID: header lines.
3434 4. The first argument for a ${extract expansion item is the key name or field
3435 number. Leading and trailing spaces in this item were not being ignored,
3436 causing some misleading effects.
3438 5. When deliver_drop_privilege was set, single queue runner processes started
3439 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3440 same command in the process it spins off) were not dropping privilege.
3442 6. When the daemon running as "exim" started a queue runner, it always
3443 re-executed Exim in the spun-off process. This is a waste of effort when
3444 deliver_drop_privilege is set. The new process now just calls the
3445 queue-runner function directly.
3451 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3454 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3455 ignored. Though the use of -oP was forcing the writing of a pid file, it
3456 was always written to the default place.
3458 3. If the message "no IP address found for host xxxx" is generated during
3459 incoming verification, it is now followed by identification of the incoming
3460 connection (so you can more easily find what provoked it).
3462 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3464 5. Added some features to "harden" Exim a bit more against certain attacks:
3466 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3467 be put in Local/Makefile. This is like the never_users runtime option,
3468 but it cannot be overridden. The default setting is "root".
3470 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3471 prefix string with which any file named in a -C command line option
3474 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3475 is retained for -C and -D only if the caller of Exim is root. Without
3476 it, the exim user may also use -C and -D and retain privilege.
3478 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3479 command line option is disabled.
3481 6. Macro names set by the -D option must start with an upper case letter, just
3482 like macro names defined in the configuration file.
3484 7. Added "dereference=" facility to LDAP.
3486 8. Two instances of the typo "uknown" in the source files are fixed.
3488 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3489 the Configure-Makefile script screwed up while processing it.
3491 10. Incorporated PCRE 4.4.
3493 11. The SMTP synchronization check was not operating right at the start of an
3494 SMTP session. For example, it could not catch a HELO sent before the client
3495 waited for the greeting. There is now a check for outstanding input at the
3496 point when the greeting is written. Because of the duplex, asynchronous
3497 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3498 way, but not yet received, when the check is performed.
3500 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3501 on TCP/IP sockets, because this apparently causes some broken clients to
3504 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3505 unchanged) from the Cygwin maintainer.
3507 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3508 was supported instead of testing for mbx. Effectively a typo.
3510 15. The spa authenticator server code was not checking that the input it
3511 received was valid base64.
3513 16. The debug output line for the "set" modifier in ACLs was not showing the
3514 name of the variable that was being set.
3516 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3518 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3519 lookup on its own. Something else was needed to trigger the lookup. For
3520 example, a match in host_lookup or the need for a host name in a host list.
3521 Now, if $sender_host_name is referenced and the host name has not yet been
3522 looked up, a lookup is performed. If the lookup fails, the variable remains
3523 empty, and $host_lookup_failed is set to "1".
3525 19. Added "eqi" as a case-independent comparison operator.
3527 20. The saslauthd authentication condition could segfault if neither service
3528 nor realm was specified.
3530 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3531 error message that was logged was misleading, and incoming SMTP
3532 connections were dropped. The message is now more accurate, and temporary
3533 errors are given to SMTP connections.
3535 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3536 (except RSET) with a 421 error, until QUIT is received. However, it was
3537 failing to send a response to QUIT.
3539 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3540 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3541 ACL is now run after the helo_try_verify_hosts code.
3543 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3544 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3545 case-independent, so other case variants are also recognized). Apparently
3546 some systems use these upper case variants.
3548 25. If more than two messages were waiting for the same host, and a transport
3549 filter was specified for the transport, Exim sent two messages over the
3550 same TCP/IP connection, and then failed with "socket operation on non-
3551 socket" when it tried to send the third.
3553 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3555 27. The extern definition of crypt16() in expand.c was not being excluded when
3556 the OS had its own crypt16() function.
3558 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3559 as a preferred synonym for return_size_limit, both as an option and as an
3562 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3564 30. Changed the default configuration ACL to relax the local part checking rule
3565 for addresses that are not in any local domains. For these addresses,
3566 slashes and pipe symbols are allowed within local parts, but the sequence
3567 /../ is explicitly forbidden.
3569 31. SPA server authentication was not clearing the challenge buffer before
3572 32. log_message in a "warn" ACL statement was writing to the reject log as
3573 well as to the main log, which contradicts the documentation and doesn't
3574 seem right (because no rejection is happening). So I have stopped it.
3576 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3577 However, I am unable to do any testing of this.
3579 34. Fixed an infelicity in the appendfile transport. When checking directories
3580 for a mailbox, to see if any needed to be created, it was accidentally
3581 using path names with one or more superfluous leading slashes; tracing
3582 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3584 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3585 contents are added to the log line that is written for every discarded
3586 recipient. (Previously a log_message setting was ignored.)
3588 36. The ${quote: operator now quotes the string if it is empty.
3590 37. The install script runs exim in order to find its version number. If for
3591 some reason other than non-existence or emptiness, which it checks, it
3592 could not run './exim', it was installing it with an empty version number,
3593 i.e. as "exim-". This error state is now caught, and the installation is
3596 38. An argument was missing from the function that creates an error message
3597 when Exim fails to connect to the socket for saslauthd authentication.
3598 This could cause Exim to crash, or give a corrupted message.
3600 39. Added isip, isip4, and isip6 to ${if conditions.
3602 40. The ACL variables $acl_xx are now saved with the message, and can be
3603 accessed later in routers, transports, and filters.
3605 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3606 strings in the file are not string-expanded.
3608 42. If a MAIL command specified a SIZE value that was too large to fit into an
3609 int variable, the check against message_size_limit failed. Such values are
3610 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3611 day this will have to be increased, but I don't think I want to be around
3612 when emails are that large.
3619 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3620 iconv() is not standard in FreeBSD.
3622 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3623 IPv6 enabled. The observed symptom was a segmentation fault on return from
3624 the function os_common_find_running_interfaces() in src/os.c.
3626 3. In the check_special_case() function in daemon.c I had used "errno" as an
3627 argument name, which causes warnings on some systems. This was basically a
3628 typo, since it was named "eno" in the comments!
3630 4. The code that waits for the clock to tick (at a resolution of some fraction
3631 of a second) so as to ensure message-id uniqueness was always waiting for
3632 at least one whole tick, when it could have waited for less. [This is
3633 almost certainly not relevant at current processor speeds, where it is
3634 unlikely to ever wait at all. But we try to future-proof.]
3636 5. The function that sleeps for a time interval that includes fractions of a
3637 second contained a race. It did not block SIGALRM between setting the
3638 timer, and suspending (a couple of lines later). If the interval was short
3639 and the sigsuspend() was delayed until after it had expired, the suspension
3640 never ended. On busy systems this could lead to processes getting stuck for
3643 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3644 process, before it forks any delivery processes. The open lookup caching
3645 mechanism meant that the open file or database connection was passed into
3646 the delivery process. The problem was that delivery processes always tidy
3647 up cached lookup data. This could cause a problem for the next delivery
3648 process started by the queue runner, because the external queue runner
3649 process does not know about the closure. So the next delivery process
3650 still has data in the lookup cache. In the case of a file lookup, there was
3651 no problem because closing a file descriptor in a subprocess doesn't affect
3652 the parent. However, if the lookup was caching a connection to a database,
3653 the connection was closed, and the second delivery process was likely to
3654 see errors such as "PGSQL: query failed: server closed the connection
3655 unexpectedly". The problem has been fixed by closing all cached lookups
3656 in a queue runner before running a delivery process.
3658 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3659 seem to have the "const" qualifier which it has on other OS. I've
3662 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3663 *of the same type* (client or server) with the same public name that an
3664 error should be diagnosed.
3666 9. When Exim looked up a host name for an IP address, but failed to find the
3667 original IP address when looking up the host name (a safety check), it
3668 output the message "<ip address> does not match any IP for NULL", which was
3669 confusing, to say the least. The bug was that the host name should have
3670 appeared instead of "NULL".
3672 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3673 user that is built into the binary, and the -C or -D options is used, root
3674 privilege is dropped before the configuration file is read. In addition,
3675 logging is switched to stderr instead of the normal log files. If the
3676 configuration then re-defines the Exim user, the unprivileged environment
3677 is probably not what is expected, so Exim logs a panic warning message (but
3680 However, if deliver_drop_privilege is set, the unprivileged state may well
3681 be exactly what is intended, so the warning has been cut out in that case,
3682 and Exim is allowed to try to write to its normal log files.
3688 1. smtp_return_error_details was not giving details for temporary sender
3689 or receiver verification errors.
3691 2. Diagnose a configuration error if two authenticators have the same public
3694 3. Exim used not to create the message log file for a message until the first
3695 delivery attempt. This could be confusing when incoming messages were held
3696 for policy or load reasons. The message log file is now created at the time
3697 the message is received, and an initial "Received" line is written to it.
3699 4. The automatically generated man page for command line options had a minor
3700 bug that caused no ill effects; however, a more serious problem was that
3701 the procedure for building the man page automatically didn't always
3702 operate. Consequently, release 4.20 contains an out-of-date version. This
3703 shouldn't happen again.
3705 5. When building Exim with embedded Perl support, the script that builds the
3706 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3707 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3709 6. The freeze_tell option was not being used for messages that were frozen on
3710 arrival, either by an ACL or by local_scan().
3712 7. Added the smtp_incomplete_transaction log selector.
3714 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3715 was accepting AUTH without a new EHLO.
3717 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3718 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3721 10. Logging of TCP/IP connections (when configured) now happens in the main
3722 daemon process instead of the child process, so that the TCP/IP connection
3723 count is more accurate (but it can never be perfect).
3725 11. The use of "drop" in a nested ACL was not being handled correctly in the
3726 outer ACL. Now, if condition failure induced by the nested "drop" causes
3727 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3728 or "require"), the connection is dropped.
3730 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3731 that yield "discard" can now be used with an "accept" or a "discard" verb,
3732 but an error is generated for any others (because I can't see a useful way
3733 to define what should happen).
3735 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3736 are now processed in the same way as lines in the Exim configuration file.
3737 In particular, continuation lines are supported.
3739 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3741 15. Added -ti meaning -t -i.
3743 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3744 domains, and warn by logging if others are found.
3746 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3747 returned from GIFCONF when Exim is trying to find the list of interfaces on
3748 a host. The code in os.c has been modified to copy each ifreq to an aligned
3749 structure in all cases.
3751 Also, in some cases, the returned ifreq's were being copied to a 'struct
3752 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3753 means the last couple of bytes of an IPv6 address could be chopped if the
3754 ifreq contained only a normal sockaddr (14 bytes storage).
3756 18. Named domain lists were not supported in the hosts_treat_as_local option.
3757 An entry such as +xxxx was not recognized, and was treated as a literal
3760 19. Ensure that header lines added by a DATA ACL are included in the reject log
3761 if the ACL subsequently rejects the message.
3763 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3764 MD5 (which is deprecated).
3766 21. When testing a filter file using -bf, Exim was writing a message when it
3767 took the sender from a "From " line in the message, but it was not doing so
3768 when it took $return_path from a Return-Path: header line. It now does.
3770 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3771 with a valid header line field name (a series of printing characters
3772 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3774 23. Changed "disc" in the source to "disk" to conform to the documentation and
3775 the book and for uniformity.
3777 24. Ignore Sendmail's -Ooption=value command line item.
3779 25. When execve() failed while trying to run a command in a pipe transport,
3780 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3781 could be confused with a return value of 69 from the command itself. This
3782 has been changed to 127, the value the shell returns if it is asked to run
3783 a non-existent command. The wording for the related log line suggests a
3784 non-existent command as the problem.
3786 26. If received_header_text expands to an empty string, do not add a Received:
3787 header line to the message. (Well, it adds a token one on the spool, but
3788 marks it "old" so that it doesn't get used or transmitted.)
3790 27. Installed eximstats 1.28 (addition of -nt option).
3792 28. There was no check for failure on the call to getsockname() in the daemon
3793 code. This can fail if there is a shortage of resources on the system, with
3794 ENOMEM, for example. A temporary error is now given on failure.
3796 29. Contrary to the C standard, it seems that in some environments, the
3797 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3798 program. Exim now does this explicitly; it affects the formatting of
3799 timestamps using strftime().
3801 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3802 complaints. I've now initialized all the variables, to avoid this.
3804 32. Header lines added by a system filter were not being "seen" during
3805 transport-time rewrites.
3807 33. The info_callback() function passed to OpenSSL is set up with type void
3808 (*)(SSL *, int, int), as described somewhere. However, when calling the
3809 function (actually a macro) that sets it up, the type void(*)() is
3810 expected. I've put in a cast to prevent warnings from picky compilers.
3812 34. If a DNS black list lookup found a CNAME record, but there were no A
3813 records associated with the domain it pointed at, Exim crashed.
3815 35. If a DNS black list lookup returned more than one A record, Exim ignored
3816 all but the first. It now scans all returned addresses if a particular IP
3817 value is being sought. In this situation, the contents of the
3818 $dnslist_value variable are a list of all the addresses, separated by a
3821 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3822 to accept a host name and all its aliases if the forward lookup for any of
3823 them yielded the IP address of the incoming connection. Now it accepts only
3824 those names whose forward lookup yields the correct IP address. Any other
3825 names are discarded. This closes a loophole whereby a rogue DNS
3826 administrator could create reverse DNS records to break through a
3827 wildcarded host restriction in an ACL.
3829 37. If a user filter or a system filter that ran in a subprocess used any of
3830 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3831 the wrong values were passed to the pipe command ($thisaddress had the
3832 value of $0, $0 had the value of $1, etc). This bug was introduced by
3833 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3835 38. Improved the line breaking for long SMTP error messages from ACLs.
3836 Previously, if there was no break point between 40 and 75 characters, Exim
3837 left the rest of the message alone. Two changes have been made: (a) I've
3838 reduced the minimum length to 35 characters; (b) if it can't find a break
3839 point between 35 and 75 characters, it looks ahead and uses the first one
3840 that it finds. This may give the occasional overlong line, but at least the
3841 remaining text gets split now.
3843 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3844 file descriptors might be low, and that setting 1000 would always raise it.
3845 It turns out that in some environments, the limit is already over 1000 and
3846 that lowering it causes trouble. So now Exim takes care not to decrease it.
3848 40. When delivering a message, the value of $return_path is set to $sender_
3849 address at the start of routing (routers may change the value). By an
3850 oversight, this default was not being set up when an address was tested by
3851 -bt or -bv, which affected the outcome if any router or filter referred to
3854 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3855 log only when "message" or "log_message" are set. However, if one of the
3856 conditions was an address verification, or a call to a nested ACL, the
3857 messages generated by the underlying test were being passed through. This
3858 no longer happens. The underlying message is available in $acl_verify_
3859 message for both "message" and "log_message" expansions, so it can be
3860 passed through if needed.
3862 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3863 new expansion $bh_ to give the encoded byte string without charset
3864 translation. Translation happens only if iconv() is available; HAVE_ICONV
3865 indicates this at build time. HEADERS_CHARSET gives the charset to
3866 translate to; headers_charset can change it in the configuration, and
3867 "headers charset" can change it in an individual filter file.
3869 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3870 that creates RFC 2047 encoded "words" labels them as that charset instead
3871 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3872 expansion operator; (ii) when Exim creates a From: line for a local
3873 message; (iii) when a header line is rewritten to include a "phrase" part.
3875 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3876 buggy, causing it to skip the first lines of messages whose message ID
3877 ended in 'D'. This would not have bitten before Exim release 4.14, because
3878 message IDs were unlikely to end in 'D' before then. The effect was to have
3879 incorrect size information for certain domains.
3881 45. #include "config.h" was missing at the start of the crypt16.c module. This
3882 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3885 46. If there was a timeout during a "random" callout check, Exim treated it as
3886 a failure of the random address, and carried on sending RSET and the real
3887 address. If the delay was just some slowness somewhere, the response to the
3888 original RCPT would be taken as a response to RSET and so on, causing
3889 mayhem of various kinds.
3891 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3892 when I implemented it. It didn't allow for the fact that some option values
3893 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3894 the right test for positive values.
3896 48. Domain names in DNS records are case-independent. Exim always looks them up
3897 in lower case. Some resolvers return domain names in exactly the case they
3898 appear in the zone file, that is, they may contain uppercase letters. Not
3899 all resolvers do this - some return always lower case. Exim was treating a
3900 change of case by a resolver as a change of domain, similar to a widening
3901 of a domain abbreviation. This triggered its re-routing code and so it was
3902 trying to route what was effectively the same domain again. This normally
3903 caused routing to fail (because the router wouldn't handle the domain
3904 twice). Now Exim checks for this case specially, and just changes the
3905 casing of the domain that it ultimately uses when it transmits the message
3908 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3911 50. If a filter generated a file delivery with a non-absolute name (possible if
3912 no home directory exists for the router), the forbid_file option was not
3915 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3916 the existing equality matching.
3918 52. Exim was using ints instead of ino_t variables in some places where it was
3919 dealing with inode numbers.
3921 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3922 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3923 called TMPDIR, and if it finds it is different, it changes its value.
3925 54. The smtp_printf() function is now made available to local_scan() so
3926 additional output lines can be written before returning. There is also an
3927 smtp_fflush() function to enable the detection of a dropped connection.
3928 The variables smtp_input and smtp_batched_input are exported to
3931 55. Changed the default runtime configuration: the message "Unknown user"
3932 has been removed from the ACL, and instead placed on the localuser router,
3933 using the cannot_route_message feature. This means that any verification
3934 failures that generate their own messages won't get overridden. Similarly,
3935 the "Unrouteable address" message that was in the ACL for unverifiable
3936 relay addresses has also been removed.
3938 56. Added hosts_avoid_esmtp to the smtp transport.
3940 57. The exicyclog script was not checking for the esoteric option
3941 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3942 will work only if exicyclog is run under the appropriate euid.
3944 58. Following a discussion on the list, the rules by which Exim recognises line
3945 endings on incoming messages have been changed. The -dropcr and drop_cr
3946 options are now no-ops, retained only for backwards compatibility. The
3947 following line terminators are recognized: LF CRLF CR. However, special
3948 processing applies to CR:
3950 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3951 nor a local message in the state where . is a terminator.
3953 (ii) If a bare CR is encountered in a header line, an extra space is added
3954 after the line terminator so as not to end the header. The reasoning
3955 behind this is that bare CRs in header lines are most likely either
3956 to be mistakes, or people trying to play silly games.
3958 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3959 was being incorrectly given as 18 bytes larger than it should have been.
3960 This is a VOB (very old bug).
3962 60. This may never have affected anything current, but just in case it has:
3963 When the local host is found other than at the start of a list of hosts,
3964 the local host, those with the same MX, and any that follow, are discarded.
3965 When the list in question was part of a longer list of hosts, the following
3966 hosts (not currently being processed) were also being discarded. This no
3967 longer happens. I'm not sure if this situation could ever has previously
3970 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3973 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3974 header. This is something that is recommended in a new Internet Draft, and
3975 is something that is documented as being done by Sendmail. There are two
3976 possible values. For messages generated by the autoreply transport, Exim
3979 Auto-Submitted: auto-replied
3981 whereas for all other generated messages (e.g. bounces) it adds
3983 Auto-Submitted: auto-generated
3985 63. The "personal" condition in filters now includes a test for the
3986 Auto-Submitted: header. If it contains the string "auto-" the message it
3987 not considered personal.
3989 64. Added rcpt_include_affixes as a generic transport option.
3991 65. Added queue_only_override (default true).
3993 66. Added the syslog_duplication option.
3995 67. If what should have been the first header line of a message consisted of
3996 a space followed by a colon, Exim was mis-interpreting it as a header line.
3997 It isn't of course - it is syntactically invalid and should therefore be
3998 treated as the start of the message body. The misbehaviour could have
3999 caused a number of strange effects, including loss of data in subsequent
4000 header lines, and spool format errors.
4002 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4003 client host had authenticated. This control can now be exercised by an ACL
4004 for more flexibility.
4006 69. By default, callouts do not happen when testing with -bh. There is now a
4007 variant, -bhc, which does actually run the callout code, including
4008 consulting and updating the callout cache.
4010 70. Added support for saslauthd authentication, courtesy of Alexander
4013 71. If statvfs() failed on the spool or log directories while checking their
4014 size for availability, Exim confusingly gave the error "space shortage".
4015 Furthermore, in debugging mode it crashed with a floating point exception.
4016 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4017 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4018 serious problem, Exim now writes to the main and panic logs when this
4019 happens, with details of the failure. It then refuses to accept the
4020 incoming message, giving the message "spool directory problem" or "log
4021 directory problem" with a 421 code for SMTP messages.
4023 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4024 0, 1, and 2 exist, because some OS complain for execs without them (see
4025 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4026 descriptors. However, the code omitted to check that the open succeeded,
4027 causing mysterious errors if for some reason the permissions on /dev/null
4028 got screwed. Now Exim writes a message to the main and panic logs, and
4029 bombs out if it can't open /dev/null.
4031 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4032 interact so that it is all more flexible. It is supposed to remain
4033 backwards compatible. Also added extra_local_interfaces.
4035 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4036 to bomb out with an assertion failure - to the client this appears as a
4037 connection drop. This problem occurs in the part of the code that was taken
4038 from the Samba project. Fortunately, the assertion is in a very simple
4039 function, so I have fixed this by reproducing the function inline in the
4040 one place where it is called, and arranging for authentication to fail
4041 instead of killing the process with assert().
4043 75. The SPA client code was not working when the server requested OEM rather
4044 than Unicode encoding.
4046 76. Added code to make require_files with a specific uid setting more usable in
4047 the case where statting the file as root fails - usually a non-root-mounted
4048 NFS file system. When this happens and the failure is EACCES, Exim now
4049 forks a subprocess and does the per-uid checking as the relevant uid.
4051 77. Added process_log_path.
4053 78. If log_file_path was not explicitly set, a setting of check_log_space or
4054 check_log_inodes was ignored.
4056 79. If a space check for the spool or log partitions fails, the incident is now
4057 logged. Of course, in the latter case the data may get lost...
4059 80. Added the %p formatting code to string_format() so that it can be used to
4060 print addresses in debug_print(). Adjusted all the address printing in the
4061 debugging in store.c to use %p rather than %d.
4063 81. There was a concern that a line of code in smtp_in.c could overflow a
4064 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4065 initially expressed, the concern was not well-founded, because trailing
4066 spaces are removed early. However, if the trailing spaces were followed by
4067 a NULL, they did not get removed, so the overflow was possible. Two fixes
4070 (a) I re-wrote the offending code in a cleaner fashion.
4071 (b) If an incoming SMTP command contains a NULL character, it is rejected
4074 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4075 runs initgroups(), so that if the Exim user is in any additional groups,
4076 they will be used during message reception.
4082 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob